Effect of Physical Weathering on - DiVA...

DOCTORA L T H E S I S

Department of Civil, Environmental and Natural Resources EngineeringDivision of Mining and Geotechnical Engineering

Effect of Physical Weathering on Mechanical Properties of Tailings

Juan Rodríguez

ISSN 1402-1544ISBN 978-91-7583-735-2 (print)ISBN 978-91-7583-736-9 (pdf)

Luleå University of Technology 2016

Juan Rodríguez E

ffect of Physical Weathering on M

echanical Properties of Tailings

Soil Mechanics

EFFECT OF PHYSICAL WEATHERING ON MECHANICAL PROPERTIES OF

TAILINGS

DOCTORAL THESIS SUBMITTED BY

JUAN MANUEL RODRÍGUEZ ZAVALA

DEPARTEMENT OF CIVIL, ENVIRONMENTAL AND NATURAL RESOURCES ENGINEERING

DIVISION OF MINING AND GEOTECHNICAL ENGINEERING LULEÅ UNIVERSITY OF TECHNOLOGY

IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY IN

SOIL MECHANICS AND FUNDATION ENGINEERING

Printed by Luleå University of Technology, Graphic Production 2016

ISSN 1402-1544 ISBN 978-91-7583-735-2 (print)ISBN 978-91-7583-736-9 (pdf)

Luleå 2016

www.ltu.se

iii

DEDICATORIA

Estos años que mi familia y yo hemos estado en tierra vikinga, extraña en un inicio para nosotros de climas cálidos, nos han dado mucho aprendizaje. Aprendizaje cultural adaptándonos a las actividades diarias como cualquier familia Sueca, incluso disfrutando del invierno; aprendizaje en nuestro núcleo familiar que se ha hecho más fuerte al estar alejados del resto de la familia que añoramos.

Como en toda gran historia no podían faltar los grandes amigos que hemos hecho, con

ellos hemos viajado, convivido, reído, llorado y han hecho de Suecia también nuestro hogar, tengo en lo personal mucho que agradecerles. A quien más agradezco por estar siempre a mi lado, creciendo como familia es a mi…

Esposa Ada

Por el amor, paciencia y soporte

Hija Vielka

Por la satisfacciones que “mi pequeña” me da todos los días

v

DEDICATORY My family has been living for the last years in Viking’s land from where we have learned a lot; it was strange in the beginning for this “warm climate” citizens. Learning about Swedish culture and adapting to the daily activities like any local family; even enjoying the cold Swedish winter. Family’s matures becoming stronger been away from the rest of the relatives we love. As in all great stories there could not miss the great friends we've made, we've traveled, shared, laughed, cried and they have made of Sweden also our home. I personally thank you very much. To whom I most thank to be on my side maturing as a family is to my …

Wife Ada

For the love, patience and support

Daughter Vielka

For the joy my ‘little love’ gives me everyday

vii

ACKNOWLEDGMENT I would like to acknowledge everyone that has assisted me throughout my doctoral studies over the years. I would like to thanks first and specially to my supervisors, Professor Sven Knutsson and Dr. Tommy Edeskär for them patience and feedback I always received from them. The dissertation would have been not possible without them guidance. Very special thanks to Boliden Mineral AB and its operations in AITIK from where tailings were obtained to be used during this research. I appreciate everyday conversations and suggestions for all my colleges and laboratory technicians it has improve undoubtedly my work and make it more enjoyable. I am very grateful to the Civil and Mining Department at the Engineering Division, University of Sonora and also to the Division of Mining and Geotechnical Engineering at the Department of Civil, Environmental and Natural Resources Engineering, Luleå University of Technology for the scholarship, financial support and especially for the opportunity to carry out this research. I would like to thanks also the staff of Luleå University of Technology. Finally to whom I most love and dedicate this thesis… my family. Juan Rodriguez November, 2016 Luleå, Sweden.

ix

ABSTRACT The mining activity produces minerals to supply to the modern society with commodities. During the ore refining process large amounts of tailings are generated as waste. Tailings are the result of crushing and wet milling, typically in the size range of sand to silt and angular in shape. Tailings are in general deposited in tailings dams for permanent storage. Tailings dams are usually considered as walk-away solutions and needs to be designed and constructed to be safe in a long time perspective. Several incidents around the world have occurred in tailings dams both during construction and operation and after closure of the activities. The consequences of the failures may be fatal to the local society and harm the surrounding environment. Considering the consequences of failures and relatively few studies on tailings properties in a long term perspective there is a need for research. As a consequence of the operation and raising procedures of tailings dams, the conditions in the tailings dams could be considered to be dynamic in a longer perspective. Grain size distribution, formation of layers, pore pressures and stress states are continuously changing during the operation. Tailings may be susceptible to weathering in the deposit environment. The change of these factors needs to be addressed in the design of walk-away solutions. In this work image analysis, oedometer-, triaxial-, direct shear- and attrition tests has been carried out to study the tailings particle influence on fundamental geotechnical properties in a case study. The parameters have been compared to similar-sized natural materials. The effect of loading and physical weathering has been studied and evaluated by image analysis and sieving. The comparative test by natural materials shows that tailings, probably due to the irregular shape, initially arranges in looser interparticle structures. The looser arrangement makes tailings fills more susceptible to settlement compared to natural deposits in the same size range. The two most identified factors affecting the tailings on a particle level was the type of physical weathering and grain size. Static load, shearing and milling decreases the grain size but the effect on the individual particles were different. Large grains tend to maintain the shape or get rounded by physical degradation and finer grains get more angular by milling but rounded by shearing. An attempt has been done to predict the effect on friction angle as a function of angularity, grain size and initial void ratio. The increase in angularity in the model suggests an increase in the friction angle and so the increase of regularity (decrease of elongation relation). However the reduction of size will either reduces or compensates this effect. More studies are needed to verify this.

xi

SAMMANFATTNING Gruv industrin producerar mineralråvaror till det moderna samhället. Under malmens utvinningsprocess genereras stora mängder som en restprodukt. Anrikningssanden är ett resultat av krossning och våtmalning; materialet har en kantig form och består av partiklar som tillhör fraktionerna silt och sand. Allmänhet deponeras anrikningsand i gruvdammar för permanent lagring. Gruvdammar brukar betraktas som walk-away-lösningar och måste således vara konstruerade så att de är säkra i ett långtidsperspektiv. Flera incidenter relaterade till gruvdammar har inträffat runt om i världen; både under uppförande och drift, samt efter avslutad verksamhet. Konsekvenserna vid ett dambrott kan vara allvanligt för samhället med hänsyn till förlorade liv samt skada på den omgivande miljön. Med tanke på de allvarliga konsekvenserna vid dammbrott och relativt få studier om anrikningssanden egenskaper i ett långsiktigt perspektiv finns det ett behov av forskning inom detta område. Som en följd av ökade mängder av restprodukten samt höjningen av gruvdammar, kan förhållandena anses vara dynamisk i ett längre perspektiv. Kornstorleksfördelning, bildandet av skikt, portryck och spänningstillstånd förändras kontinuerligt under en gruvdams uppförande och drift. Vid lagring i gruvdammar kan den deponerade anrikningssanden vara känslig för vittring. Hänsyn tas till anrikningssandens med tiden förändrade egenskaper måste tas vid utformning av gruvdammar. I detta arbete har bildanalys, öedometer-, triaxial- och direkt skjuvförsök och nötningstest utförts för att studera hur partiklar i anrikningssand inverkar på grundläggande geotekniska egenskaper. Geotekniska egenskaper för anrikningssanden och naturliga material med samma kornstorleksfördelning har jämförts med varandra. Effekten av belastning och fysisk vittring har studerats och utvärderats genom bildanalys och siktning. De jämförande försöken av naturmaterial visar att anrikningssand, troligen på grund av oregelbunden form, först lagras i en lösare partikelstrukturer. Den lösare lagringen är orsaken till att fyllningar av anrikningssand är mer sättningskänsliga jämfört med fyllningar av naturligt material. De två faktorer som mest påverkar anrikningssand på partikelnivå är fysisk vittring och partiklarnas storlek. Statisk belastning, skjuvning och malning minskar kornstorleken dock beror effekten på de enskilda partiklarna på kornstorleken. Stora korn tenderar att bibehålla formen eller avrundas genom fysisk nedbrytning; finare korn blir kantigare vid malning och mer rundade vid skjuvning. Ett försök har gjorts för att förutsäga påverkan på friktionsvinkeln som en funktion av kantighet, kornstorlek och initiala portalet. En ökning av partiklarnas kantighet i modellen antyder en ökning av friktionsvinkeln liksom en ökning av regelbundenheten. (minskning töjning förhållande). Reduktion av partiklarnas storlek kommer varken reducera eller kompenser för denna effekt. Det behövs fler studier för att verifiera detta.

xiii

TABLE OF CONTENTS

PART I

1 INTRODUCTION .................................................................................................................. 1

1.1 Aim and Scope of Work / objectives and achievements of the work .......................... 3

1.2 Thesis outline ............................................................................................................... 4

2 PARTICLE SHAPE ............................................................................................................... 7

2.1 The particle shape ........................................................................................................ 7

2.1.1 Background .......................................................................................................... 7

2.1.2 Scale dependence classification ........................................................................... 7

2.2 Measurement techniques ............................................................................................. 8

2.3 Arrangement and contact of the particles .................................................................... 9

2.3.1 The arrangement of particles ................................................................................ 9

2.3.2 The inter-particle contact ................................................................................... 10

2.3.3 Breakage ............................................................................................................. 11

2.4 The influence of the shape ......................................................................................... 11

2.4.1 Void ratio and porosity ....................................................................................... 12

2.4.2 Angle of repose .................................................................................................. 13

2.4.3 Shear strength ..................................................................................................... 13

2.4.4 Sedimentation ..................................................................................................... 14

2.4.5 Hydraulic conductivity, Permeability ................................................................ 15

2.4.6 Liquefaction ....................................................................................................... 15

2.4.7 Groundwater and seepage modeling .................................................................. 16

3 MATERIALS AND METHODS – EXPERIMENTAL PART ........................................ 17

3.1 Tailings location ........................................................................................................ 17

3.2 The ore and tailings composition ............................................................................... 17

3.3 The quantities or shape descriptors ........................................................................... 17

3.4 The tailing sampling and samples ............................................................................. 18

3.5 Image acquisition and Analysis ................................................................................. 19

3.5.1 Statistics ............................................................................................................. 19

3.6 The laboratory tests on tailing properties. ................................................................. 20

3.6.1 Triaxial (Paper III) ............................................................................................. 21

xiv

3.6.2 Oedometer (Paper IV) ........................................................................................ 21

3.6.3 Wearing/attrition/degradation by mill (Paper V) ............................................... 21

3.6.4 Direct shear (Paper VI) ...................................................................................... 21

4 RESULTS .............................................................................................................................. 23

4.1 Triaxial test and friction angle empirical relations .................................................... 23

4.2 Vertical load effects ................................................................................................... 27

4.3 Ball and autogenous milling ...................................................................................... 30

4.4 Shear strength in uniformed sized particle ................................................................ 33

4.4.1 Suggested empirical relation .............................................................................. 36

5 DISCUSSION ....................................................................................................................... 39

6 SUMMARY AND MAIN CONCLUSIONS ...................................................................... 43

7 SUGGESTED FURTHER WORK ..................................................................................... 45

8 BIBLIOGRAPHY ................................................................................................................ 47

APPENDIX

PART II – APPENDED PAPERS

I. Particle Shape Quantities and Measurement Techniques – A Review. Rodriguez, J.M.; Edeskär, T. and Knutsson, S. Electronical Journal of Geotechnical Engineering. Vol. 18 (A), 2013, 169-198.

II. Particle Shape Determination by Two-Dimensional Image Analysis in

Geotechnical Engineering. Rodriguez, J.M.; Johansson, J.M.A. and Edeskär, T. Danish Geotechnical Society. (2012) DGF Bulletin 27, 207-218. Proceedings of the 16th Nordic Geotechnical Meeting. Copenhagen, 9-12 May 2012, 207-218.

III. Case of Study on Particle Shape and Friction Angle on Tailings.

Rodriguez, J.M. and Edeskär T. Journal of Advanced Science and Engineering Research. Vol. 3 (4) 2013, 373-387.

IV. Effect of Vertical Load on Tailings particles.

Rodriguez, J.M.; Bhanbhro, R.; Edeskär, T. and Knutsson, S. Journal of Earth Science and Geotechnical Engineering. Vol. 6, No. 2, pp. 115-129. 2016. ISSN: 1792-9040 (printed) 1792-9660 (online). Scienpress Ltd, 2016.

xv

V. Mechanical weathering effect on tailings particles Rodriguez, J.M.; Edeskär, T. and Knutsson, S. Proceedings of to the 17th Nordic Geotechnical Meeting (NGM2016). Reykjavik, Iceland. 25th - 28th of May 2016. ISBN 978-9935-24-002-6

VI. Shear strength in uniformed sized tailings particles

Rodriguez, J.M.; Bhanbhro, R.; Edeskär, T. and Knutsson, S. To be submitted to Journal of Earth Science.

VII. Basic description of tailings from Aitik focusing on mechanical behavior

Bhanbhro, R.; Knutsson, R.; Rodriguez, J.M.; Edeskär, T. and Knutsson, S. International Journal of Emerging Technology and Advanced Engineering. Vol. 3, No. 12, pp. 65-69. 2013. ISSN: 2250-2459.

VIII. Evaluation of primary and secondary deformations and particle breakage of

tailings Bhanbhro, R.; Rodriguez, J.M.; Edeskär, T. and Knutsson, S. Pan-American conference of Soil Mechanics and Geotechnical Engineering., 15-18 November, 2015. Buenos Aires, Argentina. Ed. Diego Manzanal; Alejo O. Sfriso, IOS Press, 2015, pp. 2481-2488.

1

1 INTRODUCTION The first mining vestiges can be found as early as in the paleolithic era and it is known that mining appears and develop as the civilization does (Armengot, et al., 2006). From the industrial revolution to the modern world the technology had advance enabling to the mining industry to increase maybe exponentially the amount of ore extracted, even at lower grades than before. The mined ore rock is crushed and milled to liberate and concentrate the precious mineral from the host rock; the valueless rock debris left after wet concentration is the so called tailings and since ore extraction increases also tailings do. Typical amount of tailing materials produced by the ore extraction is 43% for iron ore (EPA, 1994) and 99% for copper ore (Northey, et al., 2014). To store tailings, ‘tailings dams’ are constructed and they widely vary in storage capacity, coverage area or height. The highest constructed tailings dams are up to 100 meters (Rico, et al., 2008) and covering areas rounding the 13 km2 (e.g. Aitik Tailings dam, Sweden and Mission Tailings dam, USA). Historically incidents as Aznalcollar (Spain) in 1998, Baia Mare (Romania) and Aitik (Sweden) both in year 2000 and the public opinion had encouraged to the representatives involved to work in the prevention instead of react after an incident. Tailings dams are usually considered as walk-away solutions and needs to be designed and constructed to be safe in a long time perspective. The consequences of the failures may be fatal to the local society and harm the surrounding environment. Tailings mechanical properties are site specific and from an engineering perspective in general the strength and deformation properties are regarded as a natural soil material in the same size range and size distribution. Tailings materials range usually from sand to silt and angular shaped (Garga, et al., 1984) due to the mechanical process involved but strongly dependent in the genesis itself as rock structure, mineralogy and hardness among others (Wentworth, 1922b). Tailings dams should be constructed to provide structural and environmental safety for long time perspectives. Considering the consequences of failures and relatively few studies on tailings properties in a long term perspective there is a need for research. In order to design safe tailing dams in a long time perspective it could be suggested not only regard in current material properties but also in further changes due to e.g. weathering/aging, particularly when the tailings are used as a construction material to rise up the structure e.g. the upstream method. As a consequence of the operation and raising procedures of tailings dams, the conditions in the tailings dams could be considered to be dynamic in a longer perspective. Grain size distribution, formation of layers, pore pressures and stress states are continuously changing

2

during the operation (Ormann, et al., 2013). Tailings may be susceptible to weathering in the deposit environment. The change of these factors needs to be addressed in the design of walk-away solutions. Weathering could be measured by determining the particle’s shape change and by studying the shape a prognosis of the properties of a tailing deposit could be developed. For this study a literature review on particles shape and measuring techniques has been compiled in a report written by Rodriguez (2012). The report identified 42 shape descriptors or quantities. However, 32 of the quantities are applicable for two dimension analysis and it was possible to used 12 shape descriptors with image analysis due to the availability of the algorithms and software. During this research the image analysis in two dimensions was adopted due to the advantages it provides. Image analysis evaluation on the effects of the resolution and magnification over the quantities were performed. Results have shown that resolution is an important factor to consider due to the effect it was over the quantities (see paper II). Laboratory tests in this work includes oedometer-, triaxial-, direct shear- and attrition tests to study the tailings particle influence on fundamental geotechnical properties. The effect of loading and physical weathering has been studied and evaluated by image analysis and sieving. The two most identified factors affecting the tailings on a particle level was the type of physical weathering and grain size. Static load, shearing and milling decreases the grain size but the effect on the individual particles were different. Large grains tend to maintain the shape or get rounded by physical degradation and finer grains get more angular. Comparison with similar sized natural materials shows that tailings, probably due to the irregular shape, initially arranges in looser interparticle structures. The looser arrangement makes tailings fills more susceptible to settlement compared to natural deposits in the same size range. An attempt has been done to predict the effect on friction angle as a function of morphology/angularity, initial consolidation load, grain size and initial void ratio. For angularity/morphology different quantities were evaluated. The reduction of size will in the model suggest a decrease or increase in the friction angle depending in the used quantity. This effect is probably related and explained by the mica minerals present in the samples that could create weak zones ideal to slide. However, the increase in angularity in the fine grain sizes will either reduces or compensates this effect. More studies are needed to verify this.

3

1.1 Aim and Scope of Work / objectives and achievements of the work

There are two main objectives for the work performed within this thesis:

1.- Investigate the effect on degradation/weathering on tailing materials. Properties of special importance are the particle shape change, void ratio and the breakage generated. This research comprises the study of tailing particles separated by sizes to determine the shape change and breakage under vertical loads, wearing by rolling using a laboratory mill and by direct shear. Results are compared to understand the effects on each partial fraction. 2.- Evaluate the shape descriptors available in image analysis to measure the degradation by shape change of the tailing particles. Shape measurement techniques were evaluated to determine the best available practice, advantages and suitability for the actual research. Image analysis in 2 dimensions as the best method available under the research conditions was tested. The study was designed to understand the aging of the tailing particles and the possible effects on the dam stability using image analysis and regarding on the following questions:

1. What is the effect of the vertical load in the tailing particles? 2. Degradation agents (load, shearing and wearing) act in a different way? 3. Is it possible to apply empirical relations found in literature relating shape and

friction angle for tailings? 4. Is possible to obtain an empirical relation among tailing parameters and the

internal friction angle? 5. What is the effect of tailing particle size in the internal friction angle? 6. Which are the available shape descriptors? 7. What shape descriptors are available to use with the image analysis? 8. What is the influence of resolution and magnification on the shape descriptors?

4

1.2 Thesis outline The thesis consists of 8 chapters (Part I) and 8 appended papers (Part II). A brief summary of the content of each chapter and paper is presented below: Chapter 1: General introduction to the thesis. Includes aim and scope, research questions and structure of the thesis Chapter 2: Describes briefly the history development and fundamentals of the particle shape as well as their classification. It also includes some of the available effects of the particle shape in soil properties as friction angle and void ratio among others. Chapter 3: Provides information of the tailings used during the research as the source location, basic mineralogy description and the single number shape descriptors also called quantities. It is also described the general treatment given to the tailings before image analysis and before the casting of disturbed samples. Chapter 4: Describes in detail the results of the experimental work by test and highlight the outcomes. Chapter 5: Discuss the most relevant findings of the research. Chapter 6: Summarize the relevant research questions and its responses as well. It is include a list of main conclusions Chapter 7: It provides suggestions to further work related to the research presented in this thesis. References

5

Summary of appended papers Paper I: This paper summarizes and compiles the available shape descriptors and the methods used to determine the shape descriptors. It is also discussed and pointed the physical meaning of the shape descriptors and image analysis as a valuable tool to maximize the data collection and the data reproducibility. Paper II: The paper focus on the two dimensions image analysis; how the shape descriptors are affected when resolution and magnification are changed. Ideal geometrical figures and its attributes were selected to compare with the image analysis results. Paper III: Investigate tailings through triaxial test and particle shape quantification using two dimensions image analysis. The evaluated shape descriptors (quantities) are used in previous published empirical relations between shape and friction angle. Also the tailings particles shape are evaluated and compare among size range. Paper IV: Investigate the effects of vertical load on tailing particles through oedometer tests. Tailings were split by four range sizes (1-0.5, 0.5-0.25, 0.25-0.125 and 0.125-0.063mm) and each size range was subject to vertical load. The effects of the load were measured through the breakage (sieving) and the particle’s shape change. Paper V: Describes the behavior of the tailing particles when they are subject to erosional process. The erosional process was achieved using an experimental mill and rolling table in two configurations, using balls and autogenous (no balls). Paper VI: Investigate the tailing particles friction angle base on particle size, void ratio and shape using direct shear tests. Paper VII: Describes basic tailing characteristics including specific gravity, face relationship, particle sizes, particle shape and direct shear behavior. Showing that particle size decreases along depth from surface for collected sample location Paper VIII: Presents the results from oedometer tests conducted on tailings materials. The study includes the stress-deformation behavior and particle breakage of tailings material of different gradations upon application of incremental loads in oedometer tests. .

7

2 PARTICLE SHAPE

2.1 The particle shape 2.1.1 Background Interest of particle shape was raised earlier in the field of geology compared to geotechnical engineering. Particle shape is considered to be the result of different agents (temperature and moisture changes, rigor of the transport, etc.) and the particle genesis itself (rock structure, mineralogy, hardness, etc.) (Wentworth, 1922b). In the early 1900’s authors as Wadell (1932), Riley (1941) and Pentland (1927) among others develop their own shape descriptors also called quantities. A quantity is defined as the relation of two or more measureable characteristic of a particle e.g. perimeter, area, volume, major axis, shortest axis, etc., Paper I includes 42 quantities. Authors have named their quantities as Sphericity (Wadell, 1932), Working sphericity (Aschenbrenner, 1956), Degree of circularity (Wadell, 1935) among others but the same name some times is used to describe different quantities bringing confusion. It has been considered not to regard in the author’s name but in the actual quantity even if names are used during the description process of the actual document. Particle shape descriptors can be classified as qualitative and quantitative. Qualitative describe in terms of words (e.g. elongated, spherical, angular, etc.); and quantitative that relates the measured dimensions; in the engineering field the quantitative description of the particle is more important due to the reproducibility. Quantitative measures on particles may be used as basis for qualitative classification. Quantities could be also divided in two: -3D (3 dimensions) Obtained from a 3D scan or from two orthogonal images -2D (2 dimensions) Particle projection where the particle outline is drawn Image analysis in 3D requires a sophisticated equipment to scan the particles surface and create a 3D model. Orthogonal images combine two 2D images to create a 3D model, orthogonality is difficult to achieve specially in particles smaller than 1mm (Fernlund, 2005). Image analysis in 2D is easy to perform and it does not require sophisticated equipment (e.g. regular camera or camera mounted in a microscope for smaller particles). In 2D image analysis the particle is assumed to lay over its more stable axis or randomly (Wadell, 1935; Riley, 1941 and Hawkins, 1993). 2.1.2 Scale dependence classification In order to describe the particle shape in detail, there are a number of terms, quantities and definitions used in the literature. Some authors Mitchell and Soga (2005) and Arasan, et al. (2010) are using three scales to classify quantities; one and each describing the shape but at

8

different scales. The terms are morphology/form, roundness and surface texture. Figure 1 shows how the scale terms are defined. At large scale the particle’s diameters in different directions are considered. At this scale, describing terms as spherical, platy, elongated etc., are used. An often seen quantity for shape description at large scale is sphericity (antonym: elongation). Graphically the considered type of shape is marked with the dashed line in Figure 1. At intermediate scale it is focused on description of the presence of irregularities. Depending on at what scale an analysis is done; corners and edges of different sizes are identified. By doing analysis inside circles defined along the particle’s boundary (Figure 1) deviations are found and valuated. A generally accepted quantity for this scale is roundness (antonym: angularity). Regarding the smallest scale, terms like rough or smooth are used. The descriptor is considering the same kind of analysis as the one described above, but is applied within smaller circles, i.e. at a smaller scale. Surface texture is often used to name the actual quantity.

2.2 Measurement techniques There are different methods that can be used to determine the shape of particles:

Hand measurement: This technique was probably the first used for obvious reasons and rapidly evolve with the use of measuring devices like the sliding rod caliper (Krumbein, 1941), convexity gage (Wentworth, 1922a), Szadeczky-Kardos apparatus (Krumbein and Pettijohn, 1938) and the camera lucida (Wadell, 1935).

Chart comparison: Basically tree charts appear in literature to compare the roundness and/or angularity Krumbein (1941), Powers (1953) and Krumbein and Sloss (1963). Chart comparison shows high variability due to the human error (Folk, 1955).

Sieving: Standards like EN 933-3:1997, ASTM D4791 among others has been developed. Sieving provides only elongation indexes and it becomes more complicated in fine grains (Persson, 1998).

Computer based image analysis: Shape descriptors or quantities can be obtained from image analysis in 3 or 2 dimensions. Algorithms have been developed to determine the particles attributes depending on the quantity to evaluate (e.g. area, longest and shortest axis, perimeter, etc.).

Figure 1 Shape describing the scale classification (Mitchell and Soga, 2005)

9

All measurement techniques present their own particularities and more detailed information is available in Paper I. Image analysis in two dimensions was chosen over the rest of the techniques due to its advantages:

- Available equipment and software - Particle size limits are beyond the actual used minimum size in the research - Capability to deal with large amount of particles in short time - Reproducibility

2.3 Arrangement and contact of the particles In laboratory, test on the effect on particle size on basic mechanical properties has been investigated. Particle size as basis has been discussed and various mechanisms had been proposed to explain the behavior of the soil. Basically there are two mechanisms proposed: The arrangement of particles and the inter-particle contact (Santamarina & Cho, 2004). 2.3.1 The arrangement of particles Arrangement of the particles can be presented in three different forms, loose, dense and critical; this arrangement determines the soil properties (e.g. density increase with more dense arrangement). Figure 2 shows the difference while in the upper part of the figure the particles are arranged using the minimum space needed, in the lower part a span is created using the flaky particle as a bridge this phenomena is known as “bridging”. Bridging can produce different geotechnical results when just the shape of the particle is changed, e.g. void ratio (Santamarina & Cho, 2004). Particles are able to rearrange, this could be done applying pressure (energy) to the soil, the pressure will create such forces that soil particles will rotate and move (see Figure 5) finishing in a more dense state. A loose soil (continues line in Figure 3), will contract in volume on shearing and may not develop any peak strength. In this case the shear strength will increase gradually until the residual shear strength is revealed, once the soil has ceased contracting in volume (see Figure 3). A dense soil (dotted line in Figure 3) may contract slightly (Figure 3, right) before granular interlock prevents further contraction (granular interlock is dependent on the shape of the grains and their initial packing arrangement). In order to continue shearing once

Figure 2 Bridging effect when flaky particles are combined in the bulk material (Santamarina & Cho, 2004)

10

granular interlock has occurred, the soil must dilate (expand in volume). As additional shear force is required to dilate the soil, peak shear strength occurs (Figure 3, left). Once this peak shear strength caused by dilation has been overcome through continued shearing, the resistance provided by the soil to the applied shear stress reduces (termed strain softening). Strain softening will continue until no further changes in volume of the soil occur on continued shearing. Peak shear strengths are also observed in overconsolidated clays where the natural fabric of the soil must be destroyed prior to reaching constant volume shearing. Other effects that result in peak strengths include cementation and bonding of particles. The distinctive shear strength, called the critical state, is identified where the soil undergoing shear does so at a constant volume (Schofield & Wroth, 1968). 2.3.2 The inter-particle contact For frictional soil, i.e. coarse grained soil (diameter >0.063mm), the friction between particles is the dominating factor for strength. Materials usually consisting of coarse grains behave as a frictional soil. It means that the strength of coarse soils (silt, sand, gravel, etc.) comes from an inter-particle mechanical friction thus, ideally they do not have traction strength (Axelsson, 1998). In Figure 4 the inter-particle contact is illustrated, here the pressure (P) is applied and

Figure 4 Inter-particle contact and forces acting (Axelsson, 1998)

Figure 3 The left part of the figure show a typical behavior of loose and dense material over shear stress, while at the right the figures illustrate the typical volume changes (Schofield & Wroth, 1968).

11

two more components are found, the normal load (N) and the tangential load (T) described as the friction coefficient (μF). The forces stand in equilibrium. When particles equilibrium is disturbed (friction coefficient is not enough to keep particles unmoved) the rotation is imminent, and it is necessary in order to compact the soil, in Figure 5 can be seen that the arrangement is a fact that inhibit or allow this rotation, and the shape in the 3 different scales are also factors because the more spherical and/or more rounded and/or less roughness more easy is the rotation. (Santamarina & Cho, 2004). 2.3.3 Breakage Breakage is the effect of the stresses when they exceed the strength of the rock particle. Breakage depend on physical aspects e.g. stress path, coordination number (number of contact points), temperature changes, among others and on the particle genesis e.g. rock structure, mineralogy, hardness, etc. (Wentworth, 1922b). Yoginder, et al. (1985) notice that the angular particle break during his experiments and they turn more rounded changing the original size and form configuration at the same time there was a soil properties loosening. The bridging effect (described early in this chapter) contributes to the breakage due to the voids below the elongated particle do not provide support cracking and disrupting the bridge as the pressure increases (Santamarina & Cho, 2004). The number of contact points (coordination number) and area of contact defines the pressure thus, if there are more contact point (and contact area) the pressure per area unit will be lower making clear why the angular particles are more susceptible to breakage even at low stresses (Wentworth, 1922b).

2.4 The influence of the shape There are five types of forces which may act between the particles in soils (Qazi, 1975): 1. Force of friction between the particles 2. Force due to presence of absorbed gas and/or moisture of particle 3. Mechanical forces, caused by interlocking of particles of irregular shape 4. Electrostatic forces arising from friction between the particles themselves and the surface

with which they come in contact 5. Cohesion forces operating between neighboring particles

Figure 5 Rotation inhibition by the particles compaction at low void ratio (Santamarina & Cho, 2004)

12

Research has shown that particle shape influence soil properties as void ratio and porosity, angle of repose, shear strength, sedimentation properties, hydraulic conductivity and risk for liquefaction. 2.4.1 Void ratio and porosity Several studies shows that there is a relation between morphology and roundness on void ratio Holubec & D'Appolonia (1973), Rousé, et al. (2008), Youd (1973) and Cho, et al. (2006) among others. The results show that the maximum and the minimum void ratio decreases as the particles are more uniformed in morphology and rounded in corners (see Figure 6 and Figure 7).

Figure 6 Example of behavior of the minimum void ratio. Values close to 1 represent uniform particles and rounded edges while 0 is the opposite. Data from Holubec and D’Appolonia and

Sukumaran & Ashmawy were modified due to the difference in the working range (range 1 to 2 and 0 to 100 respectivelly)

Figure 7 Example of behavior of the maximum void ratio. Values close to 1 represent uniform particles and rounded edges while 0 is the opposite. Data from Holubec and D’Appolonia and

Sukumaran & Ashmawy were modified due to the difference in the working range (range 1 to 2 and 0 to 100 respectivelly)

13

The morphology and roundness have bigger influence on the maximum void ratio. The effect on the maximum void ratio is more pronounced than the change of the minimum void ratio when the form and roundness changes, as shown in Figure 8. Particles internal arrangement and interlocking are probably the main factor that effects the void ratio; bridge effect permit the existence of void among the particles while interlocking allowed the particles to form arches avoiding the possibility to rotate and stay in a more stable configuration e.g. as it happens with marbles. 2.4.2 Angle of repose The angle of repose of a granular material is the steepest angle of descent or dip of the slope relative to the horizontal plane when material on the slope face is on the verge of sliding. Incremental roundness decrees the angle of repose (Rousé, et al., 2008) Rousé et. al. (2008) proposed the empirical relation: where R is the roundness value, defined by Wadell (1935) 2.4.3 Shear strength In engineering, shear strength is the strength of a material or component against the type of yield or structural failure where the material or component fails in shear. The Mohr–Coulomb failure criterion represents the linear envelope that is obtained from a plot of the shear strength of a material versus the applied normal stress. The internal friction angle is represented by the slope of the plotted envelop. Studies show that the shear strength (under drained triaxial tests) increases more rapidly on those materials having higher angularity increasing the relative density. The internal friction angle is a function of the relative density and the particle shape (Holubec & D'Appolonia,

Figure 8 Example of behavior of the maximum minus minimum void ratio. Values close to 1 represent uniform particles and rounded edges while 0 is the opposite. Data from Holubec and D’Appolonia and Sukumaran & Ashmawy were modified due to the difference in the working range (range 1 to 2 and 0

to 100 respectivelly)

R4.147.41rep

14

1973). More studies as Chang & Page (1997) made with dry copper in direct shear tests and Shinohara, et al. (2000) with steel powder using a triaxial cell conclude practically the same. The following empirical relations have been proposed to the behavior of the internal friction angle: Cho, et al. (2006): R is the roundness value, Krumbein and Sloss (1963) chart Rousé, et al. (2008): R is the roundness value, defined by Wadell (1935) In Figure 9 the suggested empirical relations above and data from Holubec & D'Appolonia (1973) and Sukumaran & Ashmawy (2001) were plotted together to display the general trend on the particle shape and friction angle relation. Sukumaran reports two lines, one based on the shape factor (SF) and the second referring the angularity factor (AF). 2.4.4 Sedimentation A particle released in a less dense Newtonian fluid initially accelerate trough the fluid due to the gravity. Resistances to deformation of the fluid, transmitted to the particle surface drag, generate forces that act to resist the particle motion. Particle shape has been assumed to be spherical when equations are applied on the settling velocity. Correlation deviates when particle shape departs from spherical form and it is known that natural particles depart from spherical form (Dietrich, 1982). Empirical relations have

Figure 9 The change of the internal friction angle shows a general increase when the particle becomes more angular or less spherical. Holubec & D’Applonia and Sukumaran & Ashmawy original

morphology/roundness ranges (1 to 2 and 0 to 100 respectively) were modified to fit the figure.

R1742

R6.93.34

been suggested by Jimenez & Madsen (2003) and Dietrich (1982) to account for the particle shape into the settling velocity.

2.4.5 Hydraulic conductivity, Permeability Darcy’s Law: Permeability is one component of Darcy’s law. Darcy's law is a simple proportional relationship between the instantaneous discharge rate through a porous medium, the viscosity of the fluid and the pressure drop.

Reynold’s number (Laminar and turbulent Flow): Typically any laminar flow is considered to have a Reynold’s number less than one, and it would be valid to apply Darcy's law. Experimental tests have shown that flow regimes with Reynolds numbers up to 10 may still be Darcian (laminar flow), as in the case of groundwater flow.

Permeability is affected by the shape and texture of soil grains. Elongated or irregular particles create flow paths which are more tortuous than those spherical particles. Particles with a rough surface texture provide more frictional resistance to flow. Both effects tend to reduce the water flow through the soil (Head & Epps, 2011). the amount of soil detached during laminar and turbulent flow is dependent on each soil and also greater on turbulent flow due the greater shear strength generated during this kind of flow, this could suggest the greater erosion when turbulent flow is present (Nearing & Parker, 1994).

Kozeny-Carman empirical relation accounts for the dependency of permeability on void ratio in uniformly graded sands; serious discrepancies are found when it is applied to clays due the lack of uniform pores (Mitchell and Soga, 2005). There are various formulations of the Kozeny-Carman equation; one published by Head & Epps (2011) relates the angularity among other factors with the permeability. Kane & Sternheim (1988) suggest that the inclusion of the shape has probably the background on the Reynolds number due this factor is dependent significantly on the shape of the obstacles increasing the turbulence and flow resistance when morphology of the particles in the soil is more irregular.

2.4.6 Liquefaction Soil liquefaction is a phenomenon in which soil loses much of its strength or stiffness for a generally short time by earthquake shaking or other rapid loading. Static and dynamic liquefactions occur been the second one the most regular known. Liquefaction often occurs in saturated soils, typical condition in tailing dams. This water exerts a pressure on the soil particles that influences how tightly the particles themselves are pressed together. Shaking or other rapid loading can cause the water pressure to increase to the point where the soil particles can readily move with respect to each other (Jefferies & Been, 2000). At low confining pressure angular material is more resistant to liquefaction. Probably the breakage of the corners on the angular particles in tailings is ruling the lost in resistance at high confining pressures, Sieve analysis after test identified the breakage of angular particles while on rounded particles the sieve analysis was practically with no change (Yoginder, et al., 1985).

15

16

2.4.7 Groundwater and seepage modeling In groundwater flow the particle shape affects the soil pore size distribution, hence, the flow characteristics. Current models incorporating the effects of particle shape have failed to consider irregular particles such as those that would prevail in a natural porous medium. Research conclusions suggest that particle size and porosity are more important predictors for hydraulic conductivity explaining the 69% of the variability but particle shape appears to be the next most important. The interaction effect of the particle size and particle shape suggests a different packing configuration for particles of the same shape but different size (Sperry & Peirce, 1995). Tortuosity and permeability (also see section 2.4.5) are two significant macroscopic parameters of granular medium that affect the passing flow. It was suggested based on laboratory results that tortuosity effect converge when the porosity increases indicating that the shape have dominance at low and middle porosity ranges (Hayati, et al., 2012). For more detailed information in the shape effects consult the report by Rodriguez (2012).

17

3 MATERIALS AND METHODS – EXPERIMENTAL PART

3.1 Tailings location In this thesis samples from the Aitik mine belonging to Boliden AB has been used. The Aitik tailing dam is located about 100 km north of the Arctic Circle in the boreal parts of Northern Sweden (Figure 10) about 15 km from the community of Gällivare. The mine was opened in 1968 and since then tailings had been deposited, at present the tailing dams covers an area of 13 km2. Since all tailings used for testing were taken from the same site this study is considered as a case of study.

3.2 The ore and tailings composition The target minerals are mainly copper and gold and the ore is of sulphide type. The valuable copper mineral is in the form of chalcopyrite, CuFeS2. Main sulphides are pyrite, chalcopyrite and sphalerite. Main gangue minerals are quartz, feldspar, plagioclase and mica (Lindvall, M. and Eriksson, N., 2003).

3.3 The quantities or shape descriptors Used quantities are described in Table 1 and a graphical description can be found in the appendix. Quantities were chosen due to its availability to be used in the commercial and freeware software respectively, Image pro-plus® (Image Pro Plus v. 7.0, 2011) and imagej (ImageJ, 2013).

Figure 10 Circle shows the location of the Aitik tailing dam in Sweden (left) and the dam (right). (Google map, 2014).

18

Table 1 Quantities available to be use in image analysis

Quantity number

Quantity description Working range

Reference

1 #Major axis/Minor axis 0 - 1 (Hawkins, 1993) 2 +4πArea/Perimeter2 0 - 1 (Cox, 1927) 3 #4Area/πMajor axis2 0 - 1 (ImageJ, 2013) 4 +Area/Convex Area 0 - 1 (Mora and Kwan, 2000) 5 +Fractal dimension 1 - 2 (Image Pro Plus v. 7.0, 2011) 6 #Square root of Maximum inscribed/Minimum

circumscribed, circle diameters 0 - 1 (Riley, 1941)

7 #Diameter of a circle same area as particle/Minimum circumscribed circle diameter

0 - 1 (Wadell, 1935)

8 #Perimeter2/Area * 0 - 1/4π (Blott & Pye, 2008) 9 #Perimeter of a circle with same area/Perimeter 0 - 1 (Wadell, 1935)

10 #Area/Area of the minimum circumscribed circle 0 - 1 (Tickell, 1938) 11 +Perimeter/Convex perimeter * 0 - 1 (Janoo, 1998) 12 +Perimeter /πAverage feret diameter* 0 - 1 (Kuo, et al., 1998)

*Inverse is used to obtain a working range between 0 and 1 # Quantity describing the large scale “form” +Quantity describing the intermediate scale “roundness”

3.4 The tailing sampling and samples Disturbed and undisturbed samples were used for laboratory tests. The disturbed samples from the tailing dam were obtained from test pits at different locations. Disturbed samples were used for oedometer, direct shear and mill attrition tests. The undisturbed samples were provided by the mine administration (Boliden Minerals AB) using a typical piston sampler. (SGF, 1996) Undisturbed samples were used only for triaxial tests. Reference material was taken from undisturbed samples for geotechnical and shape characterization. Sieving size ranges used during this thesis can be seen in Table 2. Further on the thesis, for convenience size ranges are called by their lower limit, e.g. 0.063mm instead 0.125-0.063mm. Particle size used for each specific test is defined in section 3.6. Wet sieving was used with Sodium Diphosphate decahydratate (Na4P2O7·H2O) as a dispersant to enhance the particle separation. After sieving tailings were dried for 24 hours at 105°C. Image analysis data collection requires the particles to be dry and splitted by size in order to focus and avoid flocculation, this is the reason to separate by size, and the other reason to split by size is to perform tests with a single particle size range.

Table 2 Particle size ranges Range sizes (mm)

Upper limit Lower limit 2 1 1 0.5

0.5 0.25 0.25 0.125

0.125 0.063

19

Disturbed samples were remolded and casted by using single particle size e.g. 0.5mm. Remolded samples made of a single particle size were used in the oedometer and direct shear tests. Casting was made in 50mm diameter and 170mm length sampling tubes by using the methodology describes by Dorby (1991). Dorby’s procedure includes the filling of the tube specimen by steps, usually 5 to 6 in total, where each step comprises 2-3cm of the tube height. Water is added until the step is reached followed by the addition of the tailings sample and posterior self-settlement for at least 6 hours. Same procedure is followed for every step until the tube is filled up. This methodology simulates the natural sedimentation process leading the tailings settle in beds with natural segregation. Since the test specimens are uniformly graded the segregation should be based on the grain density differences and not in the particle size. Basic geotechnical properties were determined for each sample used during the test such as particle density, bulk density, saturated density, void ratio and degree of saturation (more detailed information about each test is described in the appended papers).

3.5 Image acquisition and Analysis The image acquisition was performed through a microscope (Motic B1). Lightening sources can be set from below and from the side of sample. The camera mounted on top of the microscope (Infinity 2) for image acquisition has 2 megapixel resolution. Magnification lenses 4x and 10x were used for particles size range 2-1, 1-0.5, 0.5-0.25mm and 0.25-0.125, 0.125-0.063mm respectively. Image acquisition in two dimensions was selected due to the simplicity of the method and also due the availability of the equipment. The magnification and resolution used for each size range was preferred based on the results on Paper II were the effects of the resolution and magnification in a set of tests were studied. A more detailed description of the image acquisition is described Paper II and III. Image analysis applied to the actual work is the examination of digital files (photographs) to measure different properties as perimeter, area, axis, etc. Shape descriptors or quantities combine one or more of this parameter to describe the shape of a particle (see Table 1). ImageJ (ImageJ, 2013) and Image Pro-Plus® (Image Pro Plus v. 7.0, 2011) are the image analyzer tools (software) chosen to develop the research due to ImageJ is a freeware and in combination with Image Pro-Plus ® (software available in the market) could provide the measurements required to obtain the major amount of quantities, 12 in this case (see quantities in Table 1). 3.5.1 Statistics Certain number of measured particles is required to obtain a stable margin of error that should only depend on the distribution itself. The number of particles (n) measured is determined by the Equation 1:

20

Where: n is the number of particles, z is the critical value for the desired level of confidence (in our case 95%, z=1.96), σ is the standard deviation and m is the margin of error. To calculate the number of particles to be measured a database of 300 measured particles was obtained. An initial calculus with 20 particles was done to determine the margin of error with respect of the 300 particles. Furthermore same test was performed adding 20 particles each time to the database until all 300 particles were included. Final margin of error were chosen when there was no significant variation and Equation 1 was applied to obtain the number of particles to measure. Finally it was decided to measure from 180 to 200 particles. Two normal distributions can be compared using the Two sample T-test (Snedecor & Cochran, 1989) but if the distribution is not normal there are options like transform the data (make it normal) or use non-parametric test. In the present thesis Johnson transform (Johnson, 1949), non-parametric Mann-Whitney test (Lincoln, 2014) and Two samples t-test were used. Results showed that the three tests agree in the 97% of the cases thus, two samples t-test was adopted for simplicity. Even when the number of particles measured in some cases was not the ideal, the two samples t-test can determine the margin of error in between distributions to compare and conclude its similarity or difference.

3.6 The laboratory tests on tailing properties. The laboratory tests were chosen to answer questions related to physical erosion of the tailing particles and how changes can affect the strength. Among the performed tests are triaxial were the internal friction angle and particle shape was obtained to evaluated them in published empirical relation to identify those that could be applicable in tailings. Furthermore oedometer test were performed to obtain in first instance the strain and settlement aptitude but also to determine the breakage behavior and shape change of the tailings due to one dimensional load. Physical weathering by mill degradation has the intention to determine the erosional effects trough the time in breakage and also in shape change. Direct shear and evaluation of factors as consolidation, initial void ratio, particle shape and particle size have the intention to obtain a relation with the tailing strength internal friction angle. Finally a general evaluation of the tests was carried out to find connections or contrasts in the physical degradation tests performed.

Equation 1 Determination of the minimum number of measured particles

mzn

21

3.6.1 Triaxial (Paper III) Active triaxial test were performed on three samples in a confining stress-ranges of 90-150 kPa in drained conditions (see Table 3). One sample was executed in undrained conditions. The samples used were retrieved by undisturbed coring. The samples had, incremental load step, isotropic confined. During this consolidation a backpressure of 100 kPa has been applied. In the case a maximal deviatoric peak stress was present during shearing the friction angle was evaluated at both the maximal deviatoric stress and at the residual stresses. For those samples where a maximal deviatoric stress was not present the friction angle was evaluated at 15 % compression was used. Empirical relations Table 4 found in literature were compared with tailings triaxial test results.

Table 3 Triaxial test sample conditions Samples Sampling

depth (m) Triaxial test Confining

pressure (kPa) a 13.3 Drained 140 b 13.3 Undrained 150 c 8.3 Drained 90 d 16.8 Drained 170

Table 4 Empirical relations suggested in literature relating the friction angle and the shape quantity

3.6.2 Oedometer (Paper IV) Oedometer tests according to ASTM D2435 were performed in disturbed samples doubling the weight every 24 hr starting with 10, 20, 40, 80, 160, 320 and 640 kPa load steps under saturated and drained conditions. Samples were made of uniformed size range particles of 0.5, 0.25, 0.125 and 0.063mm 3.6.3 Wearing/attrition/degradation by mill (Paper V) Laboratory milling in wet conditions was used to physically weather the tailings, smooth drum mill was used to accomplished attrition using iron balls and also in autogenous set up during 2 and/or 3 time periods (see Table 12). A mill speed of 60 rpm was set to create a constant rebuilding cascade able to erode the material and avoid the impact generated at higher revolutions (Sponenburgh, 2006). Degradation was identified measuring shape change (image analysis) and breakage using traditional sieving (en each particular size). Sample particle sizes were same as describe in point 3.6.2. 3.6.4 Direct shear (Paper VI) Direct shear tests were performed on disturbed samples. Samples were mounted by surrounding reinforced latex membrane and porous filter spikes were placed on top and bottom. Rubber tape at the end of membrane edges was used to avoid any leakage from membrane edges (see Figure 11). NGI (Norwegian Geotechnical Institute) Direct simple shear

EQ. # DEFINITION REFERENCE

2 Cho et. al. (2006)

3 Rousé (2008)

Q, quantity value (0 to 1)

Q1742'

Q4.147.41'

22

apparatus was used for this thesis. The apparatus has been rebuilt and modified with electronic sensors which enable to record applied load, specimen height and pore pressure continuously during shearing. The logged data is then transferred to computer program which helps with the monitoring of stresses and deformations during the test. Tests were performed in saturated and drained conditions. Sample particle sizes ranges were 0.25, 0.125 and 0.063mm. Consolidation loads for the samples were 50. 100. 150, 300 and 500 kPa. According with the Swedish criteria (SGF, 2014) the shear strength is obtained at 0.15 radians.

Additional information about the test conditions can be found in the referred papers .

Figure 11 Direct shear apparatus and sample mounting (Bhanbhro, et al., 2013)

23

4 RESULTS By visual inspection using Powers (1953) roundness chart of the tailings samples, the particle shape range are subjectively classified from sub angular to very angular. The smaller sizes appear to be more angular compared to larger. Figure 12a) shows the particles counted and the distribution based on the figure12b) Powers (1953) chart.

4.1 Triaxial test and friction angle empirical relations Quantities presented in Figure 13 are arranged by size they show that larger particles are more uniform in shape. In general, for all quantities, the uniformity showed in big particles diminishes for smaller particles. It is also evident (in quantities 1-1, 2, 3 and 4) that in the box plot (between 1st and 3th quartile), the average and the minimum values move downwards in the vertical axis (from bigger to smaller size) while the maximum values seem to have no apparent change (close to the upper limit). Figure 13 Box-plot for analyzed tailings material grouped by size (entire triaxial database)

a) b)

Figure 12. a)Visual inspection results of the tailing particles b)Powers (1953) comparative chart

24

In sample c (Figure 14) some oxidation is recognized consisting of reddish colored zones. The red color is presumed to be the result of the iron oxidation coming from pyrite and chalcopyrite that is commonly present in the tailings. The X-ray tests show that only sample c contains iron oxide in the form of FexOy. Quantity 4 in Figure 14 is able to show the value increase when particle size increases in all samples while for the rest of the quantities it is not that defined for sample c. Sample c (oxidation presence as FexOy) does not show clear increase or decrees in values when quantity number 1, 1-1 and 3 are applied. Quantity number 2 is showing a clear value increase on sample d while for the rest of the samples a higher peak value appears in some cases in small sizes and in others in medium sizes thus, from quantity 2 (except sample d) it is not clear the size-shape increasing or decreasing behavior. Quantities number 1-1 and 3 have the same result and represents the invers of the quantity number 1 (see the mirror image comparing quantity 1 and quantity 1-1). Samples values on quantity 3 and 1-1 (except sample c) increase as particle size increase and for quantity 1 the values decrees as particle size increase (except sample c). The results from the basic geotechnical characterization and the triaxial tests are compiled in Table 5. As seen in the table the result of an evaluation of the friction angle by Mohr-Coulomb failure criterion will generate high difference depending on a cohesion intercept is used or if the cohesion intercept is omitted (c=0 kPa). By inspection of the samples the material could be considered to have low or moderate cohesion depending on the clay content. Thus is the approximation of the failure envelope to be straight, as in the Mohr-Coulomb failure criterion, not valid for high stress intervals, or at least for the lower ranges of effective stress. The initially loose specimens has equally high, or higher friction angle compared to the firm. Dilatant behavior during shearing was observed for all the initially firm specimens including the undrained specimen and contractant behavior for the loose samples.

Figure 14 Particle shape by quantity and sample

25

Table 5 Results from the basic geotechnical characterization and triaxial tests. Sample (-) is in this study only used for evaluation of strength properties

ID a - b c d Sample location I I I II II Sampling depth [m] 13.3 20.0 13.3 8.3 16.8 Ocular classification saSi clSa clSi siSa siCl Initial state Firm Firm Firm Loose Loose Condition Drained Drained Undrained Drained Drained ρ [t/m3] 1.96 1.97 1.93 1.76 1.89 w [%] 31-39 33-36 29-34 49-54 40-50 Initial conditions [kPa]

'3

'1

31 u

140 250 110

200 300 100

150 270 120

90

200 110

170 280 110

Compression rate. [mm/h] 1.0 0.9 2.0 0.6 0.6 Failure criterion Max

dev.str. Max

dev.str. Max

dev.str. 15%

compr. Max

dev.str. Compr. at failure [%] 10.4 9.6 14.2 15.0 14.4

'1 at failure [kPa] 564 826 1170 484 794

'3 at failure [kPa] 141 201 253 90 170

u at failure [kPa] 109 99 17 110 110 ’ (c = 0) 40º 37º 40º 43º 40º ’ and c c = 79 kPa, ’ = 29º c = 39 kPa, ’ = 37º

Figure 15 represent sample d of the table 5 where Quantities 1-1, 2, 3 and 4 were used to measure the shape of particles of such sample d. Each subset of figures are separated by size (horizontal axis) and the relative amount of material (vertical axis). Figure 15 shows how the irregular particles (Lower the value in the legend represent more elongated and angular particles) populates the fine section of the figures; this is more empathized in quantities 1-1, 2 and 3 with 40%, 50% and 40% with quantity value of 0.5. The results from samples a and b

Figure 15 Percentages of the total particles by size and quantity values for sample d.

Quantity 1-1 Quantity 2

Quantity 3 Quantity 4

26

(Table 5) are similar, always appearing low values in the small fraction of size. Sample c is the exception (where oxidation is present) due to contains more regular shape along the size fractions resulting in a flat graphic behavior. Since a sample grains were analyzed the output was a range of numerical values of each shape descriptor. In this thesis the average, minimum and maximum values for each of the four quantities (1-1, 2, 3 and 4) were tested in the equations 2 and 3 (Table 4). As reference to the output friction angle was the results from the triaxial tests on the samples in Table 5. In Figure 16 a line and a dotted line represent the internal friction angle obtained by sample testing (ϕ’triaxial), the points (x ex, ∆ triangle, + plus and ◊ rhomb) represent the equation 2 and 3 output when using the quantity (1, 2, 3 and 4) values. Figure 16 shows that the minimum of each shape descriptor presents less difference between the expected empirical friction angle (ϕ’empirical) and the laboratory tests output (ϕ’triaxial). Among the quantities number 2 has the highest accuracy of predicted friction angle (see Table 6).

QUANTITIES min=minimum, max=maximum, avg=average

Q S

Figure 16 Friction angle results, lines (dotted and pointed) represent the laboratory test results (ϕ’triaxial) and the points are the empirical relation output (ϕ’empirical) using the equation 2 and 3

27

Table 6 Main internal friction angles data differences among lab test and empirical relations Equations

Quantity 2 3

1-1

average 13.0 8.8

maximum 14.5 11.1 minimum 10.7 5.4

2

average 12.3 7.8 maximum 14.3 10.8 minimum 8.6 2.3

3


4


4.2 Vertical load effects Table 7 summarizes the basic properties of the tailings specimens before and after the oedometer test. Similar test were performed to natural sands, the basic properties for sand are in Table 8.

Table 7 Basic tailing properties Specimens (size, mm)

Particle Density (g/cm3)

Void ratio (e)

Δe Porosity n (%)

Δn (%)

Dry Density,ρ (gr/cm3)

Δ ρ

0.5 Initial

2.881 1.070

0.227 51.7

6.0 1.392

0.171 Final 0.843 45.7 1.563

0.25 Initial

2.904 0.849

0.141 45.9

4.4 1.571

0.129 Final 0.708 41.5 1.700

0.125 Initial

2.873 0,762

0.093 43.3

3.2 1.630

0.092 Final 0.669 40.1 1.722

0.063 Initial

2.943 0.847

0.113 45.9

3.6 1.589

0.108 Final 0.734 42.3 1.697

Table 8 Basic sand properties

Specimens (size, mm)

Particle density (g/cm3)

Void ratio (e)

Δe Porosity n (%)

Δn (%)

Dry Density, ρ (gr/cm3)

Δ ρ

0.5 Initial

2.656 0.700

0,148 41.2

5.7 1.563

0.149 Final 0.552 35.5 1.712

0.25 Initial

2.651 0.740

0,118 42.5

4.2 1.524

0.111 Final 0.622 38.3 1.635

0.125 Initial

2.673 0.751

0,096 42.9

3.3 1.527

0.088 Final 0.655 39.6 1.615

0.063 Initial

2.684 0.954

0,156 48.8

4.4 1.374

0.119 Final 0.798 44.4 1.493

28

In Table 9 and Table 10 bold-gray highlighted paired numbers represent those values where the mean of both populations (before and after test) show that they are different. Control specimens (natural sand) have shown no shape change (see Table 10). All changes in shape indicate that particles become more rounded except in Table 10 where “↓” indicates that particles became more angular/irregular.

Table 9 Mean quantity values for samples. Highlighted marked results show statistically significant changes in the shape quantities.

Table 10 Mean quantity values for the control sample, Natural sand. Simbol “↓” indicates that particles became

more angular/irregular.

Figure 17 (left) shows the stress-strain behavior where specimens 0.5 and 0.25mm are weaker. Figure 17 (right) show that all specimens have initially high deformation, probably a result of initial rearrangement of the loose structures and re-distribution of stresses in the test specimens, especially for the smaller fraction 0.063mm where the initial strain was not expected. The step strain for the test specimens 0.5, 0.25, and 0.125mm differs from the specimen 0.063mm behavior (Figure 17, right) because this last specimen shows a high initial deformation and no significant difference in deformation between the two last load steps.

Quantity

2 3 4 5 6 7 8 9 10 11 12 Sample (size, mm)

0.5 Before 0.665 0.753 0.9303 1.0339 0.6451 0.8011 0.1534 0.8458 0.6449 0.9197 0.2930

After 0.660 0.742 0.9269 1.0339 0.6368 0.7918 0.1631 0.8484 0.6302 0.9290 0.2960

0.25 Before 0.690 0.735 0.9275 1.0414 0.6375 0.7945 0.0840 0.8626 0.6361 0.9405 0.2998

After 0.712 0.746 0.9315 1.0390 0.6515 0.8049 0.0815 0.8804 0.6518 0.9530 0.3039

0.125 Before 0.702 0.713 0.9406 1.0292 0.6292 0.7853 0.0433 0.8795 0.6224 0.9599 0.3060

After 0.695 0.744 0.9363 1.0286 0.6478 0.8035 0.0474 0.8770 0.6492 0.9513 0.3032

0.063 Before 0.694 0.693 0.9329 1.0492 0.6098 0.7759 0.0251 0.8729 0.6083 0.9603 0.3065

After 0.722 0.738 0.9413 1.0439 0.6460 0.7973 0.0266 0.8919 0.6395 0.9638 0.3076

Quantity 2 3 4 5 6 7 8 9 10 11 12 Sample (size,

mm)

0.5 Before 0.730 0.705 0.952 1.022 0.637 0.789 0.135 0.890 0.628 0.966 0.307

After 0.735 0.732 0.954 1.016 0.657 0.801 0.186 0.895 0.645 0.966 0.307

0.25 Before 0.736 0.723 0.948 1.034 0.643 0.794 0.081 0.895 0.634 0.969 0.309

After 0.743 0.735 0.951 1.030 0.651 0.800 0.085 0.898 0.644 0.968 0.309

0.125 Before 0.742 0.731 0.954 1.023 0.653 0.815 0.047 0.031 0.698 0.970

↓ 0.309 ↓

After 0.734 0.736 0.953 1.022 0.656 0.804 0.051 0.033 0.650 0.967 0.308

0.063 Before 0.744 0.717 0.951 1.040 0.643 0.793 0.027 0.903 0.633 0.974 0.311

After 0.736 0.701 0.949 1.039 0.632 0.791 0.028 0.899 0.630 0.972 0.310

29

In Figure 18 stress-strain curves of the tailings and control sand are plotted together. Dark lines represent the tailings while light lines are the control sands. In this figure if size by size are compared it can be seen that the secondary deformation is more stepped in the tailings. The degradation of the specimen was measured by quantifying the presence of finer material after oedometer-test. The amount of generated fines (in weight), determined by sieving, is presented in Table 11. The amount of degraded material decrees by grain size

Figure 18 Semi-logarithmic stress-strain curve for tailings and reference sand by particle size range (s = sands and t = tailings).

Figure 17 Semi-logarithmic stress-strain curve (left) and incremental stress-strain by load step (right) from the oedometer test by particle size range.

30

Table 11 The amount fines generated in the test specimens after the odometer test Specimen

tailings % fine content

Control specimen natural sand

% fine content

0.5 14 0.5 19 0.25 10 0.25 11

0.125 12 0.125 4 0.063 <1 0.063 2

4.3 Ball and autogenous milling Figure 19 represents the amount of fines generated after the autogenous and ball milling. This figure shows that the use of balls increases the abrasion speed (dotted lines). Smaller fraction (0.063mm) in both cases results in a lower weight lost for both attrition agents. Data from Table 12 was plotted into Figure 19. Figure 20 is the overall representation of the shape changes; gray lines and markers represent the shape change. In the left, ball degradation shows a rapid shape change especially for the 0.5 and 0.25mm fraction. In the right, autogenous degradation is not showing any shape change (except one for fraction 0.125mm). Figure was built with the data in Table 13 showing the gray markers and lines when more than 50% of the quantities account for a shape change.

Ball milling

Autogenous milling

Figure 19 Fines generation in mill test

31

Table 12 Time-fine content table for tailings milling Autogenous milling Ball milling

Sample Size (mm)

Time period (min)

% fine content

Sample Size (mm)

Time period (min)

% fine content

0.5 1260 7.6 0.5 100 32.0 4085 13.9 370 47.5 7207 14.7 1140 67.2

0.25 1442 6.7 0.25 100 33.9 4158 18.4 1046 69.0 1558 82.8

0.125 1344 11.3 0.125 103 19.8 4252 14.8 343 33.7 1441 90.6

0.063 1470 4.8 0.063 100 8.7 4222 6.4 324 17.2 1406 62.5

Table 13 is the entire comparison database where it can be seen the quantities used, quantity values and the shape change (marked in gray color), this table account for the step shape change. Shape changes colored suggest that the particles become more rounded/uniform with exception when “↓” appears (it means change is to be less rounded, less uniform).

Figure 20 Degradation and shape change by milling agent. Gray leyends indicate shape change. Left, ball milling. Right Autogenus milling (Fraction sizes are represented by the lower limit)

. .

32

Table 13 Shape change by time-step. Gray shadow is used when shape change was detected.

Size Balls Autogenous (mm) Q 2hr 6hr 24hr 24hr 72hr 120hr 0.5 2 0.665 0.681 0.681 0.702 0.702 0.732 0.665 0.668 0.668 0.673 0.673 0.672 3 0.728 0.742 0.742 0.735 0.735 0.760 0.728 0.743 0.743 0.741 0.741 0.748 4 0.924 0.935 0.935 0.943 0.943 0.951 0.924 0.932 0.932 0.932 0.932 0.931 5 1.041 1.031 1.031 1.025 1.025 1.022 1.041 1.033 1.033 1.032 1.032 1.031 6 0.787 0.802 .0802 0.803 0.803 0.819 0.787 0.796 0.796 0.796 0.796 0.801 7 0.786 0.800 0.800 0.799 0.799 0.818 0.786 0.792 0.792 0.792 0.792 0.798 8 0.126 0.161 0.161 0.160 0.160 0.160 0.126 0.159 0.159 0.153↓ 0.153 0.159 9 0.847 0.856 0.856 0.873 0.873 0.887 0.847 0.852 0.852 0.855 0.855 0.851 10 0.620 0.642 0.642 0.641 0.641 0.672 0.620 0.631 0.631 0.631 0.631 0.640 11 0.929 0.930 0.930 0.946 0.946 0.949 0.929 0.931 0.931 0.933 0.933 0.927↓ 12 0.296 0.296 0.296 0.301 0.301 0.303 0.296 0.297 0.297 0.297 0.297 0.295↓ 0.25 2 0.676 0.705 0.705 0.745 0.745 0.746 0.676 0.687 0.687 0.689 3 0.710 0.748 0.748 0.740 0.740 0.771 0.710 0.728 0.728 0.730 4 0.922 0.932 0.932 0.949 0.949 0.948 0.923 0.924 0.924 0.928 5 1.045 1.041 1.041 1.033 1.033 1.033 1.047 1.042 1.042 1.042 6 0.779 0.800 0.800 0.812 0.812 0.822 0.779 0.789 0.789 0.794 7 0.780 0.798 0.798 0.810 0.810 0.823 0.780 0.787 0.787 0.790 8 0.079 0.079 0.079 0.082 0.082 0.082 0.079 0.082 0.082 0.080 9 0.852 0.872 0.872 0.900 0.900 0.902 0.852 0.858 0.858 0.862 10 0.612 0.640 0.640 0.659 0.659 0.680 0.612 0.624 0.624 0.629 11 0.937 0.947 0.947 0.964 0.964 0.963 0.937 0.940 0.940 0.940 12 0.299 0.302 0.302 0.307 0.307 0.307 0.299 0.300 0.300 0.300 0.125 2 0.700 0.703 0.703 0.702 0.702 0.643↓ 0.700 0.684↓ 0.684 0.696 3 0.706 0.713 0.713 0.710 0.710 0.689 0.706 0.708 0.708 0.719 4 0.936 0.935 0.935 0.939 0.939 0.919↓ 0.936 0.926↓ 0.926 0.934 5 1.034 1.032 1.032 1.031 1.031 1.038↓ 1.034 1.032 1.032 1.032 6 0.784 0.785 0.785 0.787 0.787 0.769↓ 0.784 0.778 0.778 0.789 7 0.780 0.784 0.784 0.784 0.784 0.769↓ 0.780 0.777 0.777 0.788 8 0.038 0.040 0.040 0.040 0.040 0.039 0.038 0.046 0.046 0.044 9 0.878 0.877 0.877 0.878 0.878 0.843↓ 0878 0.865↓ 0.865 0.872 10 0.614 0.619 0.619 0.620 0.620 0.597↓ 0.614 0.610 0.610 0.625 11 0.961 0.960 0.960 0.960 0.960 0.936↓ 0.961 0.953↓ 0.953 0.953 12 0.307 0.306 0.306 0.306 0.306 0.298↓ 0.307 0.304↓ 0.304 0.304 0.063 2 0.700 0.702 0.702 0.713 0.713 0.690↓ 0.698 0.713 0.713 0.713 3 0.685 0.684 0.684 0.714 0.714 0.688 0.685 0.692 0.692 0.693 4 0.931 0.933 0.933 0.937 0.937 0.925↓ 0.931 0.939 0.939 0.936 5 1.051 1.051 1.051 1.053 1.053 1.061↓ 1.051 1.051 1.050 1.050 6 0.772 0.773 0.773 0.786 0.786 0.769↓ 0.772 0.777 0.777 0.781 7 0.768 0.771 0.771 0.782 0.782 0.766↓ 0.768 0.772 0.772 0.778 8 0.024 0.023 0.023 0.022↓ 0.022 0.018↓ 0.024 0.023 0.023 0.024 9 0.875 0.876 0.876 0.885 0.885 0.869↓ 0.875 0.885 0.885 0.883 10 0.596 0.600 0.600 0.616 0.616 0.593↓ 0.596 0.602 0.602 0.609 11 0.964 0.965 0.965 0.967 0.967 0.959↓ 0.964 0.971 0.971 0.967↓ 12 0.308 0.308 0.308 0.309 0.309 0.306↓ 0.307 0.310 0.309 0.308↓ Q=Quantity number, ↓=less spherical, less rounded

Differences of the samples compared with the initial shape are represented in Table 14. In this table ball milling seems to have an immediate influence on the shape in fractions 0.5 and 0.25mm from the initial milling (2hrs) making them more rounded/uniformed. For the rest of the fractions in ball milling the shape change seems to be more chaotic even making them more angular/irregular. Table 14 contains side-by-side Mann-Withney and two sample t-test results where gray colored cells indicates if the shape change was recognized by the corresponding statistical method. The two statistical methods agree in the results in 96%. There is no skew evidence between the methods since the lack of shape change recognition occurs in both methods. There is also no skew evidence related to any quantity.

33

Table 14 Shape change compared with initial state Balls Autogenous Size 2hr 6hr 24hr 24hr 72hr 120hr (mm) Q MW t-T MW t-T MW t-T MW t-T MW t-T MW t-T 0.5 2 3 4 5 6 7 8 9 10 11 12 0.25 2 3 4 5 6 7 8 9 10 11 12 0.125 2 ↓ ↓ 3 4 ↓ ↓ 5 ↓ 6 ↓ 7 8 9 ↓ ↓ 10 11 ↓ ↓ ↓ 12 ↓ ↓ ↓ 0.063 2 3 4 ↓ 5 ↓ 6 7 8 ↓ ↓ 9 10 11 ↓ 12 ↓ Q=Quantity number, MW= Mann-Withney test, t-T=two samples t-Test

Autogenous milling (Table 14) seems to have a more smooth transition during the milling time (compare with ball milling fraction 0.5 and 0.25mm) however, the finest fraction behaves in the same chaotic way (compare with ball milling).

4.4 Shear strength in uniformed sized particle Table 15 shows the percentages of broken material after tests depending on the normal load applied to consolidate the sample. The amount of broken particles in size 0.25mm seems to increase. For size 0.063mm is stable for all loads and for 0.125mm size there is a peak at 150kPa with high breakage compare with the rest loads. Graphical display can be seen in

34

Figure 21 where fraction 0.25mm has a more clear increase of breakage in relation with the specimens 0.125 and 0.063mm.

Table 15 Particle fines generated by sample and test load size load % fine

content size Load % fine

content size load % fine

content mm kPa mm kPa mm kPa 0.25 50 26.6 0.125 50 21.0 0.063 50 10.3

100 32.8 100 24.1 100 9.2 150 29.5 150 29.6 150 10.9 300 34.8 300 22.3 300 12.1 500 39.6 500 23.9 500 12.7

Shape change by fraction size is display in Table 16. In relation with the original shape only fraction size 0.063mm have different shape for all quantities while fraction 0.125mm has change in half of them but in both cases for the highest consolidation load 500kPa. Table 17 and Figure 22 show the initial void ratio for the test performed, in this data a higher void ratio is recognized for bigger fraction size. The void ratio diminish as the confining pressure increases.

Figure 21 Fine generation percentages by weight after direct shear test

35

Table 16 Quantity values for particles by size and load state Size

(mm) Q Initial value 50kPa 500kPa

0.063 1 1.513 1.494 1.422 2 0.707 0.703 0.730 3 0.698 0.698 0.729 4 0.934 0.932 0.943

0.125 1 1.472 1.434 1.386 2 0.701 0.698 0.705 3 0.714 0.733 0.750 4 0.933 0.932 0.934

0.25 1 1.452 1.397 1.395 2 0.693 0.682 0.685 3 0.718 0.743 0.739 4 0.928 0.922 0.924

Q=quantity

Table 17 Initial conditions for the shear tests 0.25mm 0.125mm 0.063mm

load initial load initial load initial kPa void ratio kPa void ratio kPa void ratio

0 1.096 0 0.978 0 0.767 50 1.061 50 0.948 50 0.744

100 1.034 100 0.943 100 0.730 150 1.015 150 0.924 150 0.709 300 0.989 300 0.884 300 0.677 500 0.917 500 0.831 500 0.679

Figure 22 Void ratio as a function of normal consolidation in direct shear tests

36

Figure 23 shows the typical behavior of the samples; dilatant for samples 50,100,150 and 300kPa and contractive for samples at 500kPa (black bold line). Lines show the height reduction during the shearing process; all of them except 500kPa present a contractive behavior due to the rearrangement of the particles with a followed increase of height (dilatant behavior) due the overlapping of the particles. Same curves present a secondary height reduction. For 500kPa sample there is only contractive behavior showing no high increase maybe due to the particles do not re-arrange but break. 4.4.1 Suggested empirical relation Table 18 contains laboratory data collected from the samples. This data has been used to estimate the internal friction angle with the help of a regression analysis.

Table 18 Quantity values for the samples

Internal friction

Consolidation kPa

Initial void ratio

Size mm

Quantity values (Q)

angle (φ) Q1 Q2 Q3 Q4 25.8 50 0.745 0.063 1.513 0.707 0.698 0.934 25.1 100 0.731 0.063 1.494 0.703 0.698 0.932 23.3 50 0.948 0.125 1.472 0.701 0.714 0.933 17.9 100 0.943 0.125 1.434 0.698 0.733 0.932 22.1 50 1.062 0.25 1.452 0.693 0.718 0.928 26.7 100 1.034 0.25 1.397 0.682 0.743 0.922

Figure 23 Typical sample height reduction trend during the test for different normal load (0.063mm)

37

Suggested empirical relations are listed below (table 18, 19 y 20); they were obtained from a regression analysis with a 95% of confidence. Additionally to the regression analysis a subset regression has been carried out to identify the factors affecting the empirical model. For all the models the best fit includes all the suggested parameters. Below empirical suggested relations, the table undergoing the empirical relation represents the r-square (fit) when using the available parameters.

Table 19 Empirical relation for Quantity 1 φ = 289 – 0.134 c – 76.3 e + 63.6 s – 134Q1-1

R-square Consolidation (c)

Initial void ratio (e)

Size (s)

Q1-1

56.4 X X 60.5 X X X 56.7 X X X 73.9 X X X X

Table 20 Empirical relation for Quantity 2

φ = 801 -0.1526 c – 43.6 e – 27.9 s – 1036 Q2 R-square Consolidation

(c) Initial void

ratio (e) Size (s)

Q2

56.4 X X 40.0 X X 94.6 X X X 60.5 X X X 96.7 X X X X

Table 21 Empirical relation for Quantity 3 φ = -5 -0,060 c - 61,1 e + 76.4 s + 109 Q3



Size (s)

Q3

56.4 X X 60.5 X X X 56.6 X X X 65 X X X X

Table 22 Empirical relation for Quantity 4 φ = 2026 – 0.1431 c +2 e – 95.3 s – 2129 Q4


Initial void Ratio (e)

Size (s)

Q4

56.4 X X 50.0 X X 93.1 X X X 83.6 X X X 93.1 X X X X

Quantities 1 and 3 are related with the form (first order scale of the Mitchel and soga classification) and quantities 2 and 4 describe the roundness (second order scale) see Figure 1 and Table 1.

39

5 DISCUSSION This thesis was developed to answer the research questions raised in chapter 1. To accomplish the task the entire work could be divided in two faces: Preparation and tailings testing. The preparation face includes a state of the art review completed on the particle shape and its effects (paper I). Furthermore, image analysis test were performed to decide the required resolution to use during the image capture (paper II). These two papers were the starting point for the research to decide and chose methods, equipment and tests to address the research questions. This part of the thesis can be used for anyone interested in the same area as a compilation of previous work and hopefully would solve some questions shortening its taking-decision time. The second part includes all tests performed on tailings material. The main reason to develop a research in tailing particles was the safety of the impoundments; mechanical weathering of the tailing particles needs to be study to determine further changes in the particles and its mechanical properties. The safety of a tailing dam in a long time perspective depends not only on the properties and conditions during the disposal and operation phase of the tailings dam but in a longer perspective on weathering and changes in environmental conditions. The attempt to introduce new empirical relations between tailings conditions and strength could help in the future to develop low-cost early warning indicators compared with the actual expensive and long-time laboratory tests. Even if some questions are addressed some more still remain and moreover new inquires have appeared. During preparation face and literature review 43 quantities were compiled, they cover two and tree-dimensions description of the particles (see paper I). For a mixture of practical and best practices reasons two dimensions image analysis was chosen (availability, avoid subjectivity, fast data collection, fast analysis, etc.). Quantities describe in table 1 are available in two-dimension image analysis. Quantities were tested using Image analysis in different resolution and magnification to understand the possible effects and reliability of the results (paper II). Geometrical figures were used to determine its exact geometric attributes and later compare with image analysis output. Magnification test was also performed to identify the stability of the quantities. Results showed that resolution increase minimize the error. Resolution and magnification are more susceptible to have deviations when length attributes are measured e.g. areas measurements have less deviation than length, but in length attributes when perimeter is measured the error is bigger compare with diameter due to its extension. Quantities were categorized using Mitchell and Soga (2005) classification using the author quantity definition. Since the results have shown that form and roundness could have different effects and also that a large amount of quantities are available in the literature it is suggested by the author to use 2 or 3 form and roundness quantities. The inclusion of more quantities into any research seems to only increase the complexity and interpretation. Unfortunately there is no agreement on which is the best quantity but some of them appear more than others in the scientific literature e.g. Wadell (1932) comparison chart (recently computerized by Zheng & Hryciw, 2015), Quantity 1 (called Aspect ratio by Hawkins, 1993)), Quantity 2 (called circularity by Cox, 1927) among others.

40

Tailings are subject to increasing loads in tailings dams due to the continue raise of the deposits. The coarse fraction is preferred to raise the dikes upon better foundation properties. Tailings had been classified as very angular to sub angular material by visual inspection base on Powers (1953) comparison chart. The results of this classification is in agreement with the conclusion of the general shape of tailings made by Garga, et al. (1984). However the loom of elongated and irregular particles in sizes is bigger than those declared by Mitchell and Soga (2005). Particles involve in this thesis are not natural geological materials but they are crushed and milled rock from the mining industry and this could be the reason to differ. The effect of the incremental load over the tailing particles has been studied in this thesis using oedometer and direct shear tests. The effect of the vertical load on tailings was studied trough the conventional sieving and shape measurement. Results have shown increase in the fines fraction during the incremental load in the direct shear tests (Table 15) due to the breakage. Furthermore fines generation increases as particle size does (bigger particles breaks more than small particles). Shearing produce more fines compared with vertical load (see Table 11 and Table 15) even at lower loads. Fines generation or breakage depends in the rock characteristics as mineralogy, hardness, structure etc. but also by the rigor of the environment (Wentworth, 1922a). Direct shear test is not only vertically loading the particles (normal load) but it is also generating some attrition due to the constant deformation (creeping), this could explain the higher fines generated during the shear test compared with oedometer tests. Shape was also measure during both tests and smaller fractions 0.063mm during oedometer (Table 9) and 0.125mm and 0.063mm in direct shear (Table 16) present changes; they become more regular in form and more rounded. The rounded shape of the particles has a direct effect on the soils dropping their strength (Cho, et al., 2006; Holubec & D'Appolonia, 1973 and Rousé, et al., 2008). Shearing is producing more shape changes in the particles making understandable the higher production of fines in this mechanical process. Degradation of the tailing particles was studied also with the help of a laboratory mill, in this case autogenous and ball mill. The intention was to compare the attrition generated in the mill with the effects of the shearing (direct shear) and loading (oedometer) results. Mill attrition results has shown opposite shape changes respect to load and shearing tests; coarse particles 0.5 and 0.25mm become more rounded and regular in form while smaller fractions are more angular and irregular in form. Also this result is in contrast with other milling studies where ball milling produces more irregular and angular particles compare with the autogenous (Ulusoy, 2008). However materials are different and milling test was designed to erode and not to break particles as regular milling does. It seems like physical erosion depends on the test and test conditions, this is in agreement with Wentworth (1922a) who declared that particle shape depend among other factors in the rigor of the transport, humidity, etc. Shearing and loading test produce rounded material in finer fraction 0.125 and 0.063mm but mill attrition produce more angular particles in same sizes. Particle size seems plays an important role in this size reduction due to the smaller fraction 0.063mm reduces its size less than the rest supportive with the milling theory that says it is needed more energy to reduce smaller particles due to the probability of striking during a regular milling process and the uniformity of the constituent materials (Sponenburgh, 2006).

41

Published empirical relationships between shape and friction angle were evaluated in order to investigate if this approach is valid for tailings. Quantities were automatically quantified by 2D-image analysis (quantity 1, 2, 3 and 4). Despite the different methodologies applied to quantify or to qualitative classify the shape of the particles the trend of the friction angle should be similar if the shape descriptors are correct describing the material of interest. Based on the general behavior of the empirical relations (eq. 2 and 3) the friction angle is likely to be higher in crushed artificial rock than for natural materials since crushed materials in general are considered to be more angular (Garga, et al., 1984). All quantities evaluated describe an underestimation of the reference friction angle from the triaxial tests (see Figure 16). Equation 3 is the most suitable empirical relation to describe the friction angle for the samples. In the same way the value or statistical parameter minimum for the four quantities present the lowest differences (see Table 6). Underestimation of the internal friction angle is possible related with the empirical relations (Table 4), the maximum value obtained from them is 42 degree and only one triaxial result was over this value (considering also the quantity value as zero). It can also explain that the minimum quantity values produce the best agreement with the empirical relations. Since the evaluated empirical relationships has been established on mainly uniformly graded sand and it is not likely that the actual resulting friction angle would be accurate predicted on tailings consisting of both smaller particles and a larger range in grain size distribution. Furthermore the empirical relationships are based on the shape descriptor or quantity that can both been defined in different ways and also be evaluated differently. Cho, et al. (2006) (Equation 2) used the roundness values based on the Krumbein and Sloss (1963) modified chart and Rousé, et al. (2008) (Equation 3) uses the roundness values as Wadell (1932) defined and based on a compilation of various authors. Empirical relations suggested in this thesis are applicable under defined conditions. Data in Table 18 was used to asses a regression analysis. Empirical relations are suggesting that friction angle would decreases if consolidation increase and void ratio decreases. It is evident that the physical changes in void ratio and consolidation are not in agreement with the known, if consolidation (relative density) increases and void ratio decreases friction angle should increase (Holubec and D’Apolonia, 1973). This can be explained if we consider the entire equations. Shape is the main driving parameter in the equations and probably the rest of the parameters are adjusting the model with no physical mining. From this conclusion it is not possible to determine the real physical effect of the consolidation, size and void ratio but shape as a main driving parameter could. Empirical relations results show that when particles become more regular in form (quantities 1 and 3) the friction angle (φ) increases but when they become more rounded (quantities 2 and 4) friction angle (φ) decreases. The methodology used to cast the direct test samples is based in the natural sedimentation process (Dorby, 1991) and it is possible that this could have an effect on the results. Results could suggest that elongated particles like mica minerals are deposited in horizontal layers creating preferential sliding horizons that have reduced strength compared with regular in form particles. Harris, et al. (1984) found decrees in the shear resistance with increase the mica content in regular sands. Furthermore the void ratio increases with mica content (and flat-elongated particles) resulting in a looser and weaker soil

42

(Santamarina & Cho, 2004 and Chen & Lin, 2005). Deposition methods in tailing dams as spigotting could create this preferential layering deposition. The mica minerals presence during further experimentation should be account to understand the effects in tailings. Tailings dam research should consider deposition method effects over the particle settlements and layering in the tailing impounds that could drive the strength of the materials. Roundness increases the soil strength decrease. In this case angular tailing particles are providing more interlocking strength been necessary to break the corners to be able to rearrange the soil structure. Figure 21 shows the fine generation increase while increasing stresses and Figure 23 is the behavior of the height sample during shearing. Considering both (figures 21 and 23) it seems like samples have a primary and secondary consolidation separated for a height sample increase; this height increase could represent a major rearrangement of the particles (overlapping). The increases of stresses in the sample avoid the rearrangement generating more fines. The breakage of the particles is not necessary to produce more angular or rounded material, corners can always break and keep its angularity as direct shear results showed it (table 16 sizes 0.25 and 0.125mm). Particle size increase in literature has been identified as a possible soil strength parameter modifier inherent to the material strengthening the soil in some studies (Lewis, 1956 and Kolbuszewski & Frederick, 1963) but weakening in others (Kirkpatric, 1965 and Marschi, et al., 1972) or even having no effect (Bishop, 1948 and Vallerga, 1957). Particle size influence in the friction angle is ambiguous for the thesis results. Quantities seem to have the main influence on the size. For form quantities, if the size increases the strength increases but for roundness quantities if the size increases the strength decreases. However as it has been discussed in the above paragraphs this thesis can conclude that the particle shape can be considered as a strength parameter for tailing particles. The limited amount of data shrinks the action area of the empirical relations but state an initial relation for further research and new data acquisition. In perspective if particles become more rounded the strength of the tailings could be reduced. If particles are more regular in form the strength should not be compromised. However the reduction of the sizes (fine soil production) could have an effect that it is still unknown. There are still a lot of questions regarding not only on the long-time perspective stability of the tailing dams but also in some unsolved thesis questions e.g. the influence of the particle size. During the work new research questions has been raised but not answered. The opposite effect of the morphology and the roundness of the particles on the strength of the tailings are presumed to be the result of the alignment of some minerals as mica due to the layering during the samples casting. The numbers of inter-particle contacts is a factor that drives the breakage of the particles and during this thesis uniformed particle size were used. Thus, effect of mica and the use of different particle size distributions are suggested for further research.

43

6 SUMMARY AND MAIN CONCLUSIONS The study concerns the degradation of tailings due mechanical agents. Mechanical degradation was conducted in laboratory by milling, oedometer compression and shearing. Tailings divided by sizes were studied by breakage and shape changes using regular sieving and image analysis respectively. Several shape descriptors are found in literature and but available quantities were classified according with the scale dependence (see Table 1). In the following, the addressed research questions stated in the section 1.1 are answered; then some major conclusions from the study base on papers are highlighted:

1. What is the effect of the vertical load in the tailing particles? Particles breakage was identified in all samples. The extent of breakage decreases as the particle size decreases. The breakage of the particles decreases the size but not the shape, except for the finest studied particles. Particle fraction 0.063mm becomes more irregular and angular by breakage.

2. Degradation processes (load, shearing and wearing) act in a different way? Yes they do, all agents break the particles but the shape changes are not similar, wearing in general produce more rounded/regular particles in all sizes while loading and shearing only act over smaller fraction (0.063mm).

3. Is it possible to apply empirical relations found in literature to relating shape and friction angle for tailings?

It is possible to apply the studied empirical relations. But quantities used and applied in the empirical relations underestimate the tailings shear strength.

4. Is possible to obtain an empirical relation among tailing parameters and the internal friction angle?

Four empirical relations were suggested in this thesis. Two of them are related to form quantities and two in roundness quantities. They are based on limited amount of data and needs further development and replantation.

5. What is the effect of tailing particle size in the internal friction angle? Unknown. Contrarious results were obtained when particle size was evaluated in the empirical relation suggested. If quantities related to the form are applied in the empirical relation then when the particle size increases the friction angle increases. But if quantities related to the roundness are applied in the empirical relation the increase in particle size decreases the friction angle.

6. Which are the available shape descriptors? Shape descriptors are summarized in Paper I

44

7. What shape descriptors are available to use with the two dimensional image analyses?

Available shape descriptors for two dimensions image analysis are presented in Table 1

8. What is the influence of resolution and magnification on the shape descriptors? Resolution increase reduces the error but also quantities related with the perimeter are more susceptible to error increase. The major conclusions based on this work are: Tailing particles are more irregular in form as the particle size decreases for particles < 0.063mm while literature suggests it should be under clay sized (0.002mm). All physical degradation processes the tailings were subject to in this work produced a distinctive breakage and shape. Shape changes are not always producing more rounded particles especially in the sizes 0.125 and 0.063mm but more angular during mill attrition. More rounded material was obtained in smaller fraction (0.125 and 0.063mm) during shearing and loading. Larger fractions (0.5 and 0.25mm) present no change. Empirical relations were suggested to relate the normal load, void ratio, particle size and particle shape to the internal friction angle. According to the empirical relation suggested the regular morphology and decrees of roundness of the tailing particles increase the friction angle. Effects of the size on the internal friction angle were not possible to determine. Tailing are more susceptible to settlements under static loading compared to natural geological materials.

45

7 SUGGESTED FURTHER WORK Since generated fine particles seems to have influence in the final breakage and probably shape it would be of interest to set experiments with a configured size distribution. If mica is creating slippery surfaces while shearing it could be suggested to test tailings at variable mica content. Obtain the shape configuration for the fines produced during shearing and loading. This may require more sophisticated equipment depending in the ultimate size wanted to evaluate.

47

8 BIBLIOGRAPHY Arasan, S., Hasiloglu, A. S. & Akbulut, S., 2010. Shape particle of natural and crushed aggregate using image analysis. International Journal of Civil and Structural Engineering, 1(2), pp. 221-233. Armengot, J., Espi, J. & Vazquez, F., 2006. Origenes y desarrollo de la mineria. Industria y mineria, Volume 365, pp. 17-28. Aschenbrenner, B., 1956. A new method of expressing particle sphericity. Journal of Sedimentary Petrology, 26(1), pp. 15-31. Axelsson, K., 1998. Introduktion till jordmekaniken jämte jordmaterialläran, Luleå: Skrift 98:4, Luleå: Avdelningen för Geoteknologi, Luleå Tekniska Universitet. Bhanbhro, R. et al., 2013. Basic description of tailings from Aitik focusing on mechanical behavior. International Journal of Emerging Technology and Advanced Engineering, 3(12), pp. 65-69. Bishop, A., 1948. A large shaer box for testing sand and gravels. Proceedings of the 2nd international conference of soil mechanics and foundation engineering, Volume 1, pp. 207-211. Blott, S. & Pye, K., 2008. Particle shape: a review and new methods of characterization and classification. Sedimentology, Volume 55, pp. 31-63. Chang, L. C. & Page, N. W., 1997. Particle fractal and load effects on internal friction in powders. Powder Technology, Volume 90, pp. 259-266. Chen, Chang & Lin, 2005. Influence of coarse aggregate shape on the strength of asphalt concrete mixtures. Journal of Eastern Asia Society for transportation studies, Volume 6, pp. 1062-1075. Cho, G., Dodds, J. & Santamarina, J., 2006. Particle shape effects on packing density, stiffness and strength: natural and crushed sands. Journal of Geotechnical and Geoenvironmental Engineering, 132(5), pp. 591-602. Cox, E. P., 1927. A method of assigning numerical and percentage values to the degree of roundness of sand grains. 1(3), pp. 179-183. Dietrich, W. E., 1982. Settling velocity of natural particles. Water Resources Research, 18(6), pp. 1615-1626. Dorby, R., 1991. Soil properties and earthquake ground response.. Florence, Italy, s.n. EPA, 1994. Extraction and beneficiation of ores and minerals, Volume 3: IRON, Washington: U.S. Environmental Protection Agency, EPA 530-R-94-030, NTIS PB94-195203. Fernlund, J., 2005. Image analysis method for determining 3-D shape of coarse aggregate. Cement and Concrete Research, 35(8), pp. 1629-1637. Folk, R., 1955. Student operator error in determining of roundness, sphericity and grain size. Journal of Sedimentary Petrology, Volume 25, pp. 297-301. Garga, V., ASCE, M. & McKay, L., 1984. Cyclic triaxial strength of tailings. Journal of Geotechnical Engineering, 110(8), pp. 1091-1105. Harris, W., Parker, J. & Zelazny, L., 1984. Effects of mica content on engineering properties of sand. Soil Science Society of America, 48(3), pp. 501-505. Hawkins, A., 1993. The shape of powder-particle outlines. New York: Wiley.

48

Hayati, A. N., Ahmadi, M. M. & Mohammadi, S., 2012. How particle shape affects the flow through granular materials. American Physical Society, Physical review E 85(036310), pp. 036310-1 36310-4. Head, K. & Epps, R., 2011. Manual of soil Laboratory testing. 3rd ed. Scotland, UK: Whittles Publishing. Holubec, I. & D'Appolonia, E., 1973. Effect of particle shape on the engineering properties of granular soil. ASTM SPT, Volume 523, pp. 304-318. Image Pro Plus v. 7.0, 2011. http://www.mediacy.com/. [Online]. ImageJ, 1., 2013. version 1.47v. Wayn Rasband, National Institutes of Health, USA.. [Online] Available at: http://imagej.nih.gov/ij/ Janoo, V. C., 1998. Quantification of shape, angularity, and surface texture of base course materials, s.l.: s.n. Jefferies, M. & Been, K., 2000. Soil liquefaction. A critical state approach. London and New York: Taylor & Francis Group. Jimenez, J. A. & Madsen, O. S., 2003. A simple formula to estimate settling velocity of natural sediments. Journal of Waterway, 129(2), pp. 70-78. Johansson and Vall, 2011. Friktionsjords kornform: Inverkan på geotekniska egenskaper, beskrivande storheter och bestämningsmetoder. Luleå: Avdelningen för Geoteknologi, Institutionen för Samhällsbyggnad och naturresurser, Luleå tekniska universitet. Johnson, N. L., 1949. System of frequency curves generated by methods of translation. Biometrika, 36(1/2), pp. 149-176. Kane, J. W. & Sternheim, M. M., 1988. Physics. Third edition. ed. s.l.:John Wiley & Sons. Kirkpatric, W., 1965. Effects of grain size and grading on the shearing behavior of granular materials. Proceedings of the 6th International Conference on Soil Mechanics and Fundation Engineering, Volume 1, pp. 273-277. Kolbuszewski, J. & Frederick, M., 1963. The significance of particle shape and size on the mechanical behavior of granular materials. European Conference of soil mechanics and foundation engineering, Volume sec.4 paper9, pp. 253-263. Krumbein and Pettijohn, 1938. Manual of Sedimentary Petrology. New York: Appleton-Century Crofts. Krumbein and Sloss, 1963. Stratigraphy and Sedimentation. 2nd ed. San Francisco: W.H. Freeman. Krumbein, W., 1941. Measurement and geological significance of shape and roundness of sedimentary particles. Journal of Sedimentary Petrology, 11(2), pp. 64-72. Kuo, C.-Y., Rollings, R. & Lynch, L. N., 1998. Morphological study of coarse aggregates using image analysis. Journal of Materials in Civil Engineering, 10(3), pp. 135-142. Lewis, J., 1956. Shear strength of rockfill. Proceedings of the 2nd Australia-New Zeland conference of soil mechanics and foundation engineering, pp. 50-52. Lincoln, E., 2014. Wilcoxon-Mann-Whitney test: Definition and Example. s.l.:John Wiley and Sons. Lindvall, M. and Eriksson, N., 2003. Investigation of the weathering properties of tailings sand from Boliden´s Aitik copper mine – a summary of twelve years of investigations. Cairns, s.n.

49

Marschi, N., C.K., C. & Seed, H., 1972. Evaluation of properties of rock fill materials. journal of the soil mechanics and foundation division, 98(1), pp. 95-114. Mitchell and Soga, 2005. Fundamentals of soil behavior. 3rd ed. s.l.:Wiley. Mora and Kwan, 2000. Sphericity, shape factor, and convexity measurement of coarse aggregate for concrete using digital image processing. Cement and Concrete Research, 30(3), pp. 351-358. Nearing, M. & Parker, S., 1994. Detachment of soil by flowing water under turbulent and laminar conditions. Soil Science Society of American Journal, 58(6), pp. 1612-1614. Northey, S. et al., 2014. Modelling future copper ore grade decline based on a detailed assessment of copper resources and mining. Resources, Conservation and Recycling, Volume 83, pp. 190-201. Ormann, L., Zardari, M., Mattsson, H. & Knutsson, S., 2013. Numerical Analysis of strengthening by rockfill embankments on an upstream tailings dam. Canadian Geotechnical Journal, Volume 50, pp. 391-399. Pentland, A., 1927. A method of measuring the angularity of sands. Acta Eng. Dom. Transaction of the Royal Society of Canada, 21(Ser.3:XCIII). Persson, A.-L., 1998. Image analysis of shape and size of fine aggregates. Egineering Geology, Volume 50, pp. 177-186. Powers, M., 1953. A new roundness scale for sedimentary particles. Journal of Sedimentary Petrology, 23(2), pp. 117-119. Qazi, M., 1975. Flow properties of granular masses: A review on the angle of repose. The Arabian Journal for Science and Engineering, 1(2). Rico, M., Benito, G., Salgueiro, A. & Diez-Herrero, A. a. P. H., 2008. Reported tailing dam failures. Areview of the european incidents in the worldwide context. Journal of jazarduos materials, Volume 152, pp. 846-852. Riley, A. N., 1941. Projection Sphericity. Journal of Sedimentary Petrology, 11(2), pp. 94-97. Rodriguez, J. M., 2012. Particle shape quantities and influence on geotechnical properties: a review, Lulea, Sweden: Lulea University of Technology. Rousé, P., Fennin, R. & Shuttle, D., 2008. Influence of roundness on the vid ratio and strength of uniform sand. Geotechnique, 58(3), pp. 227-231. Santamarina, J. & Cho, G., 2004. oil behaviour: The role of particle shape. London, Skempton Conf. London.. Schofield & Wroth, 1968. Critical state soil mechanics. s.l.:McGraw Hill. SGF, 1996. Allmänna råd och metodbeskrivningar, Svenska Geotekniska Föreningen, SGF, Linköping (1996) (Rapport 1:96. 147 pp., in Swedish). s.l.:s.n. SGF, 2014. Direkta skjuvförsök – en vägledning (in Swedish), Linköping: Swedish Geotechnical Society, SGF Notat 2:2004, SGF:s Laboratoriekommitté. Shinohara, K., Oida, M. & Goldman, B., 2000. Effect of particle shape on angle of internal friction by triaxial compression test. Powder Technology, Volume 107, pp. 131-136. Snedecor, G. W. & Cochran, W. G., 1989. Statistical methods. 8th ed. Iowa: State University Press. Sperry, J. M. & Peirce, J., 1995. A model for estimating the hydraulic conductivity of granular material based on grain shape, grain size and porosity. Ground Water, 33(6), pp. 892-898.

50

Sponenburgh, L. E., 2006. Theroy and practice for the amateur pyrotechnician. Bunnell, Florida: Bridge City Enterprises. Sukumaran, B. & Ashmawy, A., 2001. Quantitative characterisation of the geometry of discrete particles. Geotechnique, 51(7), pp. 619-627. Tickell, F. G., 1938. Effect of the angularity of grain on porosity and permeability. bulletin of the American Association of Petroleum Geologist, Volume 22, pp. 1272-1274. Ulusoy, U., 2008. Application of ANOVA to image analysis results of talc particles produced by different milling. Powder Technology, 188(2), pp. 133-138. Vallerga, B., 1957. Effect of shape, size and surface roughness of aggregate particles on the strength of granular materials. ASTM Special technical report No. 212. Wadell, H., 1932. Volume, shape and roundness of rock particles. Journal of Geology, Volume 40, pp. 443-451. Wadell, H., 1935. Volume, shape, and roundness of quartz particles. Journal of Geology, Volume 43, pp. 250-279. Wentworth, W., 1922a. A method of measuring and plotting the shape of pebbles. U.S. Geological Survey Bulletin, Volume 730C, pp. 91-114. Wentworth, W., 1922b. The shape of the beach pebbles. U.S. Geological Survey Bulletin, Volume 131C, pp. 75-83. Yoginder, P., Jing, C. & Hadai, T., 1985. Confining pressure, grain angularity and liquefaction. Journal of geotechnical engineering, 111(10), pp. 1229-1235. Youd, T., 1973. Factors controlling maximum and minimum densities of sands, "evaluation of relative density and its role in geotechnical projects involving cohesionless soils". ASTM STP, Volume 523, pp. 98-112. Zheng, J. & Hryciw, R., 2015. Ttraditional soil particle sphericity, roundness and surface roughness by computational geometry. Geotechnique, 65(6), pp. 494-506.

51

APENDIX Quantity Description Graphic description 1* Major axis/Minor axis

2* 4πArea(A)/Perimeter2(P)

3* 4Area(A)/πMajor axis2 See figures in quantities 1 and 2 4* Area(A)/Convex Area(Ca)

5 Fractal dimension use 'strides' (minimum step lengths) of various sizes. The fractal

dimension is calculated as 1 minus the slope of the regression line obtained when plotting the log of the perimeter (for various strides) against the log of the stride length. (more info in imageproplus)

6* Square root of [Maximum inscribed (Di)/Minimum circumscribed(Dc)], circle

diameters

continue….

52

7* Diameter of a circle same area as particle(Da)/Minimum circumscribed circle diameter(Dc)

8* Perimeter2(P)/Area(A) See figure in quantity 2 9* Perimeter of a circle with same area

(Pa)/Perimeter(P)

10* Area(A)/Area of the minimum

circumscribed circle (Ac)

11* Perimeter/Convex perimeter

12 Perimeter(P)/πAverrage Feret

Average feret box is obtained rotating two parallel lines (two degrees each time) and measuring the distance, finally the average feret is the average distance of all the feret boxes distance measured

* Figures were taken and modified from Johansson and Vall (2011) 1 and 3 are the inverse for individual values (ImageJ, 2013), similar values can result from quantities 11 and 12 (Kuo, et al., 1998).

APPENDED PAPERS

Paper I Rodriguez, Juan M.; Edeskär, Tommy and Knutsson, Sven. (2013). Particle Shape Quantities and Measurement Techniques – A Review. Electronical Journal of Geotechnical Engineering, Vol. 18/A. pp. 169-198.

- 169 -

Particle Shape Quantities and Measurement Techniques–A Review

Juan M. Rodriguez Ph.D. Student

Department of Civil, Environmental and Natural resources engineering, Luleå University of Technology, Tel: +46 920 491523, SE – 971 87, Luleå, Sweden

e-mail: [email protected]

Tommy Edeskär Assistant Professor

Department of Civil, Environmental and Natural resources engineering, LuleåUniversity of Technology, Tel: +46 920 493065, SE – 971 87, Luleå, Sweden


Sven Knutsson Professor

Department of Civil, Environmental and Natural resources engineering, LuleåUniversity of Technology, Tel: +46 920 491332, SE – 971 87, Luleå, Sweden


ABSTRACT It has been shown in the early 20th century that particle shape has an influence on geotechnical properties. Even if this is known, there has been only minor progress in explaining the processes behind its performance and has only partly implemented in practical geotechnical analysis. This literature review covers different methods and techniques used to determine the geometrical shape of the particles. Particle shape could be classifying in three categories; sphericity - the overall particle shape and similitude with a sphere, roundness - the description of the particle’s corners and roughness - the surface texture of the particle. The categories are scale dependent and the major scale is to sphericity while the minor belongs to roughness. The overview has shown that there is no agreement on the usage of the descriptors and is not clear which descriptor is the best. One problem has been in a large scale classify shape properties. Image analysis seems according to the review to be a promising tool, it has advantages as low time consumption or repeatability. But the resolution in the processed image needs to be considered since it influences descriptors such as e.g. the perimeter. Shape definitions and its potential role in soil mechanics are discussed. KEYWORDS: Particle shape, Quantities, Image analysis.

Vol. 18 [2013], Bund. A 170

INTRODUCTIONEffects on soil behavior from the constituent grain shape has been suggested since the earliest

1900’s when Wadell (1932), Riley (1941), Pentland (1927) and some other authors developed their own techniques to define the form and roundness of particles. Into the engineering field several research works conclude that particle shape influence technical properties of soil material and unbound aggregates (Santamarina and Cho, 2004; Mora and Kwan, 2000). Among documented properties affected by the particle shape are e.g. void ratio (porosity), internal friction angle, and hydraulic conductivity (permeability) (Rousé et al., 2008; Shinohara et al.,2000; Witt and Brauns, 1983). In geotechnical guidelines particle shape is incorporated in e.g. soil classification (Eurocode 7) and in national guidelines e.g. for evaluation of friction angle (Skredkommisionen, 1995). This classification is based on ocular inspection and quantitative judgment made by the individual practicing engineer, thus, it can result in not repeatable data. The lack of possibility to objectively describe the shape hinders the development of incorporating the effect of particle shape in geotechnical analysis.

The interest of particle shape was raised earlier in the field of geology compared to geotechnical engineering. Particle shape is considered to be the result of different agent’s transport of the rock from its original place to deposits, since the final pebble form is hardly influenced by these agents (rigor of the transport, exfoliation by temperature changes, moisture changes, etc.) in the diverse stages of their history. Furthermore, there are considerations regarding on the particle genesis itself (rock structure, mineralogy, hardness, etc.) (Wentworth 1922a). The combination of transport and mineralogy factors complicates any attempt to correlate length of transport and roundness due that soft rock result in rounded edges more rapidly than hard rock if both are transported equal distances. According to Barton & Kjaernsli (1981), rockfill materials could be classified based on origin into the following (1) quarried rock; (2) talus; (3) moraine; (4) glaci-fluvial deposits; and (5) fluvial deposits. Each of these sources produces a characteristic roundness and surface texture. Pellegrino (1965) conclude that origin of the rock have strong influence determining the shape.

To define the particle form (morphology), in order to classify and compare grains, many measures has been taken in consideration (axis lengths, perimeter, surface area, volume, etc.). Furthermore, corners also could be angular or rounded (roundness), thus, the authors also focus on develop techniques to describe them. Additionally corners can be rough or smooth (surface texture). Nowadays some authors (Mitchell & Soga, 2005; Arasan et al., 2010) are using these three sub-quantities, one and each describing the shape but a different scale (form, roundness, surface texture).

During the historical development of shape descriptors the terminology has been used differently among the published studies; terms as roundness (because the roundness could be apply in the different scales) or sphericity (how the particle approach to the shape of a sphere) were strong (Wadell, 1933; Wenworth, 1933; Teller, 1976; Barrett 1980; Hawkins, 1993), and it was necessary in order to define a common language on the particle shape field; unfortunately still today there is not agreement on the use of this terminology and sometimes it make difficult to understand the meaning of the authors, that’s why it is better to comprehend the author technique in order to misinterpret any word implication.

Several attempts to introduce methodology to measure the particle’s shape had been developed over the years. Manual measurement of the particles form is overwhelming, thus, visual charts were developed early to diminish the measuring time (Krumbein, 1941, Krumbein and Sloss, 1963; Ashenbrenner, 1956; Pye and Pye, 1943). Sieving was introduced to determine

Vol. 18 [2013], Bund. A 171

the flakiness/elongation index but it is confined only for a certain particle size due the practical considerations (Persson, 1988). More recently image analysis on computer base has been applied on sieving research (Andersson, 2010, Mora and Kwan, 2000, Persson, 1998) bringing to the industry new practical methods to determine the particle size with good results (Andersson, 2010). Particle shape with computer assisted methods are of great help reducing dramatically the measuring time (Fernlund, 2005; Kuo and Freeman 1998a; Kuo, et al., 1998b; Bowman, et al.,2001).

In the civil industry e.g. Hot Asphalt mixtures (Kuo and Freeman, 1998a; Pan, et al., 2006), Concrete (Mora et al., 1998; Quiroga and Fowle, 2003) and Ballast (Tutumluer et al., 2006) particle’s shape is of interest due the material’s performance, thus, standards had been developed (e.g. EN 933-4:2000 Tests for geometrical properties of aggregates; ASTM D 2488-90 (1996) Standard practice for description and identification of soils).

Sieving is probably the most used method to determine the particle size distribution. This traditional method, according to Andersson (2010) is time consuming and expensive. Investigations shows that the traditional sieving has deviations when particle shape is involve; the average volume of the particles retained on any sieve varies considerably with the shape (Lees, 1964b), thus, the passing of the particles depend upon the shape of the particles (Fernlund, 1998). In some industries the Image analysis is taking advantage over the traditional sieving technique regardless of the intrinsic error on image analysis due the overlapping or partial hiding of the rock particles (Andersson, 2010). In this case the weight factor is substitute by pixels (Fernlund et al.,2007). Sieving curve using image analysis is not standardized but after good results in the practice (Andersson, 2010) new methodology and soil descriptions could raise including its effects.

Describing the particle’s shape is the main objective, there are 42 different quantities in this document, and it is required to review the information about them to comprehend and interpret the implication of each quantity to determine them usability and practice.

DESCRIPTION OF SHAPE PROPERTIES Particle shape description can be classified as qualitative or quantitative. Qualitative describe

in terms of words the shape of the particle (e.g. elongated, spherical, flaky, etc.); and quantitative that relates the measured dimensions; in the engineering field the quantitative description of the particle is more important due the reproducibility.

Quantitative geometrical measures on particles may be used as basis for qualitative classification. There are few qualitative measures in contrast with several quantitative measures to describe the particle form. Despite the amount of qualitative descriptions none of them had been widely accepted; but there are some standards (e.g., ASTM D5821, EN 933-3 and BS 812) specifying mathematical definitions for industrial purposes.

Shape description of particles is also divided into two: -3D (3 dimensions): it could be obtained from a 3D scan or in a two orthogonal images and -2D (2 dimensions) or particle projection, where the particle outline is drawn.

3D and 2D image analysis present challenges itself. 3D analysis requires a sophisticated equipment to scan the particle surface and create the 3D model or the use of orthogonal images and combine them to represent the 3 dimensions. The orthogonal method could present new challenges as the minimum particle size or the placing in orthogonal way of the particles (Fernlund, 2005). 2D image analysis is easy to perform due the non-sophisticated equipment

V

requir2D imintermor ran1941;

Indefiniusingmorphdefine

Adescridescrishapedescricornethe paFigurRegarthe sasmalleanton

Wthe p

Vol. 18 [20

red to take pimage analysimediate axis lndom, some ; Hawkins, 19

n order to desitions used inthree sub-qu

hology/form, ed.

FigurAt large scaleibing terms aiption at large

e is marked iption of the rs and edges article’s boune 1. A generding the smaame kind of aner scale. Surf

nyms are summ

Table 1:

Wentworth in particle dime

13], Bund.

ictures (e.g. rs the particlelie more or leauthors publi

993).

scribe the parn the literatureantities; one aroundness an

re 1: Shape d the particle’

as spherical, pe scale is sphwith the daspresence ofof different s

ndary, deviatierally accepteallest scale, tenalysis as theface texture imarized in tab

Sub-quantitScaleLarge scalIntermediaSmall scale

1922 (Blott aensions, this

. A

regular camere is assumedess parallel toish their own

SCALE Drticle shape ie. Some authand each descnd surface te

describing su’s diameters platy, elonga

hericity (antonshed line in

f irregularitiessizes are idenions are founed quantity erms like roue one describeis often used ble 1.

ties describin

eate scale e

FOand Pye, 200

consisted o

ra or the use d to lay overo the surface n preferences

DEPENDEin detail, therhors (Mitchellcribing the shexture. In Fig

ub quantitiesin different d

ated etc., are nym: elongat

Figure 1. As. Depending

ntified. By dond and valuatefor this scalgh or smooth

ed above, but to name the

ng the particQuantity Sphericity Roundness Roughness

RM (3D)8), was proba

on the obtain

of microscopr its more stawhile the sho

s about this i

ENCEre are a numbl & Soga, 200

hape but at difgure 1 is show

s (Mitchell &directions areused. An oft

tion). GraphicAt intermediag on at what oing analysis ied. The mentle is roundnh are used. This applied wiactual quant

cle’s morphoAntonymElongatioAngulariSmoothn

)ably one of thning of the

pe for smalleable axis (e.gortest axis is issue (Wadell

ber of terms, 05; Arasan etfferent scales.wn how the s

& Soga, 200e considered.ten seen quancally the consate scale it scale an ana

inside circlestioned circles

ness (antonymhe descriptor ithin smaller ctity. The sub-

ology and itsmonityness

the first autholength of

17

er particles). Ig. longest anperpendiculal, 1935; Riley

quantities ant al., 2010) ar. The terms arscale terms ar

5) At this scalntity for shapsidered type ois focused oalysis is dons defined alons are shown im: angularity

is considerincircles, i.e. at-quantities an

antonym

ors on measurthe tree axe

72

Inndar) y,

ndrerere

le, peofone;

ngin

y).ngt a nd

rees

V

perpespheri

Figu

Kthis iparticc/b arhe cal

Figu

Wschemsame

Vol. 18 [20

ndicular amoicity (Equatio

ure 2: Measu

Krumbein (194s done by m

cle, it can be sre located in tlled (See Figu

ure 3: Detail

Wadell (1932)matic represen

volume.

13], Bund.

ong each otheron 1).

urement of th

41) develop ameasuring theseen in Figurthe chart deveure 3). This ch

ed chart to d

) defined the ntation of the

. A

r (see Figure

he 3 axes pe

a rapid methoe longest (a), re 2 (Always eloped by his hart is an easy

determining

sphericity as sphere surfac

Ψ

2) on the tree

erpendicular

od for shape mmedium (b)

perpendicularown where it

y graphical w

Krumbein in

the specific ce and particl

c2ba +=Ψ

e dimensions

among each

measurement ) and shorterar among eacht can be foun

way to relate th

ntercept sphe

surface ratio e surface, bot

(where a b c

h other (Krum

to determine r (c) axes diah other). The

nd the Intercephe dimension

ericity (Krum

(Equation 2)th particle an

17

c) to obtain th

(

mbein, 1941

the sphericityameters of thradios b/a an

pt sphericity ans.

mbein 1941)

). Figure 4 is d sphere of th

73

he

1)

)

y;hendas

).

ahe

V

F

Tdeclar

WEquat

WEquat

Vol. 18 [20

Figure 4: Sam

This way to res, due the d

Wadell (1934)tion 3 (see Fig

Figure 5:

Wadell (1934tion 4).

13], Bund.

me volume s

obtain the sdifficulty to ge

) also definedgure 5):

Relation betcircumsc

4) used a new

. A

sphere surfacJohansson

phericity is et the surface

d the spheric

tween the vocribed sphere

w formula sim

ce (s) and pan and Vall, 2

almost impoarea on irreg

city based up

olume of the e (Johansson

mple to manag

3

CIR

P

VV=Ψ

CIR

SV

DD=Ψ

article surfac2011).

ossible to acgular solids.

pon the partic

particle andn and Vall, 2

ge using the d

ce (S). (mod

hieve, as Ha

cle and spher

d the volume2011).

diameters (se

17

(2

dified after

awkins (1993

re volumes, a

(3

e of the

e Figure 6 an

(4

74

2)

3)

as

3)

nd

4)

V

Figu

Zeasy tsummthe Fi

Inseen,

Vol. 18 [20

ure 6: The reof a spher

Zingg (Krumbto find out th

marized on Figigure 3.

Figure 7:

n Figure 8 thit is an easy w

13], Bund.

elation betwre of the samein, 1941) de

he main formgure 7. Zingg

Zingg’s cla

e Figures 3 away to unders

. A

een the diamme volume asevelop a class

m of the particg’s classificat

ssification o(Krum

and 7 are comstand the mor

meter of a cirs the particleification base

cles as a disktion is related

of pebble shambein 1941)

mbined, the rrphology rega

rcumscribede (Johanssoned on the 3 axks, spherical, d with Krumb

ape based on).

elation in thearding on the

d sphere and n and Vall, 2xes relation, iblades and r

bein intercept

n ratios b/a a

e two classifia, b and c dim

17

the diameter2011).in this way it rod-like; this

sphericity an

and c/b

ications can bmensions

75

r

isis

nd

be

V

Pycompspheriby Krwith eFolkProjec

InSpher

FpublisDobk

Vol. 18 [20

Figure 8:

ye and Pye (are the Wadicity” based orumbein (194exception of in (1958) d

ction Spheric

n a similar waricity”

orm or shapeshed on 1949

kins & Folk (o

=Ψ

13], Bund.

Classificati

(1943), in thdell’s sphericon an ellipse,41). Axis meaEquation 8 w

describes a rity” (Equation

ay Ashenbren

e factor name, Williams (sh

oblate-prolate

(c/b)(11+=

. A

ion made by (Krumbein

e article “sphcity develope, this last Equasurement is where the origrelation betwn 6).

nner (1956) sh

s are used byhape factor, e index, eq. 11

6))a/b(

( 12,8 3

++

∗

Zingg’s andn and Sloss,

hericity detered in 1934 (uation (numbdone as Figuginal docume

ween the tree

howed his Eq

y authors like eq. 9) in 19651) in 1970 (Bl

b/c(16

/b()b/c( 2

+∗

∗

d chart to det1963)

rminations of(based on th

ber 5) appearsure 1 denotesent was not pe dimensiona

quation (7) at

Corey (shap5, Janke (formlott and Pye,

a/b((1)b

)a2 +

termine sphe

f pebbles andhe diameter) s two years es for Equationpossible to obal axes calle

t that time nam

pe factor, eq. m factor, eq. 1

2008).

))a 2

17

ericity

d sand grainswith “Pebb

early publishens 5 trough 1btain. Sneed &ed “Maximum

(5

(6

med “Workin

(7

8) in the pape0) in 1966 an

(8

76

s”leed12&m

5)

6)

ng

7)

ernd

8)

V

Asquare

T

Tliteratthe pa

Vol. 18 [20

Aschenbrennere of the midd

Table 2: Gen

Aspect Sphericity (3D)

The techniqueture some wayarticle project

13], Bund.

r (1956) devedle one.

neral overviehas been

Name

FlatnessTrue SpOperati

SphericZingg’sIntercepPebbleCorey sWorkinshape faMaximusphericiWilliamJanke fo

Oblate-p

e to measure ys to measuretion, some aut

. A

eloped the sha

ew over diffn compiled a

s index phericity onal sphericity

citys classificationpt sphericity chsphericity

shape factorng sphericity actorum projection ity

ms shape factororm factor

prolate index

FOthe sphericit

e the “two dimthors named “

ape factor by

ferent particland arranged

Author

WentwoWadell

y Wadell

Wadell Zingg’s1

hart KrumbePye and Corey2

AshenbrAshenbr

Sneed &r William

Janke2

Dobkins

RM (2D)ty is based inmensions sph“particle outli

using the rela

le shape defid chronologic

orth

1

ein d Pye

rennerrenner

& Folkms2

s & Folk 1) K2) B

)n three dime

hericity” whicine” or “circu

ation of the tr

initions for 3cally

Year Base

1922a 3-axe1932 Surfa1932 Volu

1934 Sphediam

1935 3-axe1941 3-axe1943 3-axe1949 3-axe1956 3-axe1956 3-axe

1958 3-axe1965 3-axe1966 3-axe

1970 3-axeKrumbein and SlosBlott and Pye, 2008

ensions, it cach is simply thularity”.

17

(9

(10

(1

ree axis but th

(12

3D sphericity

ed on

esace ume ere

metereseseseseses

eseses

esss, 1963 8

an be found ihe perimeter o

77

9)

0)

1)

he

2)

y

inof

V

W(Equathe mEquatcircumof the

Torientand th

Sown Eequal(P) timorient

RpropowerehandlWade18). Hperim

JaEquatdeviatdetermradial

Vol. 18 [20

Wadell in 193ation 4) to a 2

maximum crotions (13) shomscribed circe perimeter of

Tickell in 193tation proposehe area of sma

ome other auEquations as to longest len

me a constantation of the g

Riley (1941), osed by the ab

not computee called “ins

ell and the relHorton 1932 (

meter of a circl

anoo in 1998tion 20. Sukution of the gmine the iteml divisions).

13], Bund.

35 (Hawkins2D outline. Hss sectional aow the relatiole (DC). He a

f a circle of sa

31 (Hawkins,ed was a randallest circums

uthors has bePentland (192ngth outline (

nt, Equations grains.

realize the pbove authors

er, all was mscribed circlelation of diam(Hawkins, 19le of the same

(Blott and Pumaran and Aglobal particlems used in the

. A

, 1993) adopHe defined anarea (outline on between dlso used the t

ame area (PC)

, 1993) useddom one. It iscribed circle

en working w27) relating th(AC2), and Co16 and 17 re

problems that can carry as

made by hand sphericity”.

meters of insc993) use the re area as drain

Pye, 2008) deAshmawy (200e outline frome Equation 2

pt a conversin orientation o

of the particdiameters of aterm “degree and the actua

d his empirics described b(AC).

with the “cirche area (A) oox (Riley, 194espectively. B

t an area, pes the time con

d), and that’s He used the

cribed (DI) anrelation of thenage basin (P

evelop his ge01) develop hm a circle. F1 (N, referred

ion of his 19on the particlcle projectinga circle of saof circularityal particle per

cal relation (Eby the ratio be

cularity” concoutline and ar41) with the rBoth authors

erimeter and nsuming andwhy he dev

e same particnd circumscre drainage baPCD), see Equa

eneral ratio ofhis own shapFigure 9 can d to the num

934 3D spheles and they g the maximuame area (DAy” (Equation rimeter (P).

Equation 15)etween the ar

cept and hadrea of a circleratio area (A)did not defin

some other d tedious worvelop this Eqcle orientationribed (DC) cirasing perimeteation 19.

f perimeter (Pe factor (SF) be used as

mber of sampl

17

ericity formuwere based oum area). Th

A) and smalle14) as the rati

). The particrea outline (A

(13

(14

(15

d develop theme with diamete) and perimetene any defini

(16

(17

measuremenrk (at that timquation easy tn proposed brcles (Equatioer (PD) and th

(18)

(19)

P) to area (Adefined as tha reference t

les intervals o

78

la onheestio

le A)

3)

4)

5)

mererte

6)

7)

ntsmetobyonhe

)

)

A),hetoor

V

F

Rbesidesugge

Rconfigconce1980)superi

Wroundcircleused Eon the

Vol. 18 [20

Figure 9: De

Roundness as e the corners ested many w

Roundness is cguration and erning about t), is describe imposed in th

Wadell (1935)dness of a par diameter (RmEquation 23 (e results (Wad

13], Bund.

escription of (S

ROUdescribed pand how theyays to describ

clearly underdenotes the

the sharpnessas the third o

he corners, an

) describes hirticle using thmax-in), see (N, is the numdell, 1935).

. A

the SukumaSukumaran a

UNDNESSreviously is y are, this wabe this second

standable usisimilarities w

or the smootorder subject nd it is also a p

is methodologhe average raFigure 11 an

mber of corne

aran factors tand Ashmaw

S OR ANGthe second o

as notice by md order particl

ing the Figurewith a spherthness of the (form is the fproperty of pa

gy, calling it adius of the cnd Equation 2ers). This two

to determinewy, 2001).

GULARITorder shape most of the aule property.

e 1. Particle sre (3D) or a perimeter (2D

first and rounarticles surfac

total degree corners (r) in 22. In the sam last Equation

e the shape a

TYdescriptor. Suthors sited b

shape or formcircle (2D).

D). Surface tndness the secces between c

or roundnesn relation withme study Wadn shows sligh

17

and angularit

(20

(2

Sphericity lefbefore and the

m is the overaRoundness

texture (Barrecond), and it corners.

s to obtain thh the inscribedell (1935) hahtly difference

79

ty

0)

1)

fts ey

all is

et, is

heedases

V

ACi

PIt is idegreof get

Fi

Vol. 18 [20

Table 3:

Aspectircularity

(2D)

d

owers (1953)important to e of subjectivtting errors is

igure 11: W

13], Bund.

General chr

Name

roundness

roundness

roundness

Circularityoutline circula

degree of circulinscribed circ

sphericity

Circularity

Shape facto

) also publishhighlight tha

vity. Folk (19negligible fo

adell’s meth

. A

ronological o for 2

yarity larity cle

y Kru

or S

ed a graphic at any compa955) concludeor sphericity b

hod to estimacircle (H

(22)

overview of 2D sphericity

Author

Pentland

Cox1

Tickell2

Horton2

Wadell Wadell

Riley

umbein and Sloss Janoo

ukumaran

scale to illustaring chart toes that when cbut large for ro

ate the roundHawkins, 19

f the particle y.

Year

1927

1927

1931

1932 1935 1935

1941

1963

1998

2001 Seg

1) R2) H

trate the qualio describe parcharts are useoundness.

dness, corner993).

shape defini

Based o

area

area-perim

area

drainage bCircle diam

Perimet

Circle diam

chart

area-perimgmentation of

anglesRiley, 1941 Hawkins, 1993

itative measurticle propert

ed for classific

rs radius and

18

itions

on

meter

basin meterter

meter

t

meterparticle and s

ure (Figure 12ties has a higcation, the ris

d inscribed

(23)

80

2).ghsk

)

V

Sclassiis imp(1935

TabG

VeAnSubSubRo

We

Kround

Fa censubtenFigur

T(ANG

Vol. 18 [20

F

ome authors fication basedportant to de

5). This classi

ble 4: Degrerade terms

liery angularngular 0bangular 0brounded 0

ounded 0

ell rounded 0

Krumbein anddness paramet

ischer in 193ntral point in nded by the se 14.

To express thGPLA) on the o

13], Bund.

Figure 12: A

as Russel & Td on five andenote that thefication and c

es of roundnRussell & TaylClass

imits (R) Am

N/A0.00-0.15 0.15-0.30 0.30-0.50 0.50-0.70

0.70-1.00

d Sloss (1963)ters using com

3 (Hawkins, the outline

straight or no

he angularityoutlines and co

. A

A Roundness

Taylor in 193d six classes (e way they mclass limits ar

ness: Wadelllor (1937)

Arithmetic midpoint l

N/A0.075 00.225 00.400 00.600 0

0.800 0

) published amparison. See

1993) used aand dividing

on-curved par

y value Fischoncave (ANG

s qualitative

37, Pettijohn iHawkins, 199measure the re showed in t

l Values. (HaPettijohn

Class limits (R)

N/A0.00-0.15 0.15-0.25 0.25-0.40 0.40-0.60

0.60-1.00

a graphical che Figure 13. (C

a straightforwg the outline ts of the prof

her used theGCON) Equatio

scale (Powe

in 1957 and P93) each one roundness isthe Table 4.

awkins, 199n (1957)

Arithmetic midpoint

N/A0.125 0.200 0.315 0.500

0.800

hart easy to dCho, et al. 20

ward method tin angles ar

file were mea

e ratio of anons 24 and 25

ers, 1953)

Powers in 195with its own

s the develop

3), N/A = noPowers

Class limits (R) 0.12-0.17 0.17-0.25 0.25-0.35 0.35-0.49 0.49-0.70

0.70-1.00

determine the 006).

to quantify roround this pasured. This i

ngles standin5 respectively

18

53 developed class limits;

ped by Wade

o-applicables (1953)

Arithmetic midpoint

0.14 0.21 0.30 0.41 0.59

0.84

sphericity an

oundness usinoint that weris illustrated i

ng linear par.

81

d a it

ell

e

nd

ngrein

rts

V

Figu

Fabove

Wparticdiamein miratio oaxis (

Vol. 18 [20

ure 13: Sphethat

Figure 1

igure 14 left e Equations. (

Wentworth in cle to obtain teter of a circlenimum projeof the radius B), see Equat

13], Bund.

ericity and rappears here

4: Fischer’sA=ins

(A) and righ(Hawkins, 199

1922 (Equatithe outline ore fitting the shection (SM). Wof curvature tion 27.

. A

oundness che in the chart

methods ofscribed circl

ht (B), gives 93).

ion 26) used tr contour (Baharpest corneWentworth (Hof the most c

(24)

(26

hart. (Cho et,t is the wade

f angularity ce; B=circum

a similar ang

the maximumarret, 1980). Ter (DS) and theHawkins, 199convex part (R

)

, al., 2006). Tell’s Equatio

computation mscribed circ

gularity of ap

m projection tThe Equatione longest axis

93) also exprRCON) and the

The roundneon number 22

(Hawkins, 1cle

pproximately

to define the n reflects the s (L) plus the ressed the roue longest axis

18

ess Equation2

1993)

0.42 using th

position of threlation of thshortest axis

undness as ths (L) plus sho

(25)

(27)

82

n

he

hehe c heort

V

Apartic

Daxis ba, b an

W1980)the sh

Wthe av

C(EquacorneWentwEquat

Swcornehis Aperim

Langul

Vol. 18 [20

Actually thesecle is in its ma

Dimensions cab. The intentiond c are for 3

FWentworth 19) and it relateharpest corner

Wentworth (19verage radius

Cailleux (Barration 31). Kur (DS) and tworth roundntion 33.

wan in 1974 r(s) (DS1 and

Average roundmeter (PCON) an

Lees (1964a) darity instead

13], Bund.

e last two Equaximum proje

an be seen onon is to make D).

Figure 15: D919 has a seces the diameter (DX).

922b), used dof the pebble

rett, 1980) reuenen in 1956the breath axness with the

shows his EqDS2) and ins

dness of outlnd the actual

developed anof the round

(31

. A

(

uations are thection.

n Figure 15, Ldifference be

Description ocond way to er of the shar

define the roue (RAVG):

elates the rad6 show his roxis (B), Equarelation of sh

quation (Barrcribed circle line (Krumbeperimeter (P)

n opposite dedness, and he

(29)

)

(34)

he same, just

L and B repreetween the 2

of L and B aexpress the

rpest corner (D

undness as the

dius of the moundness indeation 32. Doharpest corner

rett, 1980) rediameter (DI

ein and Petti), Equation 35

finition to roe calls it Deg

t expressed in

esents the maand 3 dimens

axes (Hawkinroundness caDS) and the d

e ratio of the

most convex ex (Barrett, 1

obkins & Folr (DS) and ins

lating the shaI), Equation 3ijohn, 1938) 5.

oundness, it mgree of angul

(32)

n different te

ayor axis a ansions (L and B

ns, 1993) alled Shape idiameter of a

sharpest corn

part and th1980) betweelk (1970) usescribed circle

arpest (or the34. Szadeczsk

relating the

means that helarity. Figure

18

(35)

erms, when th

nd intermediaB are for 2D a

index (Barreta pebble troug

ner (RCON) an

he longest axen the sharpeed a modifie diameter (DI

e two sharpesky-Kardoss ha

concave par

e measures th 16 shows th

(28)

(30)

(33)

83

he

teas

tt, gh

nd

xis est edI),

st)asrts

hehe

V

requirthe di

Inindivi

A(1998

Tthe deEquatbut it

Vol. 18 [20

rements consiistance (x). Se

n order to apidual result w

Figure 16

A roundness i8b) it is descri

The last Equatefinition furthtion that has bis a good exa

13], Bund.

idered when Eee Equation 3

pply the Equwill add to eac

6: Degree of

index appearibed as:

tion appears ahermore sombeen used tryiample of the m

. A

Equation 36 a36.

uation 36 corh other to obt

f angularity m

rs on Janoo (

also as a 2D dme authors ha

ing to define misuse of the

applies as the

rners needs ttain the final

measuremen

(1998), Kuo

descriptor becad used to dedifferent aspequantities an

e angles ( ), i

to be entereddegree of ang

nt technique

and Freeman

cause there is efine the rougects (spherici

nd definitions.

inscribed circ

d in the formgularity.

(Blot and P

n (1998a) an

not a generalghness, this iity, roundness.

18

cle (Rmax-in) an

mula, and eac

(36

ye, 2008)

nd Kuo, et, a

l agreement ois not the onls or roughnes

(

(17)

84

nd

ch

6)

al.

onlys)

V

Sunumbshown

Subecaube use

Asp

Round

Vol. 18 [20

Table

ukumaran anber of sharpnen in Figure 9:

ukumaran anuse it is the cued to describe

pect

dness shape in

shape in

roundne

Averagoutline

roundne

roundne

roundne

roundne

roundne

roundne

roundne

degree

Angula

13], Bund.

e 5: General

nd Ashmawyess corners (E:

nd Ashmawyut off betweene the roughne

Name

ndex

ndex

ess

e roundness of

ess

ess

ess

ess

ess

ess

ess

of angularity

rity factor

. A

chronologic

y (2001) presEquation 37).

y (2001) also n angularity fass.

A

Wentwort

Wentwort

Wentwort

Fischer

Fischer

Szadeczsk

Wadell

Wadell

Russel &

Krumbein

Cailleux

Pettijohn

Powers

Kuenen

Krumbein

Lees

Dobkins &

Swan

SukumaraAshmawy

cal overview

sent an anguAngles i req

suggested usfactor and surf

Author

th

th

th

ky-Kardoss

Taylor

n

n and Sloss

& Folk

an and y

w of the parti

1) B2) H3) K4) P

ularity factor quired to obta

se not bigger face roughnes

Year

19191 diam

1922b shar

1933 conv

19332 nonc

19332 nonc

19333 conv

1935 diam

1935 diam

19372 clas

1941 char

19471 conv

19494 clas

1953 char

19561 axis

1963 char

1964acorncircl

1970 diam

19741 diam

2001 Segmangl

icle roundne

Barret, 1980 Hawkins, 1993 Krumbein and PettiPowers, 1953

(AF) calculain the angula

sampling intss. If so this E

Based on

meter of sharper

rpest corner and

vex parts

curved parts out

curved-straight p

vex parts-perime

meter of corners

meter of corners

s limit table

rt

vex parts

s limit table

rt and class limit

s-convex corner

rtners angles and ile

meter of sharper

meter of sharper

mentation of parles

18

ss

ijohn, 1938

lated from tharity factor ar

terval of N=4Equation coul

n

corner

axis

tline

parts outline

eter

t table

inscribed

corner

corners

rticles and

(37)

85

here

40ld

V

Aroundwith t

WaggresynthtimesThe leThe d

Fig

Hhere i

Odefine

Tbetwe

Vol. 18 [20

A third propedness propertithe authors, at

Wright in 195gate using stetic resin. Th. The unevenength was the

difference betw

ure 17: Mea

However, withis presented so

One techniqueed as the ratio

The convex peen the touchi

Figu

13], Bund.

ROUGHerty called ties, since thent that time, no

5 developed tudies done ohe stones wernness of the sen compared ween these tw

asurement m

h the advanceome research

e used by Janoo between per

perimeter is oing points tha

ure 18: (a) C

. A

HNESS ORexture appean, texture proot measurable

a method to qon 19 mm stre cut in thin urface was trwith an unev

wo lines was d

method for ch(Ja

e of technoloher’s ideas how

oo (1988) to drimeter (P) an

obtained usinat the Feret’s b

Convex perim(modified

R SURFAars early in operty was lone.

quantify the stones. The tesections. The

raced and theven line drawdefined as the

haracterizinganoo, 1998)

ogy it has becw this propert

define the round convex per

ng the Feret’box describes

meter (CPER)after Janoo,

ACE TEXTthe literaturnged describe

surface texturst aggregatese sections proe total length wn as a seriese roughness fa

g the surface

come easier mty should be c

ughness can brimeter (CPER)

’s box (or ds each time it

), (b) Feret m, 1998)

TUREe with the sed but it was

re or roughnes were first eojection was of the trace w

s of chords (sfactor. (Janoo

e texture of a

measure the calculated.

be seen in Fig).

diameter) tendis turn (Figur

measurement

18

sphericity an in accordanc

ess of concretembedded in magnified 12was measuredsee Figure 171998).

an aggregate

roughness an

gure 18a and

ding a line ire 18b).

t

(38)

86

ndce

tea

25d.

7).

nd

is

in

V

Kratio p

EErosiosurfacthe redimencycleserosio

Mratio,invest

TThis iperimconve

Haccuraworks

Vol. 18 [20

Kuo and Freemperimeter (P)

Erosion and don is a morphce, which leaeverse processnsion by addis are not stanon-dilatation (

Mora and CR” (Equati

tigation, they

FigurThe convex aris illustrated

meter but in thex area

TECHN

Hand measureacy special ds placing the

13], Bund.

man (1998a) and average

dilatation imahological procaves the objecs of erosion ang pixels arou

ndardized. A (Equation 40)

ion 41) and tare:

re 19: Evalurea is the areain the Figure

his case the a

NIQUES

ment techniqdevices develsample on th

. A

and Kuo et adiameter (DA

age processincess by whichct less dense and a single dund its boundrepresents the).

Kwanthe “fullness

ation of areaa of the minie 19. The conarea between

S TO DE

HAND Mque was the filoped as the he sliding roa

(39)

(41)

al., (1998b) uAVG), Equation

g techniques h boundary imalong the pe

dilatation cycdary (Pan et ae original are

(2000) uratio, FR” (E

a and convexmum convex

nvex area is othe original

ETERMIN

MEASUREirst used by o“sliding rod

ad calliper as

)

)

use the roughn 39.

are used to mage pixels a

erimeter or oucle increases al., 2006). Thea and A1 is

used the Equation 42)

x area (Morax boundaries cobtained in a outline and t

NE PART

EMENTobvious reasod caliper” use

show Figure

hness (RO) de

obtain the suare removed uter boundarythe particle s

he “n” erosionthe area after

“conin

a and Kuan, circumscribin

a similar way the convex pe

TICLE S

ons, in order ted by Krumbe 20b the leng

18

efinition as th

urface texturfrom an objey. Dilatation shape or imagn and dilatatior “n” cycles o

nvexity the

2000)ng the particl

as the conveerimeter is ou

SHAPE

to improve thbein (1941), gth in differen

(40)

(42)

87

he

re. ct is

geonof

eir

le.exur

heit nt

V

positiactualshape20a shcurvarock fshow

F(W

(Kru

Athe papartic

Fi

Spropeand eaggrevalue

Vol. 18 [20

ons can be oblly used by op

e analysis (Wehows (the tw

ature; or the “fragment, so, equipment.

Figure 20: (aWenworth, 19

umbein, 194

Another helpfuarticle’s contocle’s diameter

igure 21: Ci

ieving, e.g. aerties. By comelongation indgate particlessuch as 5:1.

13], Bund.

btain by usinpticians to mentworth, 192

wo adjacent pi“Szadeczky-K

the outline tr

a) convexity922b), (b) sli41) and c) Sz

ul tool to deteour over a cirr.

ircle scale us

according to mbining meshdex, ASTM Ds that have aThe index re

. A

ng the scale prmeasure the cu22b) works mivots are inva

Kardoss’s appraced is then

y gage, used iding rod calzadeczky-Ka

the paermine the parcle scale app

sed by Wader

SIEVEEN 933-3:1

h geometries tD4791 (Flat

a ratio of lengepresents the p

rovided in thurvature of len

measuring the ariable) as maparatus” devel

analyzed (Kr

to determineliper, deviceardoss (1933article outlinarticle dimensearing in Figu

ell (1935) to oundness

E ANALYS1997, can bethe obtained rand elongate

gth to thicknpercentage on

he handle; thenses but easimovement ofany the centrlop in 1933 trumbein and P

e the curvatue to measure 3) apparatus,ne. sions was the

gure 21, thus i

determine p

SISe used to detresults can beed particles a

ness equal to n weight of th

e “convexity gly applicable f the central pral pivot movthat traces thePettijohn, 193

ure in particlthe particle it was utiliz

e “camera lucit is possible

particle’s dia

termine simpe used to quaare defined aor greater th

hese particles

18

gage” that wa to the particpivot as Figurves more is the profile of th38) Figure 20

le corners axis length

zed to obtain

cida” to projeto measure th

ameter and

ple large scaantify flakinesas those coarshan a specifies). The metho

88

aslerehehe0c

n

ct he

lessseedod

V

is notthe sieEN 9from fractiothe pafoundstandausingcompKwanthe teis theadvan

Crequirfor pdifficu

Acharacroundchart

Atwo mthat edefini

Fdetermthat thhe conhis stu

Vol. 18 [20

t suitable for eve, and the g33-3:1997 re4 mm and uon and the searticles is ob

d, finally witards related wsieve analys

uters age andn, 2000; Perssesting time coe error due thntages are mo

Charts develored when meebbles whichult to measur

Fig

A qualitative ccteristics, it w

ded, rounded was prepared

A new chart inmean propertieliminated thitions. (Krum

olk (1955) wmination of sphe sphericity ncluded, it waudy.

13], Bund.

fine materialgreat amount lated to flaki

up to 63 mm. econd use a bbtain and witth this two pwith the partisis to determid image analyson, 1998). Inompare with the overlappinre compare w

ped over theasuring each h were mease due to the s

gure 21: Kru

chart by Powewas divided oand well rou

d with photogr

ncluding spheies of particlhe subjectivit

mbein and Slos

worried abouphericity anddeterminatioas necessary t

. A

ls. This due tof particles i

iness index.two sieving

bar sieve, afteth the secondparameters thicle shape butine particle’sysis sieving rndustry is alsothe traditionang or hiding

with disadvant

CHART Ce necessity oparticle. Kru

sured by Waecond order s

umbein (194

ers (1953) tryon six roundnunded) and twraphs to enha

ericity and roe’s shape, futy of qualitass, 1963). (Se

ut the persoangularity (h

n by chart coto carry out a

to the difficuln relation to tThe test is poperations a

er the first sied sieving (bahe flakiness t, this above geometrical

research is tako applying thl sieving metof the partic

tages (Anders

COMPARof faster resuumbein (1941adell’s methoscale that roun

41) comparis

y to include bness ranges (vwo sphericityance the reade

oundness appeurthermore, thative descriptee Figure 13).

on’s error onhe used the Poomparison hasa more wide r

lty to get the the area of th

performed on are necessary,eving the avear sieving) th

index is depresented areproperties. S

king place (Ae image analythod. An incocles during thson, 2010).

RISONults because 1) present a cod because thndness repres

son chart for

both (sphericitvery angular,y series (higher perspective

ear, this timehere was incltion. The ch

n the chart’owers 1953 cs a negligibleresearch due t

fine grains phe sieve (Persn aggregates w, the first sep

erage maximuhe shortest axetermined. The probably thSieve analysiAndersson, 20ysis sieving wonvenient of ithe capture p

the long timcomparison rohis property sents. (See Fi

r roundness

ty and roundn, angular, subh and low spe. (See Figure

e it was easierluded the numhart is based

s comparisocomparison che error while the high varia

18

passed througson, 1998) e.with grain sizparates on sizum diameter oxis diameter here are morhe most knows is facing th010; Mora anwith decrees oimage analys

process but th

me consuminoundness chawas the mogure 22).

ness) particleb-angular, subphericity). The 12)

r to handle thmerical value

d on Wadell

on studied thhart), he founthe roundnes

ability show b

89

ghg. zezeofisre

wnhendonsis he

ngart ost

’sb-

his

hees’s

hendss, by

V

Imautom

-

-

-

Fth

Vol. 18 [20

mage analysismated. Differe

Feret Diamcan rotate 22 (left) tto determin

Fourier Teindividual textural feentrant ang(right).

Fractal Dimand Vallejof the fract

Figure 22: (lehe particle to

te

13], Bund.

s is a practicaent techniques

meter: the Feraround one phese method ne diameters

echnique: It particles (Eq

eatures for grgles in order

mension: Irreo, 1997), Figtal line can be

eft) Feret meadefine the shoechnique with

. A

IMAGEal method to us appear to pr

ret diameter iparticle, or ouis not a fine (Janoo, 1988

produces maquation 43). anular soils. to complete

egular line at gure 23 showse defined as E

asurement techortest and lonh two radiuses

E ANALYSuse for shaperocess these im

is the longituutline, to defidescriptor, bu)

athematical reThis methodThe problemthe revolutio

any level ofs fractal analyEquation 44.

hnique is defingest Feret dias at one angle

YSISe classificatiomages, among

ude between tfine dimensionut as it was s

elations that d favours the m in the methon (Bowman

f scrutiny is bysis by the d

ined by two pameter (Janooe (Bowman et

n since it is fg them are:

two parallel lns, as it is shay above it is

characterize analysis of

hodology remet al., 2001),

by definition dividing meth

parallel lines to, 1988), (right al., 2001)

19

fast and can b

lines, this linehown in Figurs a helpful too

the profile oroughness an

mains in the re see Figure 2

fractal (Hysliod. The lengt

turning arounht) Fourier

(43)

(44)

90

be

esreol

ofnde-22

ipth

nd

V

-

-

-

Lcan se

Adigitaanalyparticresolu

Wbetter

Vol. 18 [20

Figure 2

Orthogonabetween thtechniques

Laser Scantechniquesparticles hthe lower Tolppanen

Laser-Aidesurveying as to use lcertain per

Figure 2Last two 3D teee in Figure 2

All these preval way obtainisis regarding

cles; orientatioution have an

When resolutir with the real

13], Bund.

23: Fractal an

al image analyhem to acquirs can be used

nning Technis. In Figures have control p

part, in Fign, 2002).

ed Tomograp(see Figure 2iquid with sarcent of light

4: a) Scanniechniques obt25 (right).

vious techniqing the desireon the errors

on is not releinfluence on

ion is increasl form, in the

. A

nalysis by th(Hyslip a

ysis: This tecre the three pain this orthog

ique: this kin24a) is show

points in ordegure 24b) it

phy (LAT), in25, left)). Thiame refractivego through. (M

ing head, b) tain the partic

ques are easiled measuremes involve, amvant when it the accuracy

se more accuother hand, m

he dividing mnd Vallejo,

chnique is basarticle dimensgonal way.

nd of laser scwed the laserer to keep a r

can be see

n this case a lis technique ie index as theMatsushima e

scanning patcle shape that

ly written in ent, but there

mong them areis random an. (Zeidan et a

uracy is obtamore resolutio

method at dif1997)

sically the usesions (Fernlu

canning 3D isr head scann

reference pointhe laser pa

laser sheet is is different ane particles, paet al., 2003).

th (Lanaro at is later used

codes or scrare some inte

e image resolnd large numbal., 2007)

ain and the oon means mor

fferent scrut

e of two imagund, 2005), an

s one of the mning the rocknt when moveath followed

used to obtaind has speciaarticles must

and Tolppaned to achieve m

ripts to be ineresting pointlution and oriber of particl

object represere spending o

19

tiny scale

ges orthogonny of the abov

most advancek particles, the them to sca. (Lanaro an

in the particleal requiremenlet the laser o

en, 2002) measures as w

nterpreted in ts in the imagientation of thes are involv

entation matcon memory an

91

alve

edheannd

esntsor

we

agehee;

chnd

V

time;(Schä

Srelativbut nsimilaoutpuperim

IndefiniKrumusedArasadifferthreesingledescriliteratis notroundProbato rec

Manalyagreemother

T(3-dim

Vol. 18 [20

thus, resolutäfer, 2002).

Figure 25: c

chäfer (2002ve high errorot for perimear results wheut for those tmeter should b

Tn order to desitions (qualit

mbein, 1941; Sto simplify t

an et al., 201rent scales. Thsub-quantitie

e definition iptors are expture that manyt a clear meadness, does it ably they coulcognize in mo

Many image asis, fractal dment on the applications.

There are sevemensions, 3-d

13], Bund.

tion needs to

(left) LAT scompleted an

) conclude ths. It can be veter that keepen 3 differentterms/quantit

be treated with

TERMS, Qscribe the partative and qSneed & Folkthe complexit10) are usinghe terms are es are probabcan interpret

plained, and ty of the shap

aning on whameans that th

ld be on theorost of the case

analysis techndimension, tousage or con

eral shape desdimension ort

. A

o be accordin

scaning partind mesh gen

hat attributes vanish or at lp the error ast resolutions ties that invoh care.

DIS

QUANTITrticle shape i

quantitative)k, 1958). Allty of shape d

g three sub-qumorphology/

bly the best wt the whole these three sce descriptors t this descriphe angularityry but not in s.

IMAGEniques had beomography, nclusion to en

criptors and athogonal and

ng with the g

icles (Matsunerated. (Mat

like length east diminishs big as initiwere used in

olve the peri

SCUSSI

TIES ANDin detail, therused in the

l mathematicadescription. Suantities; one/form, roundnway to classif

morphologyales represenare presented

ptor defines, ey never ends?

reality. Physi

E ANALYeen used to detc., (Hyslipnsure the bes

also various td 2-dimension

goal and prec

ushima et al.,tsushima et a

when measurh using high rially. Johanss

n the same paimeter. Thus

ON

D DEFINIre are a numb literature (al definitionsSome authore and each dness and surfafy and describy. Common nt an option. Id with the same.g. when the Could they bical meaning

SISdescribe the

p and Vallejost particle des

techniques to ns). Each tech

cision needed

, 2003), (righal., 2003)

ring digital iresolution jusson and Vall article obtainin

all quantitie

ITIONSber of terms, (e.g. Wadell,s (quantitatives (Mitchell &

describing theace texture (Fbe a particle language is

It is evident inme name but ere is no uppbe more and of the quanti

particle shapo, 1997) butscriptor for g

capture the phnique presen

19

d in any work

ht) 3D scan

images presenst for diamete(2011) obtai

ng an unstabes relating th

quantities an, 1932, 1934es) are mode& Soga, 2005e shape but Figure. 1). Th

because not needed whe

n the reviewealso that ther

per limit in thmore angular

ities is difficu

pe, e.g. Fouriet there is no

geotechnical o

particles profints advantage

92

k.

nterinlehe

nd4;

els5;at hea

enedreher?

ult

erotor

lees

Vol. 18 [2013], Bund. A 193

and disadvantages. 3-dimensions is probably the technique that provide more information about the particle shape but the precision also lies in the resolution; the equipment required to perform such capture could be more or less sophisticated (scanning particles laying down in one position and later move to complete the scanning or just falling down particles to scan it in one step). 3-dimensions orthogonal, this technique use less sophisticated equipment (compare with the previous technique) but its use is limited to particles over 1cm, also, information between the orthogonal pictures is not capture. 2-dimensions require non sophisticated equipment but at the same time the shape information diminish compare with the previous due the fact that it is possible to determine only the outline; as the particle measurements are performed in 2-dimensions it is presumed that they will lie with its shortest axis perpendicular to the laying surface when they are flat, but when the particle tends to have more or less similar axis the laying could be random.

Advantages on the use of image analysis are clear; there is not subjectivity because it is possible to obtain same result over the same images. Electronic files do not loose resolution and it is important when collaboration among distant work places is done, files can be send with the entire confidence and knowing that file properties has not been changed. Technology evolutions allowed to work with more information and it also applies to the image processing area were the time consumed has been shortened (more images processed in less time).

One important aspect in image analysis is the used resolution in the analysis due the fact that there are measurements dependent and independent on resolution. Thus, those dependent measurements should be avoided due the error included when they are applied, or avoid low resolution to increase the reliability. Among these parameters length is the principal parameter that is influences by resolution (e.g. perimeter, diameter, axis, etc.). Resolution also has another aspect with two faces, quality versus capacity, more resolution (quality) means more storage space, a minimum resolution to obtain reasonable and reliable data must be known but it depend on each particular application.

APPLICATIONSQuantify changes in particles, in the author’s thought, is one of the future applications due the

non-invasive methods of taking photographs in the surface of the dam’s slope, rail road ballast or roads. Sampling of the material and comparing with previous results could show volume (3D analysis) or area (2D analysis) loss of the particles as well as the form, roundness and roughness. This is important when it has been suggested that a soil or rock embankment decrees their stability properties (e.g. internal friction angle) with the loss of sphericity, roundness or roughness.

Seepage, stock piling, groundwater, etc., should try to include the particle shape while modelling; seepage requires grading material to not allow particles move due the water pressure but in angular materials, as it is known, the void ratio is great than the rounded soil, it means the space and the possibilities for the small particles to move are greater; stock piling could be modelled incorporating the particle shape to determine the bin’s capacity when particle shape changes (void ratio changes when particle shape changes) Modelling requires all information available and the understanding of the principles that apply.

Industry is actually using the particle shape to understand the soil behaviour and transform processes into practical and economic, image analysis has been included in the quality control to determine particle shape and size because the advantages it brings, e.g. the acquisition of the sieving curve for pellets using digital images taken from conveyor, this allows to have the

Vol. 18 [2013], Bund. A 194

information in a short period of time with a similar result, at least enough from the practical point of view, as the traditional sieving.

CONCLUSIONS• A common language needs to be built up to standardize the meaning on geotechnical field

that involve the particle shape.

• Based on this review it is not clear which one is the best descriptor.

• Image analysis tool is objective, make the results repeatable, obtain fast results and workwith more amount of information.

• Resolution needs to be taken in consideration when image analysis is been carried outbecause the effects could be considerable. Resolution must be set according to the necessities. Parameters as perimeter can be affected by resolution.

• There are examples where particle shape has been incorporated in industries related togeotechnical engineering, e.g. in the ballast and asphalt industry for quality control.

FURTHER WORK Three main issues have been identified in this review that will be further investigated; the

limits of shape descriptors (quantities) influence of grading and choice of descriptor for relation to geotechnical properties.

Shape descriptors have low and high limits, frequently the limits are not the same and the ability to describe the particle’s shape is relative. The sensitivity of each descriptor should be compare to apply the most suitable descriptor in each situation.

Sieving curve determine the particle size in a granular soil, particle shape could differ in each sieve size. There is the necessity to describe the particle shape on each sieve portion (due to practical issues) and included in the sieve curve. Obtain an average shape in determined sieve size is complicated (due to the possible presence of several shapes) and to obtain the particle shape on the overall particle’s size is challenging, how the particle shape should be included?

Since several descriptors have been used to determine the shape of the particles but how is the shape related with the soil properties? It is convenient to determine the descriptor’s correlation with the soil properties.

REFERENCES1. Andersson T. (2010) “Estimating particle size distributions based on machine vision”.

Doctoral Thesis. Department of Computer Science and Electrical Engineering. LuleåUniversity of Technology. ISSN: 1402-1544. ISBN 978-91-7439-186-2

2. Arasan, Seracettin; Hasiloglu, A. Samet; Akbulut, Suat (2010) “Shape particle of naturaland crished aggregate using image analysis”. International Journal of Civil and StructuralEngineering. Vol. 1, No. 2, pp. 221-233. ISSN 0970-4399

Vol. 18 [2013], Bund. A 195

3. Aschenbrenner, B.C. (1956) “A new method of expressing particle sphericity”. Journal ofSedimentary Petrology. Vol., 26, No., 1, pp. 15-31.

4. Barton, Nick & Kjaernsli, Bjorn (1981) “Shear strength of rockfill”. Journal of theGeotechnical Engineering Division, Proceedings of the American Society of CivilEngineers (ASCE) Vol. 107, No. GT7.

5. Barrett, P. J. (1980) “The shape of rock particles, a critical review”. Sedimentology. Vol.27, pp. 291-303.

6. Blott, S. J. and Pye, K., (2008) “Particle shape: a review and new methods ofcharacterization and classification”. Sedimentology. Vol. 55, pp. 31-63

7. Bowman, E. T.; Soga, K. and Drummond, W. (2001) “Particle shape characterizationusing Fourier descriptor analysis”. Geotechnique. Vol. 51, No. 6, pp. 545-554

8. Cho G., Dodds, J. and Santamarina, J. C., (2006) “Particle shape effects on packingdensity, stiffness and strength: Natural and crushed sands”. Journal of Geotechnical andGeoenvironmental Engineering. May 2006, pp. 591-602.

9. Dobkins, J. E. and Folk, R. L. (1970) “Shape development on Tahiti-nui”. Journal ofSedimentary Petrology. Vol. 40, No. 2, pp. 1167-1203.

10. Folk, R. L. (1955) “Student operator error in determining of roundness, sphericity andgrain size”. Journal of Sedimentary Petrology. Vol. 25, pp. 297-301.

11. Fernlund, J. M. R. (1998) “The effect of particle form on sieve analysis: A test by imageanalysis”. Engineering Geology. Vol. 50, No. 1-2, pp. 111-124.

12. Fernlund, J. M. R. (2005)” Image analysis method for determining 3-D shape of coarseaggregate”. Cement and Concrete Research. Vol. 35, Issue 8, pp. 1629-1637.

13. Fernlund, J. M. R.; Zimmerman, Robert and Kragic, Danica (2007) “Influence ofvolume/mass on grain-size curves and conversion of image-analysis size to sieve size”.Engineering Geology. Vol. 90, No. 3-4, pp. 124-137.

14. Hawkins, A. E. (1993) “The Shape of Powder-Particle Outlines”. Wiley, New York.

15. Hyslip, James P.; Vallejo, Luis E. (1997) “Fractal analysis of the roughness and sizedistribution of granular materials”. Engineering Geology. Vol. 48, pp. 231-244.

16. Janoo, Vincent C. (1998) “Quantification of shape, angularity, and surface texture of basecourse materials”. US Army Corps of Engineers. Cold Region Research and EngineeringLaboratory. Special report 98-1.

17. Johansson, Jens and Vall, Jakob (2011) “Jordmaterials kornform”. Inverkan påGeotekniska Egenskaper, Beskrivande storheter, bestämningsmetoder. Examensarbete.Avdelningen för Geoteknologi, Institutionen för Samhällsbyggnad och naturresurser.Luleå Tekniska Universitet, Luleå. (In Swedish)

18. Krumbein, W. C. and Pettijohn, F.J. (1938) “Manual of sedimentary petrography”.Appleton-Century Crofts, Inc., New York.

19. Krumbein, W. C. (1941) “Measurement and geological significance of shape androundness of sedimentary particles”. Journal of Sedimentary Petrology. Vol. 11, No. 2,pp. 64-72.

Vol. 18 [2013], Bund. A 196

20. Krumbein, W. C. and Sloss, L. L. (1963) “Stratigraphy and Sedimentation”, 2nd ed.,W.H. Freeman, San Francisco.

21. Kuo, Chun-Yi and Freeman, Reed B. (1998a) “Image analysis evaluation of aggregatesfor asphalt concrete mixtures”. Transportation Research Record. Vol. 1615, pp. 65-71.

22. Kuo, Chun-Yi; Rollings, Raymond and Lynch, Larry N. (1998b) “Morphological studyof coarse aggregates using image analysis”. Journal of Materials in Civil Engineering.Vol. 10, No. 3, pp. 135-142.

23. Lanaro, F.; Tolppanen, P. (2002) “3D characterization of coarse aggregates”. EngineeringGeology. Vol. 65, pp. 17-30.

24. Lees, G. (1964a) “A new method for determining the angularity of particles”.Sedimentology. Vol., 3, pp. 2-21

25. Lees, G. (1964b) “The measurement of particle shape and its influence in engineeringmaterials”. British Granite Whinstone Federation. Vol., 4, No. 2, pp. 17-38

26. Matsushima, Takashi; Saomoto, Hidetaka; Matsumoto, Masaaki; Toda, Kengo; Yamada,Yasuo (2003) “Discrete element simulation of an assembly of irregular-shaped grains:Quantitative comparison with experiments”. 16th ASCE Engineering MechanicsConference. University of Washington, Seattle. July 16-18.

27. Mitchell, James K. and Soga, Kenichi (2005) “Fundamentals of soil behavior”. Thirdedition. WILEY.

28. Mora, C. F.; Kwan, A. K. H.; Chan H. C. (1998) “Particle size distribution analysis ofcoarse aggregate using digital image processing”. Cement and Concrete Research. Vol.28, pp. 921-932.

29. Mora, C. F. and Kwan, A. K. H. (2000) “Sphericity, shape factor, and convexitymeasurement of coarse aggregate for concrete using digital image processing”. Cementand Concrete Research. Vol. 30, No. 3, pp. 351-358.

30. Pan, Tongyan; Tutumluer, Erol; Carpenter, Samuel H. (2006) “Effect of coarse aggregatemorphology on permanent deformation behavior of hot mix asphalt”. Journal ofTransportation Engineering. Vol. 132, No. 7, pp. 580-589.

31. Pellegrino, A. (1965) “Geotechnical properties of coarse-grained soils”. Proceedings.International Conference of Soil Mechanics and Foundation Engineering. Vol. 1, pp. 97-91.

32. Pentland, A. (1927) “A method of measuring the angularity of sands”. MAG. MN. A.L.Acta Eng. Dom. Transaction of the Royal Society of Canada. Vol. 21. Ser.3:xciii.

33. Persson, Anna-Lena (1998) “Image analysis of shape and size of fine aggregates”.Engineering Geology. Vol. 50, pp. 177-186.

34. Powers, M. C. (1953) “A new roundness scale for sedimentary particles”. Journal ofSedimentary Petrology. Vol. 23, No. 2, pp. 117-119.

35. Pye, W. and Pye, M. (1943) “Sphericity determination of pebbles and grains”. Journal ofSedimentary Petrology. Vol. 13, No. 1, pp. 28-34.

36. Quiroga, Pedro Nel and Fowle, David W. (2003) “The effects of aggregate characteristicson the performance of portland cement concrete”. Report ICAR 104-1F. Project number104. International Center for Aggregates Research. University of Texas.

Vol. 18 [2013], Bund. A 197

37. Riley, N. A. (1941) “Projection sphericity”. Journal of Sedimentary Petrology. Vol. 11,No. 2, pp. 94-97.

38. Rousé, P. C.; Fennin, R. J. and Shuttle, D. A. (2008) “Influence of roundness on the voidratio and strength of uniform sand”. Geotechnique. Vol. 58, No. 3, 227-231

39. Santamarina, J. C. and Cho, G. C. (2004) “Soil behaviour: The role of particle shape”.Proceedings. Skempton Conf. London.

40. Schäfer, Michael (2002) “Digital optics: Some remarks on the accuracy of particle imageanalysis”. Particle & Particle Systems Characterization. Vol. 19, No. 3, pp. 158-168.

41. Shinohara, Kunio; Oida, Mikihiro; Golman, Boris (2000) “Effect of particle shape onangle of internal friction by triaxial compression test”. Powder Technology. Vol. 107,pp.131-136.

42. Skredkommisionen (1995) ”Ingenjörsvetenskapsakademinen”, rapport 3:95, Linköping1995.

43. Sneed, E. D. and Folk, R. L. (1958) “Pebbles in the Colorado river, Texas: A study inparticle morphogenesis”. Journal of Geology. Vol. 66, pp. 114-150.

44. Sukumaran, B. and Ashmawy, A. K. (2001) “Quantitative characterisation of thegeometry of discrete particles”. Geotechnique. Vol. 51, No. 7, pp. 619-627.

45. Szádeczy-Kardoss, E. Von (1933) “Die bistimmung der abrollungsgrades”. Geologie undpaläontologie. Vol. 34B, pp. 389-401. (in German)

46. Teller, J. T. (1976) ”Equantcy versus sphericity”. Sedimentology. Vol. 23. pp. 427-428.

47. Tickell, F. G. (1938)” Effect of the angularity of grain on porosity and permeability”.bulletin of the American Association of Petroleum Geologist. Vol. 22, pp. 1272-1274.

48. Tutumluer, E.; Huang, H.; Hashash, Y.; Ghaboussi, J. (2006) “Aggregate shape effects onballast tamping and railroad track lateral stability”. AREMA 2006 Annual Conference,Louisville, KY.

49. Wadell, H. (1932) “Volume, Shape, and roundness of rock particles”. Journal ofGeology. Vol. 40, pp. 443-451.

50. Wadell, H. (1933) “Sphericity and roundness of rock Particles”. Journal of Geology. Vol.41, No. 3, pp. 310–331.

51. Wadell, H. (1934) “Shape determination of large sedimental rock fragments”. The Pan-American Geologist. Vol. 61, pp. 187-220.

52. Wadell, H. (1935) “Volume, shape, and roundness of quartz particles”. Journal ofGeology. Vol. 43, pp. 250-279.

53. Wentworth, W. C. (1922a) “The shape of beach pebbles”. Washington, U.S. GeologicalSurvey Bulletin. Vol. 131C, pp. 75-83.

54. Wentworth, W. C. (1922b) “A method of measuring and plotting the shape of pebbles”.Washington, U.S. Geological Survey Bulletin. Vol. 730C, pp. 91-114.

55. Wentworth, W. C. (1933) “The shape of rock particle: A discussion”. Journal of Geology.Vol. 41, pp. 306-309.

Vol. 18 [2013], Bund. A 198

56. Witt, K. J.; Brauns, J. (1983) “Permeability-Anisotropy due to particle shape”. Journal ofGeotechnical Engineering. Vol. 109, No. 9, pp. 1181-1187.

57. Zeidan, Michael; Jia, X. and Williams, R. A. (2007) “Errors implicit in digital particlecharacterization”. Chemical Engineering Science. Vol. 62, pp. 1905-1914.

© 2013 EJGE

Paper II

Rodriguez, J. M.; Johansson, J. M. A. and Edeskär, T. (2012) Particle Shape Determination by Two-Dimensional Image Analysis in Geotechnical Engineering. Proceedings of the 16th Nordic Geotechnical Meeting. Copenhagen, 9-12 May 2012. pp. 207-218. Danish Geotechnical Society. Dgf-Bulletin 27. May 2012.

Site investigation and laboratory testing – Particle Shape Determination by Two-Dimensional Image Analysis in Geotechnical Engineering

DGF – Bulletin 27 207 NGM 2012 Proceedings

Particle Shape Determination by Two-Dimensional Image Analysis in Geotechnical Engineering Rodriguez, J.M. Luleå University of Technology, Sweden, [email protected]

Johansson, J.M.A. Luleå University of Technology, Sweden

Edeskär, T. Luleå University of Technology, Sweden

ABSTRACT Particle shape of soil aggregates is known to influence several engineering properties; such as the internal friction angle, the permeability etc. Previously shape classification of aggregates has mainly been performed by ocular inspection and e.g. by sequential sieving. In geotechnical analysis has been a lack of an objective and rational methodology to classify shape properties by quantitative measures. Recent development in image analysis processing has opened up for classification of particles by shape. In this study 2D-image analysis has been adapted to classify particle shape for coarse grained materials. This study covers a review of soil classification methods for particle shape and geometrical shape descriptors. The image analysis methodology is tested and it is investigated how the results are affected by resolution, magnification level and type of shape describing quantity. Evaluation is carried out on as well idealized geometries as on soil samples. The interpreted results show that image analysis is a promising methodology for particle shape classification. But since the results are affected by the image acquisition procedure, the image processing, and the choice of quantity, there is a need to establish a methodology to ensure the objectivity in the particle shape classification.

Keywords: Image Analysis, Laboratory test, Soil classification, Granular materials, Geomorphology.

1 INTRODUCTION

1.1 Background Particle shape is known to influence technical properties of soil material and unbound aggregates (Santamarina and Cho, 2004; Mora and Kwan, 2000). Among documented properties affected by the particle shape are e.g. void ratio (porosity), internal frictionangle, and hydraulic conductivity(permeability) (Rouse et. al., 2008; Shinoharaet. al., 2000; Witt and Brauns, 1983). Ingeotechnical guidelines particle shape isincorporated in e.g. soil classification(Eurocode 7) and in national guidelines e.g.for evaluation of friction angle(Skredkommisionen, 1995). Thisclassification is based on ocular inspection

and quantitative judgement made by the individual practicing engineer. There is today no general accepted system to apply ocular classification but there are several systems suggested (Powers, 1953; Krumbein, 1941). These systems are not coherent in definitions. The lack of possibility to objectively describe the shape hinders the development of incorporating the effect of particle shape in geotechnical analysis.

In the ballast industry there are established standardised classification systems incorporating particle shapes (e.g. EN 933-4, 2008 and ASTM D 4791, 2005). These systems have been developed basically for quality control and for industry requirements; e.g. railway ballast and concretemanufacturing and are focusing on simplegeometries. Besides these examples there are


NGM 2012 Proceedings 208 DGF - Bulletin

a number of potential areas of application of shape classification of soil materials.

Recent progress has resulted in that image acquisition and image analysis has been proven to be useful tools for two-dimensional analysis of simple geometries (Persson, 1998). The geometry of soil particles is however complex and there is a need of development to validate appropriate geometrical definitions and algorithms as descriptors for useful particle shape classification in practise. In Johansson & Vall (2011) a pre-study was performed in order to identify and compile information concerning particle shape of coarse grained soils; the impact on geotechnical properties, existing quantities and definitions, and determination by usage of image analysis. This paper incorporates the results from the mentioned pre-study.

1.2 Scope of study The scope of this study is to explore the possibilities and limitations concerning applying image analysis for soil particle shape classification.

The goals of the study are: 1) To describe a methodology for soil particleshape determination by image acquisition andanalysis.2) To enlighten results from comparisons ofdifferent geometrical definitions, includingusability and sensitivity.

The study consists partly of a literature review of as well particle shape determination as of existing definitions. Moreover, the image analysis methodology is tested and discussed.

2 PARTICLE SHAPE

2.1 Terms and quantities In this study the word shape is used to describe a grain’s overall geometry. Furthermore, in order to describe the particle shape in more detail, there are a number of terms, quantities and definitions used in the literature. Some authors (Mitchell & Soga, 2005; Arasan et al., 2010) are using three sub-quantities; one and each describing the

shape but at different scales. The terms are morphology/form, roundness and surface texture. In fig. 2-1 is shown how the scale terms are defined.

Figure 2-1 Shape describing sub quantities (Mitchell & Soga, 2005)

At large scale a particle’s diameters in different directions are considered. At this scale, describing terms as spherical, platy, elongated etc., are used. An often seen quantity for shape description at large scale is sphericity (antonym: elongation). Graphically the considered type of shape is marked with the dashed line in Figure 2-1.

At intermediate scale is focused on description of the presence of irregularities. Depending on at what scale an analysis is done; corners and edges of different sizes are identified. By doing analysis inside circles defined along the particle’s boundary, deviations are found and valuated. The mentioned circles are shown in Figure 2-1. A generally accepted quantity for this scale is roundness (antonym: angularity).

Regarding the smallest scale, terms like rough or smooth are used. The descriptor is considering the same kind of analysis as the one described above, but is applied within smaller circles, i.e. at a smaller scale. Surface texture is often used to name the actual quantity.

2.2 Geometrical definitions For description of the scale dependent quantities, there are found a large number of terms and definitions. As what is stated in Johansson & Vall (2011) expressions and terms are used arbitrary.

There are a number of different definitions within the large and intermediate scale



groups. There are also some definitions not fitting in to only one of these, but are influenced by as well the general form as by the angularity.

Regarding mathematical definitions for description of the surface texture, the authors views differ; some say that surface texture is to be determined by analogy with the intermediate scale shape, but with the scale decreased (e.g. Mitchell & Soga, 2005). In

Santamarina & Cho (2004) it is meant that the lack of a characteristic scale of which surface texture is to be analyzed makes it difficult to do direct measurements. These authors are suggesting an approach to study interparticle contact area to describe roughness.

In Table 2-1 some definitions of shape describing quantities are presented.

Table 2-1 Some shape describing quantities are listed. As well definitions and figures as references are included. The quantities are used by the authors listed as references.

EQ. QUANTITY DEF. FIGURES REF.

1 Wadell, 1935

2 Degree of circularity

Wadell, 1935

3 Roundness Tickell, 1938

4 Angularity Pentland, 1927

5 Roundness/ Circularity

Riley, 1941/ ImageJ

6 Inscribed circle sphericity

Riley, 1941

c

a

DD

p

a

PP

c

p

AA

C2

p

AA

2p

p

PA4

C

I

DD



7 Circularity Blott & Pye 2008

8 Roughness Janoo, 1998

9 Roughness Kuo, et. al. 1998

10 Roundness

Wadell, 1932; Krumbein & Sloss, 1963; Mitchell & Soga, 2005

11 Sphericity

Krumbein, 1941, Stückrath et al., 2006

12 Aspect Ratio

ImageJ and Image Analysis Pro

Ap Area of the particle outline

Da Diameter of a circle with an area equal to that of the particle outline

Dc Diameter of smallest circumscribed circle Pp Perimeter of particle outline

Pa Perimeter of a circle of the same area as particle outline

Ac Area of the smallest circumscribing circle Ac2 Area of a circle with a diameter equal to the

longest distance between two points on the particle outline

Dinsc. Diameter of the largest inscribed circle Pconv. Perimeter, convex Davg Diameter, average

This definition is almost the same as no. 3.There will be a difference if the particle is very bent, e.g. L-shaped. The average diameter may be calculated byusage of software.The dimensions a, b and c (length, width and

perpendicular to AR defined as in the some image analysissoftware.Used software.

p

2p

AP

AVGDP

*

conv.

p

PP

insc.

i

r/r N

32ac b

MinorMajor

5



2.3 Methods for particle shape determination

There are several methods used to determine the particle shape. Techniques have been developed from handmade measuring by direct scaling, convexity gauge’s, or tools developed for the specific task (Szadeczky-Kardoss 1933). Here, the use of classification chart and sieve analysis is further reviewed.

Classification chart By usage of comparison charts, measuring may be avoided. Some comparison charts are those used by Powers (1953), Krumbein (1941) or Krumbein and Sloss (1963). The latter one is represented in Figure 2-2.

Figure 2-2 Example of a comparison chart (Santamarina and Cho, 2004)

In the chart above, roundness is defined as in eq. 10 in Table 2-1. The definition of sphericity is vaguer. According to Krumbein & Sloss (1963), the sphericity is “related to the proportion between length and breadth of the image”. In Santamarina & Cho (2004) is said that “sphericity is quantified as the diameter ratio between the largest inscribed and the smallest circumscribing sphere”. Eq. 6 in Table 2-1 is a two-dimensional version of the latter definition. In Cho et al., (2006) the classification of sphericity was done by comparison of images of the analysed soil particles, and images in the chart in Figure 2-2. This subjective procedure makes itirrelevant how the quantities aremathematically defined. Folk (1955)concludes that when charts are used for

classification, the risk of getting errors is negligible for sphericity but large for roundness.

Sieve analysis Bar sieving, e.g. according to EN 933-3:1997, can be used to determine simple large scale properties. By combining mesh geometries the obtained results can be used to quantify flakiness and elongation index. The method is not suitable for fine materials. This due to the difficulty to get the fine grains passed through the sieve, and the great amount of particles in relation to the area of the sieve (Persson, 1998).

Image analysis The development of image acquisition techniques and image processing facilitates a systematic approach to use mathematical descriptors for classification of particle shape (Santamarina & Cho 2004). By using algorithms subjectivity related to e.g. ocular classification by charts, is avoided (Persson, 1998).

2.4 Standards and guidelines As already mentioned, there are present standards and guidelines related to particle shape classification, especially within in the ballast industry focusing on paving- concrete and railway applications. These standards are valid for coarse materials. The ASTM D 3398 (ASTM 2006) are regarding shape and texture characteristics that may affect the asphalt concrete mixtures performance.

Standards based on sieve analysis, e.g. ASTM D 4791 (ASTM 2005) and EN 933-3:1997 (CEN 1997), are both regarding width/length ratio; e.g. by flakiness index. EN 933-4:2000 (CEN 2000) is used to measure individual particles by slide calliper to determine the shape index.

3 EVALUATION OF SHAPE DESCRIBING QUANTATIES

To evaluate different shape describing quantities and definitions both usability and sensitivity are relevant.



3.1 Usability Regarding usability the connection between definitions and geotechnical parameters would be the most relevant factor. This is not touched in the present study. A more simplified way of looking upon the usability of different definitions is to compare images of grains. The comparison, which is fully described in Johansson & Vall (2011), involves particles which seem to have different shapes; i.e. some grains looking elongated, some others that do not. Some particles looking angular and some that are looking smooth/rounded. Furthermore, the grains are analysed with the software ImageJ (further described in section 4.4). The analysed quantities are AR (defined as eq. 12 in Table 2-1), Circularity (defined as eq. 5 in Table 2-1), and Solidity (defined as eq. 8 in Table 2-1, but with areas instead of perimeters).

3.2 Sensitivity In this study has been carried out a sensitivity analysis regarding image resolution and geometrical definitions. Five idealized well defined geometries have been used to study the effect of resolution. The known geometrical properties, i.e. area, perimeter, etc., are compared with the analysis software results. Further on, a test on soil particles has been performed.

The idealized geometries (square, triangle, circle, star and cross) are presented in Figure 3-1.

Figure 3-1 Idealized geometries.

In order to evaluate the sensitivity different image resolutions are studied. Henceforth, the resolution is defined as the side, s (expressed in pixels) as indicated in Figure 3-1. The square, the triangle and the circle wereall built using six different resolutions. Thestar was built by one central square and fourtriangles put on each of the four sides of thesquare. The cross is formed similarly as thestar but with four squares surrounding thecentral one. In Table 3-1 the layout ofinvestigated resolutions as well the areas ofthe analyzed geometries are shown. Theresolutions are grouped in orders, 1-6depending on the resolution, s.

Table 3-1 Area, A and side, s of the geometries are listed for each of the resolutions (1st, 2nd, etc.).

Figure Order

1st 2nd 3rd 4th 5th 6th

Square A [pixels2] 100 400 1600 6400 25600 102400 S [pixels] 10 20 40 80 160 320

Circle A [pixels2] 78 314 1256 5026 20106 80424 S [pixels] 10 20 40 80 160 320

Rectangle A [pixels2] 50 200 800 3200 12800 51200 S [pixels] 10 20 40 80 160 320

Star A [pixels2] 80 405 1805 7605 31205 126405 S [pixels] 5 10 20 40 80 160

Cross A [pixels2] 75 300 1200 4800 19200 76800 S [pixels] 4 9 19 39 79 159



The Image-Pro Plus software from Media Cybernetics was used to carry out the measurements. Basically there were taken seven measures; diameter of inscribed circle, diameter of circumscribing circle, area, particle diameter, perimeter, and convex perimeter. All the measures can be seen in Table 2-1. Regarding the diameter, two different techniques were used. The first used was the “maximum feret box”. In this case the diameter is defined as a straight line (the longest that can possibly be drawn) measured between two parallel tangents of the particle’s boundary. The second used was the diameter trough the centroid of the particle. By usage of the latter definition the average diameter can be determined. The convex perimeter can be defined as the length of a string stretched around the tips of all possible Feret diameters; i.e. a fictitious elastic band stretched around the particle, see eq. 8 in Table 2-1.

Based on the measures by the software seven shape quantities of large or intermediate scale and two quantities of roughness was evaluated. The definitions are presented in Table 2-1 and numbered 1-9.

For each of the definitions the analyzed measures are compared to the true values. The comparison is carried out by calculating a deviation, defined as:

IGTIGTDeviation SGT - (13)

where IGT is the Ideal Geometrical Term (pixels) and SGT is the Software Geometrical Term (pixels).

Besides the analysis on the idealized geometries, soil particles were used. As an extension on the sensitivity analysis presented in Johansson & Vall (2011), three microscope camera pictures of the same soil particle, taken using three different objectives with different magnification rates, were analyzed. The procedure regarding as well acquisition as analysis of the pictures is described in section 4.

4 STUDY OF IMAGE ANALASYS APPLIED ON SOIL PARTICLES

The laboratory work, consisted of image acquisition has been carried out by usage of as well a microscope camera as a conventional digital SLR.

4.1 Equipment The used microscope is named Motic B1; it is equipped with lightening sources from above and below. There are three lenses with magnification rates of 4x, 10x, and 40x. The camera mounted on top of the microscope is named Infinity 2 and has a 2 megapixel resolution. The SLR is a Nikon D80 equipped with a macro lens with a focal length of 55 mm. The equipment used for imageacquisition was arranged as shown in Figure4-1.

Figure 4-1 To the left is shown the microscope with the top mounted camera connected to the computer on which the soil particles are previewed. To the right is shown the SLR.

4.2 Sample preparation Samples of dried soil were used. Pictures were captured on mixed soil particles, as well as on sorted i.e. sieved material. The sieving work was carried out with a conventional stack of sieves placed on a vibrating plate , and ended up with soil samples of the fractions 0-0.063 mm, 0.063-0.125 mm, 0.125-0.25 mm, 0.25-0.5 mm, 0.5-1.0 mm and 1.0-2.0 mm.

4.3 Image acquisition Before the shooting and the analysis work was initiated, some preparations were done. Different directions of lightening were tested, pictures of different particle size were taken, and the software Infinity Capture 5.0.4, for



controlling the microscope camera, was tested. The variables contrast, brightness, white balance, gamma and gain were varied, whereupon pictures of different type, i.e. with different features, were taken. Moreover, different microscope lenses were used; tests with varying rates of magnification and different particle sizes were carried out. The acquisition of pictures taken with different settings was done in order to make further comparison and optimization possible.

4.4 Image analysis In general, there are a number of different techniques for assimilating information from a taken picture (Mora et al., 2000). In Persson (1998) is described one procedure mainly made up by six steps: capturing, normalization of the grayscale, segmentation, filtering and filling, grain separation, and definitions of outlines.

5 RESULTS AND ANALYSIS

5.1 Geometrical quantities Usability Here is shown results from the simplified usability analysis. The selection of particles is done according to the scale based definitions explained in section 2.1. The used quantities are circularity, aspect ratio (AR), and solidity and the calculations are done using eq. 5, 12, and 8 (but with area instead of perimeter), found in Table 2-1. In Figure 5-1 are seen three particles which are judged to be spherical and three more elongated, respectively.

Figure 5-1 The three grains to the left are judged to be relatively spherical, and the grains to the right to be more elongated.

In Table 5-1 result values coming from the image analysis are presented.

Table 5-1 Image analysis results regarding the particles in Figure 5-1. ID Circularity AR Solidity 1 0.764 1.202 0.943 2 0.809 1.118 0.958 3 0.822 1.159 0.960 4 0.604 1.919 0.913 5 0.590 1.999 0.894 6 0.563 2.403 0.906

To the left in Figure 5-2 are seen particles which are judged to be elongated and rounded. To the right are seen two particles judged to be more spherical but angular.

Figure 5-2 The two grains to the left is judged to be relatively elongated and rounded, and the grains to the right to be more spherical and angular.

In Table 5-2 result values coming from the image analysis are presented.

Table 5-2 Image analysis results regarding the particles in Figure 5-2. ID Circularity AR 7 0.593 2.640 8 0.578 2.562 9 0.660 1.079 10 0.589 1.419

It is concluded that AR-values can be used apart from other values and still give information about the shape of the grain. On the contrary, values of circularity have to be combined with values of other parameters, in order to be used as an indicator on a particle’s angularity or large scale shape, respectively.

Sensitivity In Figure 5-3 deviations calculated by usage of eq. 13 are presented. The deviations selected to be graphically presented origins from analysis of the triangle and rectangle-geometries. The patterns of the curves from other investigated geometries are similar to the presented.



Figure 5-3 Deviations from ideal behaviour of the different quantities, plotted versus resolution of the triangle geometry.

The overall trend is that a higher resolution results in a lower deviation. Moreover, there are other variables as well. For instance, at a resolution of s = 10, eq. 1 applied to the geometries circle, square and cross, results in deviations lower than 5 %. The same definition applied on the star and the triangle results in higher deviation values. When the resolution is increased from the 1st to the 2nd

order, usage of eq. 1, 2, and 6, results in deviations lower than 10 % for all of the geometries. For the rest of the equations the resolution needs to be increased even more (to the 3rd order) in order to get deviations lower than 10 % obtained. This shows that eq. 1, 2, and 6 are less sensitive than the others.

In Figure 5-4 and Figure 5-5 is seen to what extent the quantities defined by eq. 8 and 9 in Table 2-1 (by the reference authors called roughness), are affected by the resolution. Regarding eq. 9 both the ferret diameter and the centroid crossing one are used.

Figure 5-4 The quantity roughness’ dependence on resolution of cross geometry.

Figure 5-5 The quantity roughness’ dependence on resolution of star geometry.

The use of the centroid diameter is more unstable than the feret measurement.

Deviation is found to be influenced by as well resolution as the combination of definition and analysed geometry.

On soil material, one single soil particle, seen in Figure 5-6, was analysed in three different images. Since the images were taken at different magnification levels, the effect of the magnification on the geometrical quantities was valuated.

Figure 5-6 Soil particle for investigation of effect of magnification on geometrical quantities.

In Figure 5-7 the result from the sensitivity analysis are presented. On the x-axis are presented the three magnification rates. On the y-axis are seen the calculated values of the quantities.

Figure 5-7 Calculated values plotted for the three different magnification rates.

The results show that there is an effect on the result for the different geometrical quantities depending on the zoom-level on the particle



and that some quantities are more sensitive than others. The graphs based on eq. 2, 5 and 7 are showing values of the shape quantities, varying by changed resolution. All of these three quantities are dependent of the perimeter of the particle.

6 DISCUSSION

6.1 General The use of image analysis on particle shape classification is promising but need further development. Besides the pros and cons regarding the actual performance of image analysis, the subject itself is fraught with uncertainties. The misusing of quantities and definitions, in detail discussed in Johansson & Vall (2011), is definitely faced even during performance of this present study. The scattered way on which terms are used needs to be homogenized. Further research should aim on standardization of as well analysis procedures as terminology in order to ensure an objective particle shape classification.

6.2 The methodology It is found that images of particles with diameters of 0.125-1.0 mm, taken with the microscope camera, were successfully analyzed. Regarding usage of the SLR, a diameter of 2.0 mm, were found to be a lower limit. The gap identified between the fraction for which good quality results were retrieved by usage of the microscope camera, and by usage of the SLR should not be too problematic to eliminate. It can probably be done by usage of other microscope lenses.

Even though valuable advantages as reduced influence of subjectivity and possibility of rational efficiency in analysis are achieved, there are disadvantages not to neglect; e.g. problems related to aggregating particles, and not satisfying focus range in the image. This makes it important to get the soil dried before performance of the image analysis. Focus and angle of image acquisition is important for the analysis result. Focus affects the interpretation of the boundaries by analysis program and the

photo angle affects the analyzed projection of the particle.

Regarding the analysis part it is stated that none of the tested applications for image analysis (neither ImageJ nor Image Analysis Pro) permits determination of the intermediate scale quantity roundness as defined in eq. 10 in Table 2-1. This means that values for application of existing soil parameter relations that include the roundness cannot be done without developing the tools.

In order to get representative results, the image acquisition should be carried out aiming at getting as many particles as possible imaged simultaneously. It can be stated that there is needed some balancing work to get the procedure fast and efficient, but still get results of sufficient quality.

6.3 Expressions and definitions It is to be emphasized that the scale approach regarding quantities used for description of shape is not a general one, and that all names of quantities and all definitions are just suggestions from different authors. Still, the breakdown based on the scales is found to be quite practical and useful.

Usability It is important to reflect on the meaning of specific values of different shape describing quantities. According to what is concluded in the usability part of section 5.1, it is stated that there are quantities that do not give unambiguous information if they are used by themselves. On the other hand, these quantities might be very useful if the determined values are combined with values of other definitions.

To investigate the usability of different quantities, the possibility of getting them determined is also to be considered. In the end, usability is a matter of as well the definition of the shape describing quantity, as the possibility of getting a reliable value. It is stated that quantities depending on particle area, particle perimeter, and different types of diameters, all can be determined and valuated. Still, these are not as usable in existing relations between particle shape and



geotechnical properties, as e.g. roundness (defined as in eq. 10 in Table 2-1) is. The latter one is, on the contrary, more difficult to determine.

Sensitivity analysis Resolution has an important role when image analysis is carried out; it is necessary to determine the minimum resolution acceptable (it may vary depending on the goal of the classification) in order to obtain low deviations. It is easy to get pictures of good quality when the analysed particles are big enough; it is more complicated when the particle size diminish. In such cases it might be necessary to use more sophisticated and expensive equipment. In this study five ideal geometrical figures were tested; this is of course a small spectrum of all possible particle shapes. To get a large, rich and varying base for all type of studying, it is of course important to perform analysis on real particles (soil or not). Still, the idealized geometries are very suitable for this type of theoretical key study.

The acceptable deviation limit in this study was chosen to be <10%. To keep the deviation below 10%, the eq. 1, 2 and 6 should be limited by a minimum area of 405 square pixels. For the rest the limit is found to be 1805 square pixels. These areas correspond to soil particles of the 2nd and the 3rd order of resolution, respectively. If the sensitivity results in this study are extrapolated, the authors recommend performance of a simple resolution test, to obtain reasonable deviations.

The use of the diameter crossing through the centroid of the analysed geometry gives for the triangle deviations up to 100%. This fact makes it reasonable to avoid usage of this measurement, and also confirms that existing quantities and definitions have to be used carefully.

Regarding the sensitivity analysis carried out on actual soil particles – results shown in Figure 5-7 – it can be concluded that definitions including the particle perimeter are the most sensitive ones. The plotted results are showing that eq. 2, 5 and 7, which all contain the perimeter, are not stable when

resolution changes. Increased rate of magnification, leads to decreased focus along the boundary, which in turn results in increased length of the outline (higher number of pixels) and, furthermore, changed affected values.

6.4 Evaluation of image analysis as a method applied on soil particles

Image analysis is a promising method for shape classification on soil particles. It is objective and the procedure could in large extent be automated. Traditional methods such as subsequent sieving procedures by combining sieves of different mesh geometries and manually scaling are time consuming and are limited to describe simple geometric descriptors.

7 CONCLUSIONS

Although there are a lot of different ways of defining shape and describing quantities, the breakdown based on the scales is found to be quite practical and useful.

The fact that a soil’s tendency to aggregate is increasing with decreased particle size and increased water content, makes it important to get the soil dried before performance of the image analysis.

To permit performance of further studies on the connections between shape and geotechnical properties, used tools i.e. the image analysis applications, are to be developed and optimized.

To avoid deviations that may influence the results from image analysis processing, the minimum resolution has to be taken into consideration.

Perimeter is a key factor of changing results when the resolution changes.

8 FURTHER WORK

Further research should aim on standardization of analysis procedures and terminology usage.

The procedure from sample preparation to interpretation of the image acquisition needs to be further developed focusing to establish



a methodology that ensures objective and useful results.

The effect on chosen geometrical definitions in the particle shape classification needs to be compared to today empirical knowledge of influence of particle shape on soil properties as a first step to connect the image analysis methodology to be incorporated into geotechnical analysis.

9 ACKNOLOWDGEMENTS

The authors want to thank MSc. Jakob Vall at WSP for participating in the pre-study of image analysis of particle shape.

10 REFERENCES

Arasan, S., Hasiloglu, A. S., Akubulut, S. (2010) Shape properties of natural and crushed aggregate using image analysis. Int .J. Civil and Structural Engineering. Vol. 1, No. 2. ISSN 0976-4399.

ASTM (2005). Standard ASTM D 4791 : Standard test method for flat particles, elongated particles or flat and elongated particles in coarse aggregate.

Blott, S. J. & Pye, K. (2008). Particle shape: a review and new methods of characterization and classification. Sedimentology. 55 (1), 31-63.

Cho, G., Dodds, J. & Santamarina, J. C. (2006). Particle shape effects on packing density, stiffness and strength: nautal and crushed sands. Journal of geotechnical and geoenvironmental engineering. 132 (5), 591-602.

Folk, R. L. (1955). Student operator error in determining of roundness, sphericity and grain size. Journal of sedimentary petrology 25, 297-301.

Janoo, V. (1998). Quantification of shape, angularity and surface texture of base coarse materials. US army corps of engineers cold region research. Special report 98-1.

Johansson, J. & Vall, J (2011). Jordmaterials kornform; inverkan på geotekniska egenskaper beskrivande storheter bestämningsmetoder. Luleå university of technology. Master thesis. (In Swedish)

Krumbein, W. C. (1941). Measurement and geological significance of shape and roundness of sedimentary particles. Journal of sedimentary petrology 11 (2), 64-72.

Krumbein, W. C. and Sloss, L. L. (1963). Stratigraphy and sedimentation. Second edition. Freeman. San Francisco.

Kuo, C., Rollings, R. S. & Lynch, L. N. (1998). Morphological study of coarse aggregates using image

analysis. Journal of materials in civil engineering. 10 (3), 135-142.

Mitchell, J. K. & Soga, K. (2005). Fundamentals of soil behavoir. Third edition. Wiley. New Jersey.

Mora, C. F. & Kwan, A. K. H. (2000). Sphericity, shape factor, and convexity measurement of coarse aggregate for concrete using digital image processing. Cement and Concrete Research 30 (3), 351-358.

Pentland, A. (1927). A method of measuring the angularity of sands. MAG. MN. A.L. Acta Eng. Dom. 21. XCIII.

Persson, A. L. (1998). Image analysis of shape andsize of fine aggregates. Engineering Geology. 50 (1-2), 177-186.

Powers, M. C. (1953). A new roundness scale for sedimentary particles. Journal of sedimentary petrology 13 (1), 28-34.

Riley, N. A. (1941). Projection sphericity. Journal of sedimentary petrology. 11 (2), 94-97.

Rouse, P. C., Fennin, R. J. & Shuttle, D. A. (2008). Influence of roundness on the void ratio and strength of uniform sand. Geotechnique 58 (3), 227-231.

Santamarina, J. C. & Cho, G. C. (2004). Soil behavior: the rol of particle shape. Proceedings Skempton Conference, London.

Shinohara, K., Mikihiro, O., Goldman, B. (2000). Effect of particle shape on angle of internal friction by triaxial compression test. Powder technology 107, 131-136.

Skredkommissionen. (1995). Anvisningar försläntstabilitetsutredningar. Ingenjörs-vetenskapsakademin, rapport 3:95, Linköping. (In Swedish)

Stückrath, T., Völker, G., Meng, J. (2006). Classification of Shape and Underwater Motion Properties of Rock. Tainan: National Cheng Kung University.

Szadeczky-Kardoss, E. (1933). Die bistimmung der abrollungsgrades. Geologie und paläontologie. 34B. 389-401.

Tickell, F. G. & Hiatt, W. N. (1938). Effect of the angularity of grains on porosity and permeability. Bulleting of the American association of petroleum geologist. 22, 1272-1274.

Wadell, H. (1932). Volume, shape and roundness of rock particles. Journal of geology. 40, 443-451.

Wadell, H. (1935). Volume, shape and roundness of quartz particles. Journal of geology. 43 (3), 250-279.

Witt, K. J. & Brauns, J. (1983). Permeability- anisotropy due to particle shape. Journal of geotechnical engineering 109 (9), 1181-1187.

Paper III

Rodriguez, Juan M. and Edeskär, Tommy. Case of Study on Particle Shape and Friction Angle on Tailings. Journal of Advanced Science and Engineering Research.

Journal of Advanced Science and Engineering Research Vol 3, No 4 December (2013) 373-387

Case of Study on Particle Shape and Friction Angle on Tailings

Juan M. Rodriguez and Tommy Edeskär Luleå University of Technology, Sweden

SE-971 87 LULEÅ, Sweden [email protected], [email protected]

Article Info Received: 10/6/2013 Accepted: 28/8/ 2013 Published online:1/9/2013

ISSN 2231-8844

Abstract

Tailings are crushed and milled materials result of the mining production. Tailings need to be stored in facilities, usually tailings dams, for a long time period for mainly safety and environmental protection. In order to design tailings dams in a long term perspective not only current material properties is needed but also future changes of these properties due to e.g. weathering. On a particle level the weathering will result in shape changes and decomposition. By studying the changes in shape a prognosis of changes in properties of a tailings deposit may be established. Tailings are site specific material and are not well investigated compared to natural geological materials such as soil. Tailings materials size ranges generaly from sand to silt and the particle shape by genesis or production processes. Based on laboratory tests tailings from the Aitik mine has been investigated through triaxial tests and particle shape quantification by two dimensions image analysis. The shape descriptors Aspect Ratio, Circularity, Roundness and Solidity are used in this study. These shape descriptors are evaluated based on how well these describes talings materials. The evaluated shape descriptors are used in previous published empirical relations between shape and friction angle. As reference are friction angles evaluated by triaxial tests on the material used. The results show that the particle shape is affected by the size of the aggregates. Aggregates in small fractions are more elongated and less rounded, i. e. more angular, compared to larger. Furthermore, the Aspect Ratio and Circularity seems to be the most situable quantities to describe the tailings behaviour in relation with the empirical model. The accuracy in predicting the friction angle of the tailings by previously published relations based on uniformly graded sand material are low. But the systematic underestimation of the friction angle indicates that it would be possible to establish such empirical relations based on tailings material. Keywords: Particle shape, Quantities, Image analysis, Friction angle.


374

1 Introduction

Tailings are crushed and milled by products from ore refining and generally considered to be angular aggregates in the size range from silt to fine sand (FHA, 1997). From an engineering perspective in general the strength and deformations properties of the tailings are regarded as a natural soil material in the same size range and size distribution.

Garga et. al. (1984) classifies the shape of tailings to be in the range from angular to sub angular. The high angularity contributes to high initial friction angles and, as a consequence of the deposition methods to high void ratio and loose fills (Holubec and D’Appolonia, 1973; Rousé et. al., 2008, among others). It is also likely that physical (e.g. stresses) and chemical (e.g. oxidation) weathering in a long time perspective will result in less angular aggregates. Yoginder et. al. (1985) show that angular aggregates are more sensitive to shape changes due to edge breakage and subsequently generation of fines under increasing confining stress compared to natural geological material. Physical properties e.g. the friction angle are shape dependent and a reduction in angularity will reduce the friction angle (Cho et. al., 2006). Empirical relations have been suggested by authors as Cho et. al. (2006) and Rousé et. al. (2008) in studies where the friction angle (from triaxial tests) and particle Roundness according to Wadell (1932)-definition are correlated. Uniform graded soil samples, basically sand, were used to determine these empirical relations. The tailings investigated in this study contain a wider range of particle sizes and especially more fine graded fractions such as silt and clay compared to these studies. The selection of comparative empirical relations for predicting the friction angle is based on the laboratory test procedure (triaxial tests) and the range of friction angle covered by these suggested relations.

Depending on the mineral composition and the deposition conditions chemical weathering may be an important factor to account for in a long time perspective. e.g. sulphide rich tailings are highly susceptible to oxidize due to the large surface exposure of the grains. Furthermore the presence of oxygen and water (no saturated) provide an adequate environment for sulphuric acid production (Al-Rawahy, 2001) that could foment the increase of oxidation and also promotes the particles shape change. If tailings dams are considered to be designed as “walk-away”-solution or to be safe in a thousand year perspective it is important to account for both the change in properties and the consequential global effect on the tailings dam.

There are more than 40 quantities (shape descriptors) compiled in the report by Rodriguez et. al. (2012) able to determine the particle shape, but there has been only minor efforts to evaluate the suitability among these descriptors in relation with the effect on the soil properties. Authors such as e.g. Cheshomi et. al. (2009) and Cho et. al. (2006) had used the comparison chart develop by Krumbein and Sloss (1963). According to Folk (1955) there is an appreciation error involve in chart comparison as methodology. The lack of possibility to objectively describe the shape hinders the development of incorporating the effect of particle shape in geotechnical analysis. Computer algorithms had been developed for at least half of the quantities and the use of image analysis reduces dramatically the processing time and provides reproducibility.

In this work Aspect Ratio (AR), Roundness (R), Circularity (C) and Solidity (S) were chosen quantities to analyze trough image analysis software.


375

2 Scope of study The scope of the study is to describe tailings by shape and relate the shape as indicative

measure for physical properties. The aims of the study are to classify tailings by shape using 2D image analysis technique in a case study on tailings from Aitik. The classified tailings by different shape factors are correlated to the friction angle by applying previously published empirical relations based on soil material compared with triaxial test results on the tailings material in order to investigate if this methodology could be applicable as indicative measure of physical properties on tailings.

3 Methodology

In this study has tailings from Boliden’s mine Aitik outside Gällivare in northern Sweden

been analyzed. The target minerals are mainly copper and gold and the ore is of sulphide type. The analysis includes basic geotechnical characterization and particle analysis. The geotechnical characterization includes sieving, compact density test and active triaxial tests. The analysis of the individual particles has been done by image analysis for shape description and x-ray analysis for classification of oxidation potential. The results from the particle shape description and laboratory results are compared with previous published empirical results on the effect of particle shape on strength properties.

Undisturbed samples have been collected by coring. Four samples were analyzed from Aitik tailing dam. Two of the samples are located in the surrounding area of the corner of the tailings dam bodies where GH-EF meets as shown in figure 1.

I, Position: 61+900@150m. Samples “a” and “b”. II, Position: 62+643@150m. Samples “c” and “d”

Figure 1: Aerial view of part of the Aitik Tailings dam to the right in the picture sampling

locations (I) and (II).(Photo courtesy of Boliden Mineral AB.)

3.1 Geotechnical characterization and triaxial tests

The sampled material was characterized by sieving, ocular inspection, determination of water quotient, compact density and bulk density.


376

Active triaxial test were performed on three samples in a confining stress-ranges of 90-150 kPa in undrained conditions (see table 1). One sample was executed in drained conditions. The samples used were retrieved by undisturbed coring. After basic characterization of the cores the triaxial tests were performed. The used triaxial equipment used was GDS Instruments based on Bishop and Wesleys (1975) principle of triaxial apparatus. The samples had, been step by step, isotropic confined. During this consolidation a backpressure of 100 kPa has been applied. In the case a maximal deviatoric stress (peak) was registered during shearing the friction angle was evaluated at both the maximal deviatoric stress and at the residual stresses. For those samples where a maximal deviatoric stress was not registered the friction angle was evaluated at 15 % compression was used.

Table 1. Triaxial tests used in this study.

Location Id Sampling depth [m]

Triaxial test Confining pressure [kPa]

I a 13.3 Drained 140 I b 13.3 Undrained 150 II c 8.3 Drained 90 II d 16.8 Drained 170

3.2 Visual inspection and X-ray

Visual inspection was performed using the undisturbed samples under a traditional magnification lens. For the X-ray test the PANalytical Empyrean X-ray Diffractometer equipped with PIXcel3D detector and X-ray tube Empyrean Cu LFF HR was used.

3.3 Image analysis

One fraction of each undisturbed cored samples was used to obtain the grains subject to image generation. In order to obtain clear images it was needed to sieve each fraction sample by wet sieve in five standard sieves mesh (1, 0.5, 0.25, 0.125 and 0.063mm). The particle size separation provides better focus in the microscope. The used microscope was Motic B1, with two lenses used with magnification rates of 4x and 10x (for particle size of 0.063mm). The camera mounted on top of the microscope (Infinity 2) and has a 2 megapixel resolution. It is equipped with lightening sources from below (straight light from source to lens) and external lightening able to move and locate in any position. The external lightening source was chosen (also from the bottom of the sample) due the possibility of change the light directional angle. The ability to have a non-unidirectional light source provide good contrast in the particle’s outline specially in mica minerals (light passes through easily) that usually are not well defined and there is not enough contrast when unidirectional light is applied. A total of 160 to 300 of particles (approximated) by sample were measured (see Appendix).


377

Table 2. Shape describing quantities and their mathematical definitions

A, area AC, area convex P, perimeter Major, major axis based on fitting ellipse Minor, minor axis based on fitting ellipse

Figure 2: Definition of the quantities applied in this study. From left: Circularity (C),

Roundness (R), and Solidity (S).

Images were analyzed with software ImageJ and four descriptors were determined: Aspect Ratio (AR), Roundness (R), Circularity (C) and Solidity (S). The definitions of these quantities are presented in table 2 and figure 2.

3.4 Correlation of shape descriptors by physical properties

A set of four databases were compiled, each database correspond to one triaxial test sample. Each database contains values for the four quantities (see Appendix) and the five sieving ranges (1, 0.5, 0.25, 0.125 and 0.063mm). Similar statistical analysis was performed among the four databases as follow: the minimum and maximum values for each quantity along the five sieving sizes per sample were obtained. Average values were obtained from each size range per sample and quantity. Median values for each sieve size were also obtained per sample. The quantity values were applied in the empirical relations in table 3, to investigate if this methodology could be useful for prediction of indicative physical properties.

Table 3. Empirical relations suggested in literature relating the friction angle (ϕ’) and the

shape quantity (Q).

Eq. # Quantity Definition Reference Eq. # Quantity Definition Reference

1 Circularity (C)

Cox (1927) 2 Roundness

(R)

Ferreira And Rasband (2012)

3 Solidity (S)

Mora And Kwan (2000)

4 Aspect Ratio (Ar)

Ferreira And Rasband (2012)

A, Area Ac, Area Convex P, Perimeter Major, Major Axis Based On Fitting Ellipse Minor, Minor Axis Based On Fitting Ellipse

EQ. # DEFINITION REFERENCE

5 Cho et. al. (2006)

6 Rousé (2008)

Q, quantity value (0 to 1)

Q1742'

Q4.147.41'

MinorMajor

2PA4

2Major 4A

CAA


378

4 Results

4.1 Geotechnical characterization and triaxial tests

Figure 3. Sieving curves for the Aitik tailing samples (SWECO Geolab, 2007)

The samples are classified as silt, with the exception of a and b that is a sand-silt material

(see sieving curve fig. 3). The sieving curves a and b are based on other samples than used in the triaxial tests but from corresponding depth, 12.2m. The amount of clay, or fines, ranges up to 20 % of the samples.

Ocular inspection on the samples shows the presence of mica, except sample c. In this sample some oxidation is recognized consisting of reddish colored zones. The red color is presumed to be the result of the iron oxidation coming from pyrite and chalcopyrite that is commonly present in the tailings. The X-ray tests also show the presence of mica (biotite and muscovite).

The results from the basic geotechnical characterization and the triaxial tests are compiled in table 4. As seen in the table the result of an evaluation of the friction angle by Mohr-Coloumb failure criterion will generate high difference depending on a cohesion intercept is used or if the cohesion intercept is omitted (c=0 kPa). By inspection of the samples the material could be considered to have low or moderate cohesion depending on the clay content. Thus is the approximation of the failure envelope to be straight, as in the Mohr-Coloumb failure criterion, not valid for high stress intervals, or at least for the lower ranges of effective stress. The initially loose specimens has equally high, or higher friction angle compared to the firm. Dilatant behavior during shearing was observed for all the initially firm specimens including the undrained specimen and contractant behavior for the loose samples.

4.2 Particle shape measurement results

By visual inspection and Powers (1953) roundness chart of the tailings samples the particle shape are subjectively classified to range from sub angular to very angular. The smaller fractions appear to be more angular compared to larger fraction.


379

Figure 4 presents the condensed results from the 2D image analysis. The analysis is done on samples split by five sieve sizes based on sieving, shape descriptors and its output values. Table 4. The results from the basic geotechnical characterization and triaxial tests. Sample (-)

is in this study only used for evaluation of strength properties ID a - b C d Sample location I I I II II Sampling level [m] 13.3 20.0 13.3 8.3 16.8 Ocular classification saSi clSa clSi siSa siCl Initial state Firm Firm Firm Loose Loose Condition Drained Drained Undrained Drained Drained

[t/m3] 1.96 1.97 1.93 1.76 1.89 w [%] 31-39 33-36 29-34 49-54 40-50 Initial conditions [kPa]

'3

'1

u

140 250 110

200 300 100

150 270 120

90

200 110

170 280 110

Compression rate. [mm/h]

1.0 0.9 2.0 0.6 0.6

Failure criterion Max dev.str.

Max dev.str.

Max dev.str.

15% compr. Max dev.str.

Compr. at failure [%] 10.4 9.6 14.2 15.0 14.4 '1 at failure [kPa] 564 826 1170 484 794 '3 at failure [kPa] 141 201 253 90 170

u at failure [kPa] 109 99 17 110 110 ’ (c = 0) 40º 37º 40º 43º 40º ’ and c c = 79 kPa, ’ = 29º c = 39 kPa, ’ = 37º

4.3 Particle shape measurement results

By visual inspection and Powers (1953) roundness chart of the tailings samples the particle shape are subjectively classified to range from sub angular to very angular. The smaller fractions appear to be more angular compared to larger fraction.

Figure 4 presents the condensed results from the 2D image analysis. The analysis is done on samples split by five sieve sizes based on sieving, shape descriptors and its output values.

AR and Roundness are inverse each other for individual measurement, if mean value is applied this relation is no longer valid. AR-1 was introduced and refers to the AR but in order to obtain an AR with range between 0 and 1 the conversion took place. Appendix contains a detailed table with the quantity values.


380

Figure 4: Box-plot for analyzed tailings material grouped by size determined by sieving

Median values were omitted in figures due to there is no significant difference to the average value (see appendix).

The complete database (the four tailing samples a, b, c and d) is presented in figure 4. It is visible in all quantities that bigger particles are more uniform. In general, for all quantities, the uniformity showed in big particles diminishes for smaller particles. It is also evident (in quantities ranging from 0 to 1) that the box plot (between 1st and 3th quartile), the average value and the minimum values move downwards in the vertical axis while the maximum values seem to have no apparent change (close to the upper limit).

Figure 5 compares the shape descriptor in the individual samples an figure 6 the individual samples variation by size for each shape descriptor. AR changes are more evident but it could be relative to the values range. In general AR decrease as the particle size increase. The rest of the quantities increase as the particles size increases. Sample c is possibly the exception due to this sample has no major changes when quantities are applied, in this sample oxidation was observed. AR-1 and Roundness have exactly the same behavior, it is due to the fact both empirical relations represent the inverse of the Aspect Ratio (Ferreira and Rasband, 2012).

Figure 5. Mean value of descriptors for the different particle sizes in each tailings sample a-d.

AR is the abbreviation for Aspect Ratio.


381

Figure 6. Mean value of the quantities in sample a-d and size based on sieving.

X-ray tests show that only sample c contains iron oxide in the form of FexOy. Solidity

(fig. 6) is able to show the value increase when particle size increases in all samples while for the rest of the quantities it is not that defined for sample c. Sample c (oxidation presence as FexOy) does not show clear increase or decrees in values when quantities as Roundness, Aspect Ratio and Aspect Ratio-1 are applied. Circularity is showing a clear value increase on sample d while for the rest of the samples a higher peak value appears in some cases in small sizes and in others in medium sizes thus, from Circularity (except sample d) is not clear the size-shape increasing or decreasing behavior. Roundness and AR-1 have the same result and represents the invers of the AR (see the mirror image comparing AR and AR-1). Samples values on Roundness and Aspect Ratio-1 (except sample c) increase as particle size increase and for Aspect Ratio the values decrees as particle size increase (except sample c).

Figure 7. Percentages of the total particles by size and quantity value for sample d (A,

Solidity; B, Roundness; C, AR-1 and D, Circularity) Figure 7 present the relative amount of the analyzed particles (by size) that falls in the

different ranges for A) Solidity, B) Roundness, C) AR-1 and D) Circularity for sample d. The results from samples a and b are similar as d, always appearing low values in the small fraction of size. Sample c, where oxidation is present contains more regular shape along the size fractions (see Fig. 5). It was decided to not show samples a and b in figure 7 due to the similarity with sample d. Also to sample c plotting was avoided because it presents no change on percentages along the sieving sizes resulting in a flat behavior.


382

4.4 Comparing friction angle and empirical relations

The results from the 2D-image analysis were applied in the equations 5 and 6 in table 3. Since all grains were analyzed the output was a range of numerical values of each shape descriptor. In this study was the average, minimum and maximum value of each shape descriptor tested in the equation. As reference to the output friction angle was the results from the triaxial tests on the samples in table 4.Table 5 and figure 8 shows that the minimum value of each shape descriptor presents less difference between the expected empirical friction angle (ϕ’empirical) and the laboratory tests output (ϕ’triaxial). Among the quantities Circularity has the highest accuracy of predicted friction angle. Table 5. The difference between the reference friction angle and the estimated friction angle

(ϕ’triaxial – ϕ’empirical) ϕ’triaxial – ϕ’empirical [°]

Quantity 5 6

AR (converted to 0-1) average 13.0 8.8 maximum 14.5 11.1 minimum 10.7 5.4

Circularity average 12.3 7.8 maximum 14.3 10.8 minimum 8.6 2.3

Solidity average 15.2 12.1 maximum 15.8 13.1 minimum 13.0 8.9

Roundness same as AR-1 average 13.3 9.3 maximum 15.9 13.2 minimum 10.9 5.7

Figure 8. Friction angle results, lines represent the laboratory test (ϕ’triaxial) and the points are

the empirical relation output (ϕ’empirical).


383

5 Discussion

5.1 Particle shape measurement results Tailings had been classified as very angular to sub angular material by visual inspection

based on Powers (1953) comparison chart. The results of this classification collaborates with the conclusions of the general shape of tailings, e.g. by Garga et. al. (1984). The drawback with classification of shape by comparison chart is that the classification is subjective and difficult to quantify. There are other options to determine the angularity. Wadell (1932) is probably the most used and systematic method to determine the roundness, here the opposite to angularity, because it is focus in the corners configuration but still this methodology evolves in chart comparison. If any of the methodologies can be scripted and image analyzed roundness results would be more objective.

Small particles (<0.002mm) tend to be more platy and elongated (Mitchell and Soga, 2005) and this study shows how the elongated particles (relation 2:1, AR-1 value of 0.5) appear and populates the sample as much as 30% (fig. 7C) in the 0.063mm size, Circularity shows the increase of percentage to 50% with low circularity value of 0.5 for 0.063mm size (fig. 7D). The loom of elongated and irregular particles in sizes bigger than those declared by Mitchel and Soga (2005) could be due to that particles involve in this study are not natural geological materials but they are crushed and milled rock from the mining industry. The results of the ocular classification of the angularity also supports that the smaller particles are more irregular than the larger in this study.

Sample “c” was found to present some oxidation, presumable iron oxidation result of the acid leakage. The sample presents no change when quantities are used to determine the shape across the different sizes with exception of solidity. Solidity is probably detecting the surface changes (acid leakage dissolution) while the rest of the quantities do not, this relates the solidity with the third shape scale surface texture. Artificial breakage and fracture formation tend to follow the mineral surface along the weakest lines while acid leakage dissolves surface constituents. This could explain why Solidity is detecting changes in sample c while is not in the rest of the samples. The range that solidity involve (0.86-0.96 : 0.1 in difference) are quiet close, and the range suggest that the particles are quiet regular, this could be the result of the specific samples used and more samples should be included to determine the whole real workability range.

AR is the inverse of AR-1 and Roundness; the results show that the behavior between AR-

1 and Roundness are exactly the same corroborating the Ferreira and Rasband (2013) description of these two quantities. As a result only one of these parameters is needed due to the inverse can be calculated easily. The inverse is applicable only for individual particles measurement and is not for e.g. the mean value (the mean value of a set of data among AR and Roundness will be not inverse each other). The use of these three quantities should depend on the nature of the work (if a range from zero to one is needed Roundness or AR-1 should be preferred). Roundness vs. AR choosing concerns more on the available data to calculate the quantity (Roundness and AR have the major axis in common and they also need the area and minor axis respectively, which data is available should be used) .

Circularity as Cox (1927) state relates the area and the perimeter of a particle with the area and perimeter that a circle should have (circle have area perimeter relation equal to one).


384

There are two inconvenient detected for circularity. 1) Length is the most influenced parameter by resolution (Schäfer, 2002) and perimeter is dependent on length, the error increases as the perimeter increases, it can be added that the quantity use the perimeter to the power of two increasing exponentially the error. Due to the fact that this parameter is influenced by the resolution it should be avoid when possible or use high resolution pictures. 2) Microscope focus for a 3 dimensional particle presents some challenges (characteristics used in the present research) due to the fact that on occasion is not possible to obtain a perfect defined outline of the particle having in this kind of pictures an irregular outline (instead of a regular straight defined line) that possible generates a longer perimeter as it should be.

5.2 Comparing friction angle empirical relations

In this study has previously published empirical relationships between shape and friction angle been evaluated in order to investigate if this methodology is possible to apply on tailings. Since the evaluated empirical relationships has been established on mainly uniformly graded sand and it is not likely that the actual resulting friction angle would be accurate predicted on tailings consisting of both smaller particles and a larger range in grain size distribution. The empirical relationships are based on the shape descriptor Roundness that can both been defined in different ways and also be evaluated differently. Cho et. al (2006) use the Roundness based on the Krumbein and Sloss (1963) modified chart and Rousé et. al (2008) uses the Roundness as Wadell (1932) defined based on a compilation of various authors. In this study has Roundness been automatically quantified by 2D-image analysis applying four different mathematical definitions; Circularity (C), Roundness (R), Solidity (S),and Aspect Ratio (AR). Despite the different methodologies applied to quantify or to qualitative classify the shape of the particles the trend of the friction angle should be similar if the shape descriptors are correct describing the material of interest.

Based on the general behavior of the empirical relations (eq. 5 and 6) the friction angle is likely to be higher in crushed artificial rock than for natural materials since crushed material in general are considered to be more angular (Garga et. al., 1984)

All quantities evaluated describe an underestimation of the reference friction angle from the triaxial tests. In all cases equations 5 and 6 presents maximum to minimum (in that order) difference in relation with the actual friction angle obtained in laboratory tests, been equation 6 the most suitable empirical relation to describe the friction angle for the samples. In the same way the value or statistical parameter minimum for the four quantities present the lowest differences. Underestimation is possible related with the empirical relations (table 3), the maximum value obtained from them is 42 degree and only one triaxial result was over this value (considering also the quantity value as zero). It can also explain that the minimum shape values produce the best agreement with the empirical relations. The limited amount of data shrinks the action area of the empirical relations but state an initial relation for further research and new data acquisition. All positive differences in table 5 represent the underestimation of the tailings friction angle that if apply in the design of tailings dams from the engineering perspective represent the safety of its stability.


385

As point out before the general trend of quantities is to increases in value (from zero to one) while the particle increase in size, it is applicable to individual samples (fig. 5) and for the overall data samples (fig. 4). In the same way empirical relations (eq. 5 and 7) friction angle (ϕ’) increase while the quantities value decreases but, it is unknown if the gradient for both are related due to the lack of more triaxial test with lower friction angles. Data is only available for high friction angles and it is why the minimum values are relating better the empirical relation. Triaxial tests with material covering the range from very angular (quantity values close to zero) to rounded (quantity values close to 1) are required.

The best descriptor determined in this work depends on the final goal, e.g. the use of solidity should be limited to the third scale surface texture due to its ability to recognize differences in the surface. According to Schäfer (2002) lengths present relative error when measuring digital images the use of high resolution could diminish the error but the use of square parameters as in equation 1 an 2 could increase it.

Considering the assumed non-linearity in the Mohr-Coloumb failure envelope, especially for low effective stresses, the use of linear empirical relations between particle shape and friction angle will be valid only above, or in a specific range, of effective stress. The available data in this study is not sufficient to define any stress limits or suggest new empirical relations. But the demonstration of the methodology shows that it is possible to retrieve indicative material properties based on classification of the shape.

6 Conclusions Based on the results in this study following conclusions can be draw:

It is possible by 2D-image analysis to automatically classify particle shape by different shape descriptors. By splitting a sample into different size ranges also multi size samples can be analyzed by this technique.

The investigated tailings are classified as sub angular to very angular. The smaller fractions, silt, appear to be more angular than larger fractions, here sand.

There are empirical relations between shape and friction angle that may be useful to predict friction angle based on the shape of tailings. The investigated previous published empirical relations between shape and friction angle in this study underestimates the reference friction angle obtained by triaxial tests.

7 Acknowledgements This study was funded and supported by Boliden Ltd, Luleå University of Technology (LTU) and the University of Sonora (UNISON). The authors thank Lic. Eng. Kerstin Pousette at LTU for the triaxial testing of the tailings material.


386

8 References Al-Rawahy, Khalid. (2001). Tailings from mining activity, impact on groundwater, and

remediation. Science and Technology. 6, 35-43. Bishop, A. W. and Wesley, L. D. (1975). A hydraulic triaxial apparatus for controlled stress

path testing. Géotechnique. 25 (4), 657-670. Cheshomi, A., Faker, A. and Jones C.J.F.P. (2009). A correlation between friction angle and

particle shape metric in quaternary coarse alluvia, Quaternary journal of engineering geology and hydrogeology. 42, 145-155.

Cho G., Dodds, J. and Santamarina, J. C., (2006). Particle shape effects on packing density, stiffness and strength: Natural and crushed sands. Journal of Geotechnical and Geoenvironmental Engineering. 132 (5), 591-602.

Cox, E. P. (1927). A method of assigning numerical and percentage values to the degree of roundness of sand grains. Journal of Paleontology. 1 (3), 179-183.

Ferreira, Tiago and Rasband, Wayne (2012) ImageJ user guide. FHA (1997). User Guidelines for Waste and Byproduct Materials in Pavement Construction

Publication Number FHWA-RD-97-148, U.S. Department of Transportation, Federal Highway Administration, Washington D.C.

Folk, R. L. (1955). Student operator error in determining of roundness, sphericity and grain size. Journal of Sedimentary Petrology. 25, 297-301.

Garga. Vinod. K.; ASCE, M. and McKay Larry. D. (1984) Cyclic Triaxial Strength of Mine Tailings. Journal of Geotechnical Engineering. 110 (8), 1091-1105.

Holubec and D’Appolonia (1973). Effect of particle shape on the engineering properties of granular soils. ASTM STP 523, 304-318.

Krumbein, W. C. and Sloss, L. L. (1963). Stratigraphy and Sedimentation, 2nd ed., W.H. Freeman, San Francisco.

Mitchell, James K. and Soga, Kenichi (2005). Fundamentals of soil behaviour. Third edition. WILEY.

Mora, C. F. and Kwan, A. K. H. (2000). Sphericity, shape factor, and convexity measurement of coarse aggregate for concrete using digital image processing. Cement and Concrete Research. 30 (3), 351-358.

Powers, M. C. (1953). A new roundness scale for sedimentary particles. Journal of Sedimentary Petrology. 23 (2), 117-119.

Rodriguez, J. M.; Edeskär T. and Knutsson S. (2013). Particle shape quantities and measurement tecniques – A review. Electronical Journal of Geotechnical Engineering. 18 (A), 169-198.

Rousé, P. C., Fennin, R. J. and Shuttle, D. A. (2008). Influence of roundness on the void ratio and strength of uniform sand. Geotechnique. 58 (3), 227-231.

Shäfer, Michael (2002). Digital optics: Some remarks on the accuracy of particle image analysis. Particle & Particle Systems Characterization. 19 (3), 158-168.

Sweco GEOLAB, 2007. Jordprovsanalys. Projekt Aitik labförsök. Uppdragsnummer 216-6092-310. Löp-nr 17095. In Swedish

Wadell, H. (1932). Volume, Shape, and roundness of rock particles. Journal of Geology. 40, 443-451.


387

Yoginder, P. Vaid, Jing, C. Chern and Haidi, Tumi (1985). Confining pressure, grain angularity and liquefaction, Journal of Geotechnical Engineering. 111 (10), 1229-1235.

Appendix. Compilation of the min, max, mean and median shape quantities values in the four samples a-d.

Paper IV Rodriguez, J.M.; Bhanbhro, R.; Edeskär, T. and Knutsson, S. (2016). Effect of Vertical Load on Tailings particles. Journal of Earth Science and Geotechnical Engineering. Vol 6, No 2, pp. 115-129. ISSN: 1792-9040 (printed) 1792-9660 (online). Sciencpress Ltd, 2016.

Journal of Earth Sciences and Geotechnical Engineering, vol.6, no. 2, 2016, 115-129 ISSN: 1792-9040 (print version), 1792-9660 (online) Scienpress Ltd, 2016

Effect of vertical load on tailings particles

Juan Manuel Rodriguez1, Riaz Bahnbhro2 Tommy Edeskär3 and Sven Knutsson4

Abstract

Tailing dams could store hundreds, thousands or millions of cubic meters of tailings result of the mining extractive industry. Mechanical behavior of this man-made soil should be known in order to maintain a safe storage. Dykes rise up to form the dams and they are buildup with the same tailing material especially in the upstream method using the coarse part. The study uses oedometer classical test to determine the load effect over tailing coarse particles. Tailings are site specific and so its characteristics. It is necessary to understand the tailings degradations to achieve safe impounds. The study comprises four samples of one range-size tailing particles (e.g. 1-0.5, 0.5-0.25, 0.25-0.125, 0.125-0.063mm) subject to vertical load in traditional oedometers. Vertical load effects are measured using two dimensional image analysis and sieving. Results show that 0.063mm sample is the only one that has change in shape with low breakage (<1%) while the rest of the sizes have no shape change but high breakage is present especially in fraction 0.5mm. Settlements also are more pronounced in coarse fractions 0.5 and 0.25mm. Keywords: Tailings, Compression, Degradation, Deformation

1 Introduction Tailings are the waste from ore concentration in the mining industry and they are usually deposited in tailings dams. Tailings dams are typically huge constructions

1Luleå University of Technology. e-mail: [email protected] 2Luleå University of Technology. e-mail: [email protected] 3Luleå University of Technology. e-mail: [email protected] 4Luleå University of Technology. e-mail: [email protected]

116 Juan Manuel Rodriguez et al.

and can cover a surface of a couple of square kilometers and be raised to hundred meters high [1]. Storage facilities for tailings need to be safe, structurally and environmentally, in long time perspective. Tailing are the product of mechanical fracturing (crushing and milling) and site specific, as a result they are in general considered to be angular in the range from silt to fine sand [2,3]. Garga et al. [4]classifies the shape of tailings in the range from angular to sub angular. The most economical method to upraise tailing dams is the so called upstream method. The upstream method consists in the use of the coarse part of the tailings to rise up dykes. The discharge of tailings slurry (spigots or cyclones) develops a dike and a beach composed of coarse material (not well graded). Fines and slimes are drag by the flowing water away from the dyke and beach (into the pool). The beach becomes the foundation for the next dyke [5]. According to Harding [6] the degree of which particle breakage occurs during load and deformation affects the strength and stress-strain behavior of the elements. Tailing particles under tenths of meters of material (after a couple of dykes) subject to loads and deformations could break and change strength and stress-strain behavior. Crushing of the particles under stress (under meters of soil e.g. inside the initials tailing dam dykes) depend on factors as particle size distribution, particle shape, void ratio, particle hardness (of all its components), water presence, state of the effective stress, effective stress path [6] and number of contacts per particle [7]. Breakage of the tailing particles into dykes could head for new configuration in the material, e.g. size distribution [8] or particle shape that would change properties as void ratio [9], shear strength [10], permeability (clogging particle skeleton), etc. The scope of the study is to determine the degradation effects of tailings subject to one-dimensional load using classical oedometer test, sieving and, shape changes using image analysis software. It is accomplished by comparing results from four particle size ranges. Eleven shape descriptors are included. Results show that for all specimens except the smallest particle size, particles breakage produces a relative high amount of fines with no change of shape. Additionally for the smallest particle size sample low breakage and shape change is recognized. By comparing natural geological materials with tailings the settlements seems to be determined by the initial void ratio higher in coarse materials but also in tailings. 2 Methodology

Effect of vertical load on tailings particles 117

In this study samples from the Aitik mine has been used. The Aitik tailing dam is located about 100 km North of the Arctic Circle in the boreal parts of Northern Sweden (Figure 1) about 15 km from the community of Gällivare. The value mineral is chalcopyrite (CuFeS2). Main sulphides are pyrite, chalcopyrite and sphalerite. Main gangue minerals are quartz, feldspar, plagioclase and mica [11].

Figure 1: Circle shows the location of the Aitik tailing dam in Sweden (left) and the sampling place in the dam (right). (Google map, 2014).

Disturbed samples were taken from a depth of 0 to 0.7 meters in a trial pit. The trial pit was located at coordinates 67°04’34”N and 20°52’39”W (Figure 1, right) in the north east part of the dam. The sample from the trial pit was split into four different test specimens based on the size range by sievingTable 1). Wet sieving was used with Sodium Diphosphate decahydratate (Na4P2O7·H2O) as a dispersant to enhance the particle separation. After sieving specimens were dried for 24 hours at 105°C. Remolded samples in 50mm diameter and 170mm length sampling tubes were casted by using the methodology describes by Dorby [12]. Dorby’s procedure includes the filling of the tube specimen by steps, usually 5 to 6 in total, where each step comprises 2-3cm of the tube height. Water is added until the step is reached followed by the addition of the tailings sample and


posterior self-settlement for at least 6 hours. Same procedure is followed for every step until the tube is filled up. This methodology imitates the natural sedimentation process leading the tailings settle in beds with natural segregation. Since the test specimens are uniformly graded the segregation should be based on the grain density differences and not in the particle size. Basic geotechnical properties were determined for each sampling tube such as particle density, bulk density; saturated density, void ratio and degree of saturation (see Table 3).

Table 1: Test specimens and size ranges. Specimen Range (mm)

A 1-0.5 B 0.5-0.25 C 0.25-0.125 D 0.125-0.063

In order to study breakage due to increased vertical load the standard oedometer test (ASTM D 2435 -Method A) doubling the weight every 24 hours starting with 10, 20, 40, 80, 160, 320 and 640kPa load step under saturated and drained conditions was performed. The oedometer sample dimensions are 20mm high and 40mm in diameter obtained from the remolded sampling tubes. Upper and lower parts of the odometer were cover with end filters. The test specimens were mounted into the oedometer test rigged and submerged into water. The settlements were monitored by LVDT and the effect on the tailings particles was analyzed after the final load step.The generated amount of fines was determined by loss of mass after re-sieving the material after testing using the sieve for each test specimen respectively. Mass lost was used as measure on decomposition. An initial-state sample from each sample tube, A, B, C and D was collected and analyzed separately as basis for effects on shape change. The effects on the individual tailings particles were studied by two-dimensional image acquisition followed by image analysis where changes in shape properties were studied. As preparation for image acquisition the specimens were dried for 24 hours at 105° Celsius. The image acquisition was performed through a microscope(Motic B1)lightening sources from below and from the side of sample. Themagnification lenses 4x was used for sample A and B and 10x for samples C and D. The magnification used for each sample was chosen based on the results obtained by Rodriguez et al. [13] where it was concluded that the minimum amount of pixels


119

that a particle should comprise is around 1800 pixels to minimize analysis errors. The camera mounted on top of the microscope (Infinity 2) has 2 megapixel resolutions. A more detailed description of the image acquisition process is described in Rodriguez et al. [13].

Table 2: Quantities use to determine the particle shape. Quantity Description Reference

1 4πArea/Perimeter2 [14] 2 4Area/πMajor axis2 [15] 3 Area/Convex Area [16] 4 Fractal dimension [17] 5 Square root of Maximum inscribed/Minimum circumscribed, circle diameters [18] 6 Diameter of a circle same area as particle/Minimum circumscribed circle diameter [19] 7 Perimeter2/Area * [20] 8 Perimeter of a circle with same area/Perimeter [19] 9 Area/Area of the minimum circumscribed circle [21] 10 Perimeter/Convex perimeter * [22] 11 Perimeter/πAverage Feret* [23]

*Inverse was used to obtain a working range between 0 and 1

The 11 quantities in the Table 2were used to determine the shape of the particles; graphical descriptions of the quantities are presented in the appendix. The two dimensional image analyses to determine the shape quantity was done by two different software; Image Pro Plus [17] and ImageJ [15]. Additionally as a control natural geological sand was tested in similar conditions to compare only the breakage and the shape change. This natural sand was obtained from a local sand extraction located 1 km north-west of the Luleå airport belonging to BDX industry AB (Sweden). 3 Results Table 3 summarizes the basic properties of the tailings specimens before and after the oedometer test. Similar test were performed to natural sands, the basic properties for sand are in Table 4.

Table 3: Basic tailing properties. Specimens

tailings Particle

density (g/cm3)

Void ratio (e)

Δe Porosity n

(%)

Δn (%)

Dry Density,

ρ (gr/cm3)

Δ ρ

A Initial

2.881 1.070

0.227 51.7

6.0 1.392

0.171 Final 0.843 45.7 1.563

B Initial 2.904 0.849 0.141 45.9 4.4 1.571 0.129


Final 0.708 41.5 1.700

C Initial

2.873 0.762

0.093 43.3

3.2 1.630

0.092 Final 0.669 40.1 1.722

D Initial

2.943 0.847

0.113 45.9

3.6 1.589

0.108 Final 0.734 42.3 1.697

A number of 170 particles from each test specimens, before and after the oedometer test were analyzed for each shape quantity in Table 2. Basic statistical measures such as the mean, standard deviation and distribution curves were determined. A statistical t-test on 5 % significance level was tested for each shape quantity before and after the oedometer-test for each test specimen.

Table 4: Basic sand properties. Specimens

sand Particle

density (g/cm3)

Void ratio (e)

Δe Porosity n

(%)

Δn (%)

Dry Density,

ρ (gr/cm3)

Δ ρ

A Initial

2.656 0.700

0,148 41.2

5.7 1.563

0.149 Final 0.552 35.5 1.712

B Initial

2.651 0.740

0,118 42.5

4.2 1.524

0.111 Final 0.622 38.3 1.635

C Initial

2.673 0.751

0,096 42.9

3.3 1.527

0.088 Final 0.655 39.6 1.615

D Initial

2.684 0.954

0,156 48.8

4.4 1.374

0.119 Final 0.798 44.4 1.493

The distribution curves for quantities present a slightly skewed distribution and in approximately half of the quantities the distribution does not comply with the normality test for normal distributions. Normal distribution was achieved using Johnson transformation [24] but no differences in the results were observed thus, it was decided to continue without transform the data. In Table 5the results are summarized. Bold-gray highlighted paired numbers represent those values where the mean of both populations (before and after test) show that they are different. Control specimens (natural sand) have shown no shape change (see Table 6). All changes in shape indicate that particles become more rounded except in Table 6 where “ ↓ “ indicates that particles became more angular.


121

Table 5: Mean quantity values for tailing samples. Highlighted marked results show statistically significant changes in the shape quantities.

The degradation of the specimen was measured by quantifying the presence of finer material after oedometer-test. The amount of generated fines (in weight), determined by sieving, is presented in Table 7. Fines generated in tailings are higher in sample C but lower in A.

Table 6: Mean quantity values for the control sample, Natural sand

Quantit

y 1 2 3 4 5 6 7 8 9 10 11

Sample

A Before

0.665

0.753

0.9303

1.0339

0.6451

0.8011

0.1534

0.8458

0.6449

0.9197

0.2930

After 0.66

0 0.74

2 0.926

9 1.033

9 0.636

8 0.791

8 0.163

1 0.848

4 0.630

2 0.929

0 0.296

0

B Before 0.69

0 0.73

5 0.927

5 1.041

4 0.637

5 0.794

5 0.084

0 0.862

6 0.636

1 0.940

5 0.299

8

After 0.712

0.746

0.9315

1.0390

0.6515

0.8049

0.0815

0.8804

0.6518

0.9530

0.3039

C Before

0.702

0.713

0.9406

1.0292

0.6292

0.7853

0.0433

0.8795

0.6224

0.9599

0.3060

After 0.69

5 0.74

4 0.936

3 1.028

6 0.647

8 0.803

5 0.047

4 0.877

0 0.649

2 0.951

3 0.303

2

D Before 0.69

4 0.69

3 0.932

9 1.049

2 0.609

8 0.775

9 0.025

1 0.872

9 0.608

3 0.960

3 0.306

5

After 0.722

0.738

0.9413

1.0439

0.6460

0.7973

0.0266

0.8919

0.6395

0.9638

0.3076

Quantity 1 2 3 4 5 6 7 8 9 10 11

Sample

A Before 0.730 0.705 0.952 1.022 0.637 0.789 0.135 0.890 0.628 0.966 0.307

After 0.735 0.732 0.954 1.016 0.657 0.801 0.186 0.895 0.645 0.966 0.307

B Before 0.736 0.723 0.948 1.034 0.643 0.794 0.081 0.895 0.634 0.969 0.309

After 0.743 0.735 0.951 1.030 0.651 0.800 0.085 0.898 0.644 0.968 0.309

C Before 0.742 0.731 0.954 1.023 0.653 0.815 0.047 0.031 0.698 0.970

↓ 0.309

↓

After 0.734 0.736 0.953 1.022 0.656 0.804 0.051 0.033 0.650 0.967 0.308

D Before 0.744 0.717 0.951 1.040 0.643 0.793 0.027 0.903 0.633 0.974 0.311

After 0.736 0.701 0.949 1.039 0.632 0.791 0.028 0.899 0.630 0.972 0.310


Table 7: The amount fines generated in the test specimens after the oedometer test Specimen tailings

% fine content Control specimen

natural sand % fine content

A 14 A 19 B 10 B 11 C 12 C 4 D <1 D 2

Figure 2: Semi-logarithmic stress-strain curve (left) and stress-strain bar by load step (right) from the oedometer test for the four tailings specimens.

Figure 2 (left) shows the stress-strain behavior where specimens A and B are weaker. Figure 2 (right) show that all specimens have initially high deformation, probably a result of initial rearrangement of the loose structures and re-distribution of stresses in the test specimens. The step strain for the test specimens A, B, and C differs from the specimen D behavior (Figure 2, right) because specimen D shows a high initial deformation and no significant difference in deformation between the two last load steps. In s Figure 3 tress-strain curves of the tailings and control sand are plotted together. Dark lines represent the tailings while light lines are the control sands. It is interesting to see how the crossing of the lines of each size (sand-tailings) moves to the right for samples A, B and C but not for D. should D sample cross in higher stresses?. For coarse samples A and B there are more settlements for


123

tailings compare with sands.

Figure 3:Stress-strain cumulative curve for sand and tailing samples. Dark lines represent tailings while light lines are the sands

4 Discussion A disturbed sample of tailings was split by sieving into four, uniformly graded, test specimens and the effect on the uniformly graded test specimens upon vertical loading in an oedometer-test was compared. The properties of the four tailings specimens are different. The coarsest tailing material sample 0.5mm, (A), is arranged in a looser state during the test specimen preparation and also showed the weakest stress-strain response in both, cumulative and step load (Figure 2, right). The analysis of the remaining material after the oedometer test showed that this test specimen also showed the highest degree of degradation measured as the generation of finer particles compared to the initial distribution (Table 7). The stress-strain behavior of the tailings specimens 0.125 (C) and 0.063 mm (D) are stiffer compared to the coarser 0.5 (A) and 0.25mm (B)


(Figure 2, left). Considering the degradation of the particles there is a big difference between specimen size 0.063mm and the others. The range of degradation is similar in specimen 0.5, 0.25 and 0.125 but is by methodology used in this study very low in specimen 0.063 mm. As Lade et al. [25] explained breakage increases while particle size increases due to the fact that larger particles contain more defects (or the probability to have it is higher) and depend also in the number of contacts per particle [7] in agreement whit Table 7 results. Considering the shape factor tailings specimen D shows a statistical significant change in 8 of 11 quantities, considerable more than the rest (Table 5).In the specimen D (the finest fraction 0.063mm) shape change was observed in the majority of the shape quantities. This sample was also the most resistant to degradation, less than 1% of finer material was generated. According with the results it can be suggest that tailings sample D, due to the low % breakage (Table 7) and the quantity value change (Table 5) before and after the oedometer tests, is breaking only in the angularities or corners. Quantities 10 and 11 can also suggest this due to quantity 10 look for the concave outline change and since there is no difference it means that only corners are breaking while quantity 11 follows the same logic but in measuring two parallel line distances (feret box). In the finest fraction also a peculiar step load behavior, compared with the rest of the specimens, can be seen (Figure 2, right). Even if all the samples show an initial high compressibility the finest test sample D(0.063mm) seems to not bee according with the rest of the stepped behavior of the samples A, B and C (major to minor in the initial load). Furthermore the two last load steps show no increase of percentage strain while the rest of the samples do. It shows a high initial compressibility but low at final loads. By comparing the control specimens (natural sands) and tailing specimens it was observed that there was no general shape change in the control material maybe, because the particles already had been rounded by the time and natural elements through the years and they have lose their angularities. Even if there is a percentage in breakage in the control material there is no statistical evidence that the shape in average have change, maybe the new angular particles when generated due the breakage are no longer belonging to the specific size range and they were lost through the sieving (change of size). Stress-strain comparison of sands and tailings (Figure 3) suggest bigger settlements in tailings comparing by size for coarse samples A and B. it is interesting to see how tailings-sand lines (by size) start crossing to the right (for samples A, B and probably C) because the higher void ratios in tailings.


125

Breakage tailings-sands comparison by size in samples A and C seem to be different; for Sample A for sand and C for tailings looks higher. The high breakage of sand sample A could be due to angularities in the tailings brake (generating new angularities) and fines materials generate a stronger skeleton while sand break through the body reducing the size of the particles. Void ratio reduction also support this explanation been almost the double for tailings compared with sands. Sample C is believed that not only angularities break but the body as well explaining the higher fines generation. In this case the void ratio change is similar showing that even with the angularities tailing sample was relatively in the same initial packing state as sand sample was breaking the corners first and the body later. This study shows that the degradation driven by increased vertical stress of tailings particles is different for different grain sizes under the same vertical loads. Coarser tailings are more susceptible compared to finer tailings. Coarse tailings are usually considered from a dam construction perspective to be more favorable since consolidation settlements are more rapid and hydraulic conductivity is higher. Cyclone techniques are sometimes applied to increase the coarser fraction at the beaches as a preoperational step for raising the tailings dams. The results from this study also indicate that there is a higher settlement as a result of particle degradation compared to the finer fractions. 5 Conclusions Only fraction 0.063mm tailing particles show a shape change due to the corners becomes more rounded. Coarse particles fraction 0.5mm present more breakage than the rest of the samples. Coarse sample fractions 0.5 and 0.25mm produce higher settlements compare with the finer fractions and also compared with sands (by size). Initial porosity seems to be the major factor affecting settlements. References [1] Ormann, Linda.; Zardai, Muhammad Auchar.; Mattson, Hans.; Bjelkevik, Annika and Knutsson, Sven. Numerical analysis of strengthening by rockfill embankments on an upstream tailings dam. Canadian Geotechnical Journal,50,(2013),pp. 391-399.


[2] FHA. User Guidelines for Waste and Byproduct Materials in Pavement Construction. Department of Transportation, Federal Highway Administration, Washington D.C. USA, (1997). [3] Rodriguez J. and Edeskär T. Case of Study on particle Shape and Friction Angle on Tailings. Journal of Advanced Science and Engineering Research,3(4),(2013), pp. 373-387. [4] Garga, Vinod. K.; ASCE, M. and McKay, Larry. D. Cyclic Triaxial Strength of Mine Tailings. Journal of Geotechnical Engineering, 110(8), (1984) pp.1091-1105. [5] Vick, S.G. Planning, design and analysis of tailing dams. Richmond, BC: BiTech Publishers Ltd.(1990). [6] Harding, Bobby. O. Crushing of soil particles. Journal of Geotechnical Engineering, 111(10), (1985), pp. 1177-1192. [7] Valdes, Julio. R. and Koprulu, Eren. Characterization of fines produced by sand crushing. Journal of Geotechnical and Geoenvironmental Engineering, 113(12),(2007), pp.1626-1630. [8] Fukumoto, Takeaki. Particle breakage characteristics of granular soil. Soil and foundations, 32(1), (1992), pp.26-40. [9] Holubec and D’Apolonia. Effect of particle shape on the engineering properties of granular soils. ASTM STP, 523, (1973), pp. 304-318. [10] Cho, G.; Dodds, J. and Santamarina, J.C. Particle shape effects on packing density, stiffness and strength: Natural and crushed sands. Journal of Geotechnical and Geoenvironmental Engineering, 132(5), (2006), pp.591-602. [11] Lindvall, M. and Eriksson, N. Investigation of the weathering properties of tailings sand from Boliden´s Aitik copper mine – a summary of twelve years of investigations. Proceedings 6th International Conference on Acid Rock Drainage.Cairns, (2003). [12] Dorby, R. Soil properties and earthquake ground response. Proceedings of the 10th European Conference on Soil Mechanics and Foundation Engineering,4,(1991), Florence, Italy. [13] Rodriguez, J. M.; Johansson, J. M. A. and Edeskär, T. Particle shape determination by two-dimensional image analysis in geotechnical engineering. Danish Geotechnical society bulletin, 27,(2012), pp. 207-218. [14] Cox, E. P. A method of assigning numerical and percentage values to the degree of roundness of sand grains. Journal of Paleontology, 1(3), (1927), pp.179-183.


127

[15] Ferreira and Rasband. ImageJ user guide,(2012), http://imagej.nih.gov/ij/ [16] Mora, C. F. and Kwan, A. K. H. Sphericity, shape factor, and convexity measurement of coarse aggregate for concrete using digital image processing. Cement and Concrete Research, 30, (3), (2000), pp.351-358. [17] Image Pro Plus v. 7.0. (2011). http://www.mediacy.com/. [18] Riley N. Allen. Projection Sphericity. Journal of Sedimentary Petrology, 11, (2),(1941), pp. 94-97. [19] Wadell, H. Volume, shape, and roundness of quartz particles. Journal of Geology, 43, (1935), pp.250-279. [20] Blott, S.J. and Pye, K. Particle shape: a review and new methods of characterization and classification. Sedimentology, 55, (2008), pp.31-63. [21] Tickell, F. G. Effect of the angularity of grain on porosity and permeability. Bulletin of the American Association of Petroleum Geologist, 22, (1938), pp.1272-1274. [22] Janoo, Vincent C. Quantification of shape, angularity, and surface texture of base course materials. Special report 98-1,US Army Corps of Engineers, Cold Region Research and Engineering Laboratory, (1998). [23] Kuo, C.-Y., Rollings, R., & Lynch, L. N. Morphological study of coarse aggregates using image analysis. Journal of Materials in Civil Engineering, vol. 10, (3), (1998), pp.135-142. [24] Johnson, N. L. System of frequency curves generated by methods of translation. Biometrika, 36(1/2), (1949), pp.149-176. [25] Lade, Poul. V.; Yamamuro, Jerry. A. and Bopp, Paul. A. Significance of particle crushing in granular materials. Journal of Geotechnical Engineering, 122(4), (1996), pp. 309-316. [26] Johansson, Jens and Vall, Jakob. Friktionsjords kornform: Inverkan på geotekniska egenskaper, beskrivande storheter och bestämningsmetoder. Examensarbete. Luleå tekniska universitet, (2011), (in Swedish).


APPENDIX

Quantity Description Graphic description

1* 4πArea(A)/Perimeter2(P)

2* 4Area(A)/πMajor axis2(Major)

3* Area(A)/Convex Area(Ca)

4 Fractal dimension

Fractal dimension use 'strides' (minimum step lengths) of various sizes. The

fractal dimension is calculated as 1 minus the slope of the regression line

obtained when plotting the log of the perimeter (for various strides) against the

log of the stride length. (more info in imageproplus)

5* Square root of Maximum inscribed (Di)/Minimum circumscribed(Dc), circle

diameters


129

6* Diameter of a circle same area as particle(Da)/Minimum circumscribed circle

diameter(Dc)

7 Perimeter2(P)/Area(A) See figure in quantity 1 in this table

8* Perimeter of a circle with same area (Pa)/Perimeter(P)

9* Area(A)/Area of the minimum circumscribed circle (Ac)

10* Perimeter/Convex perimeter

11 πAverage Feret/Perimeter(P)

Average ferret box is obtained rotating two parallel lines (two degrees each

time) and measuring the distance, finally the average ferret is the average

distance of all the feret boxes distance measured

* Figures were taken and modified from Johansson and Vall [26]

Paper V Rodriguez, J.M.; Edeskär, T. and Knutsson, S. (2016) Mechanical weathering effect on tailings particles. Proceedings of the 17th Nordic Geotechnical Meeting. Reykjavik, Iceland. 25th -28th May 2016. ISBN 978-9935-24-002-6.

Investigation, testing and monitoring - Mechanical weathering effect on tailings

Mechanical weathering effect on tailings Juan M. Rodriguez Luleå University of Technology, Sweden, [email protected] University of Sonora, Mexico, [email protected] Tommy Edeskär Luleå University of Technology, Sweden Sven Knutsson Luleå University of Technology, Sweden ABSTRACT Over the last century the tailing volume generation has grown dramatically due to the mineral demand. Nowadays the mining industry is producing every year millions of tons of tailings. The storage of the tailings has become a challenge due to the increased storage capacity demanded. Physical risk associated to the tailings dams is the stability itself since tailing dams are considered a walk-away solution. Physical changes as breakage and shape occur to the tailing particles affecting the stability of the fills by reduced strength properties. In order to understand the reduction and shape changes of tailing particles degradation test by milling attrition (erosion) and image analysis was conducted. Uniform fractions 1-0.5, 0.5-025, 0.25-0.125 and 0.125-0.063mm were used. Results have shown that attrition agents e.g. ball attrition can increase the physical erosion but also change the shape of the particles compared with autogenous attrition. However particles shape has become more regular (less elongated) and rounded in coarse fractions 1-0.5 and 0.5-0.25mm while smaller fractions 0.25-0.125 and 0.125-0.063mm seems to have opposite behavior. Comparison with previous milling studies show consistent differences probably due to the breakage of the particles was the objective. In perspective if tailings become more rounded the strength could be compromised. More studies are needed to verify this. Keywords: Particle shape, particle size, attrition, tailings.

1 INTRODUCTION

Tailings are the mining industry leftovers, for its storage tailing dams are constructed. In the history of the mineral extraction tailings dam incidents has occurred e.g. Val di Stava tailings dam (1985) in Italy and Cananea tailings dam (2014) in Mexico. As a consequence of the operation and raising procedures of tailings dams, the conditions in the tailings dams could be considered to be dynamic in a longer perspective. Grain size distribution, formation of layers, pore pressures and stress states are continuously changing during the operation (Ormann, et al., 2013).

In general particle degradation occurs due the environmental factors as chemical reactions (Kossoff, et al., 2012) and mechanical factors as stresses generated due the load and creep (Valdes & Koprulu, 2007). Tailings are also affected in the same way. Mechanical behaviors of granular materials are affected by characteristics as particle size (Islam, et al., 2011), particle shape (Santamarina & Cho, 2004), size distribution (Cabalar, 2011), mineralogy (Clayton, et al., 2004) among others. Erosion changes all physical characteristics thus, its strength and behavior changes through the time. Tailing dams need to be built with environmental and structural safety. They must stand for long periods in a long time perspective, even after mining reclamation. To ensure the structural safety not only actual

1


2

soil characteristics must be taken in consideration but also further changes. Laboratory mill attrition by uniform sized particles was performed in this study to determine the size and shape changes. Shape according to Cho, et al. (2006) influence the strength of the materials and, from this point of view if particles become more rounded with the pass of time due to mechanical wearing the stability could be compromised. Results have shown that ball attrition is more effective to erode the tailing particles, ball attrition produce more size reduction (as it was expected) and more shape changes. Results in this study also indicate that coarse fraction 0.5 and 0.25mm become more rounded but smaller sizes are more angular. The effect of the size reduction as the contradictory shape changes effects should be studied in in detail in further research. The statistical methods Mann-Withney test and two sample t-test used to determine shape differences have 96% of agreement, furthermore there is no skew evidence over the results on any of the methods. Thus, it is possible to use two samples t-test for the study of the tailings.

2 THE PARTICLE SHAPE

In this study the word shape is used to describe a grain’s overall geometry. Furthermore, in order to describe the particle shape in more detail, there are a number of terms, quantities and definitions used in the literature. Some authors (Mitchell and Soga, 2005 and Arasan, et al., 2010) are using three sub-quantities describing the shape but at different scales. The sub-quantities are morphology/form, roundness and surface texture. In Figure 1 it is shown how the scale terms are defined. At large scale a particle’s diameters in different directions are considered. At this scale, describing terms as spherical, circular, platy, elongated etc., are used. An often seen quantity for shape description at large scale is sphericity. Graphically the considered type of shape is marked with the dashed line in Figure 1. At intermediate scale is focused on description of the presence of irregularities. Depending on at what scale an analysis is

done; corners and edges of different sizes are identified. By doing analysis inside circles defined along the particle’s boundary, deviations are found and valuated. The mentioned circles are shown in Figure 1; a generally accepted quantity for this scale is roundness or the antonym: angularity. Regarding the smallest scale, terms like rough or smooth are used. The descriptor is considering the same kind of analysis as the one described above, but is applied within smaller circles, i.e. at a smaller scale. Surface texture is often used to name the actual quantity.

Figure 1 Particle describing the shape scale attributes (Mitchell and Soga, 2005).

3 METHODOLOGY

In this study samples from the Aitik mine has been used. The Aitik tailing dam is located about 100 km North of the Arctic Circle in the boreal parts of Northern Sweden about 15 km from the community of Gällivare. The value mineral is chalcopyrite (CuFeS2). Main sulphides are pyrite, chalcopyrite and sphalerite. Main gangue minerals are quartz, feldspar, plagioclase and mica (Lindvall, M. and Eriksson, N., 2003). Four range sizes were used in the actual research (see Table 1). Further on the lower size in the range will be used for convenience. Wet sieving was used with Sodium Diphosphate decahydratate (Na4P2O7·H2O) as a dispersant to enhance the particle separation. After sieving specimens were dried for 24 hours at 105°C. Smooth steel drum mill with 115mm in diameter and 132mm in length over a rolling table was used to generate a gentle and a constant flow of tailings material over a constantly rebuilding slope. Speed of 60 rpm


(revolutions per minute) was used to ensure a gentle rolling of the particles down the slope and procure the attrition/wearing/abrasion of particles. Speeds close to the terminal velocity were avoided due to it could breakage the particles. Table 1 Test specimens and particle size ranges

Specimen Range (mm) 0.5x 1-0.5 0.25x 0.5-0.25

0.125x 0.25-0.125 0.063x 0.125-0.063

The tests were configured in two different forms the first was using balls to speed up the attrition of the tailing particles and the second was autogenous attrition (no balls). The subscript “x” in Table 1 represents the autogenous attrition (absence of balls) for “a” and the use of balls for “b”. A total of 200 gr of tailings were used during each batch with 1000 ml of water. For ball attrition 100 grams of 7 millimeters in diameter steel balls were included. Degradation was conducted for 2 and/or 3 time periods; for balls attrition time periods was approximately of 2, 6 and 24 hrs., and for autogenous attrition time period consisted

of 24, 72 and 120 hrs. Time periods were set after some trial tests to identify the amount of material broken and to avoid running out of material specially when using steel balls. Material broken (in percentage weight) was identified sieving the sample after the test. Total amount of tests and more detailed data can be seen in Table 3. Particle shape was measured using eleven shape descriptors or quantities (Table 2) through two dimensional image analysis. Graphical description of the quantities can be found in the appendix. Shape change was statistically identified using Mann-Whitney Test (Moses, 2014) and Two-Sample t-test (Snedecor & Cochran, 1989) with 5% significance level. Since the data distribution is not all the time normal it was decided to use the two mentioned statistical tests, the non-parametric test Mann-Whitney test applies for unknown and skew distributions while two samples t-test determine differences when the data is normally distributed. Sub-quantities were classified according with the authors (Rodriguez, 2012); form is recognized by the “#” and roundness by “+”symbols in Table 2.

Table 2 Quantities use to determine the particle shape

Quantity Description Workingrange

Reference

1 2 0-1 (Cox, 1927) 2 2 0-1 3 0-1 4 1-2 5

0-1

6

0-1

7 2 0- 0-1

9

0-1

10 0-1 ) 11 0-1

”

3


4 RESULTS

Table 3 is showing the result of each individual test arranged by sample, milling time and percentage of fines produced. Fine production percentage was measured in relation to the initial sample weight and the original sieve size. Figure 2 summarizes the results from the tests. Dashed lines are those tests that contains iron balls as degradation agent while continues lines are samples in autogenous attrition. The use of balls as an attrition material speed up the degradation process of the tailings (size change) and it is evident when comparing the general slope of dashed and continues lines (balls and autogenous grinding). Relative shape change results are shown in Figure 3, gray colored markers and lines shows when most of the shape descriptors or

quantities values have changed. Figure 3 left (ball attrition) illustrate that there is a general shape change during all steps for sizes 0.5 and 0.25mm (except last step for 0.25mm).

Figure 2 Fines generation in mill test Dashed lines are those tests that contains iron balls as degradation agent while continues lines are samples in autogenous attrition.

Figure 3 Degradation and shape change by milling agent. Gray leyends indicate shape change. Left ball milling. Right autogenous milling (Fraction sizes are represented by the lower limit) Table 3 Rolling time periods and percentage of material undergoing the size range (fines generation)

Sample Time period(min)

% fines generated

Sample Time period (min)

% fines generated

0.5 1260 7.6 0.5 100 32.0 13.9 370 47.5 7207 14.7 1140 67.2

0.25 1442 6.7 0.25 100 33.9 1046 69.0

0.125 1344 11.3 0.125 103 4252 343 33.7 1441 90.6

0.063 1470 0.063 100 4222 6.4 324 17.2 1406 62.5

a b

Autogenous milling

Ball milling

4


Smaller sizes 0.125 and 0.063mm only show shape change in the last step. From Figure 3 right (autogenous attrition) only 0.125mm size has shown shape change. In same Figure 3 there is a rapid step increase in the broken percentage in all first step time tests. Particles seem to break more rapidly during the first minutes of the tests (see the initial slope in Figure 3 right and left).

Relative shape changes by time step and quantity are shown in Table 4 represented by

becomes more elongated or angular otherwise they are more regular in form and rounded. Quantities mean values before and after the time step attrition are presented.

Table 4 Relative shape change time-step, data represent the mean value. Shape change is highligthed by a gray shadow when particles become more regular in form or rounded; the appearace of

Balls Autogenous S Q 2hr 6hr 24hr 24hr 72hr 120hr A +1 0.665 0.702 0.702 0.732 0.665 0.673 0.673 0.672 #2 0.742 0.742 0.735 0.735 0.760 0.743 0.743 0.741 0.741 +3 0.924 0.935 0.935 0.943 0.943 0.951 0.924 0.932 0.932 0.932 0.932 0.931 +4 1.041 1.031 1.031 1.025 1.025 1.022 1.041 1.033 1.033 1.032 1.032 1.031 #5 0.796 0.796 0.796 0.796 #6 0.799 0.799 0.792 0.792 0.792 0.792 #7 0.126 0.161 0.161 0.160 0.160 0.160 0.126 0.159 0.159 0.153 0.159 #9 0.620 0.642 0.642 0.641 0.641 0.672 0.620 0.631 0.631 0.631 0.631 0.640 +10 0.929 0.930 0.930 0.946 0.946 0.949 0.929 0.931 0.931 0.933 0.933 +11 0.296 0.296 0.296 0.301 0.301 0.303 0.296 0.297 0.297 0.297 0.297 B +1 0.676 0.705 0.705 0.745 0.745 0.746 0.676 #2 0.710 0.740 0.740 0.771 0.710 0.730 +3 0.922 0.932 0.932 0.949 0.949 0.923 0.924 0.924 +4 1.045 1.041 1.041 1.033 1.033 1.033 1.047 1.042 1.042 1.042 #5 0.779 0.779 0.794 #6 0.790 #7 0.079 0.079 0.079 0.079 0.900 0.900 0.902 #9 0.612 0.640 0.640 0.659 0.659 0.612 0.624 0.624 0.629 +10 0.937 0.947 0.947 0.964 0.964 0.963 0.937 0.940 0.940 0.940 +11 0.299 0.302 0.302 0.307 0.307 0.307 0.299 0.300 0.300 0.300 C +1 0.700 0.703 0.703 0.702 0.702 0.700 0.696 #2 0.706 0.713 0.713 0.710 0.710 0.706 0.719 +3 0.936 0.935 0.935 0.939 0.939 0.936 0.926 0.934 +4 1.034 1.032 1.032 1.031 1.031 1.034 1.032 1.032 1.032 #5 #6 0.777 0.777 #7 0.040 0.040 0.040 0.040 0.039 0.046 0.046 0.044 #9 0.614 0.619 0.619 0.620 0.620 0.614 0.610 0.610 0.625 +10 0.961 0.960 0.960 0.960 0.960 0.961 0.953 0.953 +11 0.307 0.306 0.306 0.306 0.306 0.307 0.304 0.304 D +1 0.700 0.702 0.702 0.713 0.713 0.713 0.713 0.713 #2 0.714 0.714 0.692 0.692 0.693 +3 0.931 0.933 0.933 0.937 0.937 0.931 0.939 0.939 0.936 +4 1.051 1.051 1.051 1.053 1.053 1.051 1.051 1.050 1.050 #5 0.772 0.773 0.773 0.772 0.777 0.777 #6 0.771 0.771 0.772 0.772 #7 0.024 0.023 0.023 0.022 0.024 0.023 0.023 0.024 #9 0.596 0.600 0.600 0.616 0.616 0.596 0.602 0.602 0.609 +10 0.964 0.965 0.965 0.967 0.967 0.964 0.971 0.971 +11 0.309 0.309 0.307 0.310 0.309

5


Table 5 Shape change compared with initial state. Mann-Withney and Two-sample t-tests comparison. Shape change is highligthed by a gray shadow when particles become more regular in form or rounded; the appearace of

Balls Autogenous 2hr 6hr 24hr 24hr 72hr 120hr Size Q MW Tt MW Tt MW Tt MW Tt MW Tt MW Tt A +1

#2

+3

+4

#5

#6

#7

#9

+10

+11

B +1

#2

+3

+4

#5

#6

#7

#9

+10

+11

C +1

#2

+3

+4

#5

#6

#7

#9

+10

+11

D +1

#2

+3

+4

#5

#6

#7

#9

+10

+11

- - Table 4 shows that tailing particles subject to ball attrition in sizes 0.5 and 0.25 mm in general become more rounded and regular in form as the time-step evolves while for particle sizes 0.125 and 0.063 mm practically no change is seen until last time-step where

increase in angularity and irregularity in form is perceived. For autogenous attrition time-step changes are not recognized in the majority of the quantities (as in ball attrition) for coarse sizes 0.5 and 0.25 mm, however fine fraction 0.125

6


and 0.063mm present changes in the first time-step with special remark on size 0.125mm that became more angular and irregular in form. Table 5 is showing the shape change with respect to the initial state of the particles. In this table gray cells identify the shape change with respect to the original shape. Mann-Whitney Test (Moses, 2014) and Two-Sample t-Test (Snedecor & Cochran, 1989) are located together to evaluate differences, practically results agree in both tests.

5 DISCUSSION

Mill attrition test were performed for different fraction sizes (Table 1) during different time periods (Table 3). Attrition degradation was identified by sieving (fines generation) and particles shape change by image analysis. Tailing particles images were also subject to visual inspection and tailings had been classified as very angular to sub angular material base on Powers (1953) comparison chart The results of this classification is in agreement with the conclusion of the general shape of tailings made by e.g. Garga, et al. (1984).Visual inspection of the original material (splitted by size) in the soil containers detects particle breakage; particles are weak enough to be eroded during the handling and sample preparation. This could affect the results since an already broken portion of material could be introduced in the mill. Lade, et al. (1996) suggest that angular particles are very susceptible to break even at low stresses because they concentrate in the angular contact points. Furthermore the particle breakage is also a function of time even under constant stresses (Yamamuro & Lade, 1993). That could explain why sample 0.25mm in figure 3 (left) shows an outstanding breakage at the end point (around 4000 minutes of the running test). It is also shown that particles are broken at higher rate in the initial step. During this study two statistical methods were used, Mann-Withney test and two sample t-test. Mann-Withney test is applicable to skew distribution where the normality test is not recognized. Two

samples t-test is used when the normality distribution of the data is accomplished. Data in the study was detected to be in some cases normal distributed and a comparison of results among these methodologies is of interest (see Table 5). Even if the distribution of the data is not normal results has shown that there is a 96% of coincidence between the two statistical methods. Furthermore the shape change detection is not skew to one of the methods or in a specific quantity. Two samples t-test could be used to determine differences in shape since there is no evidence that the skew distribution of the data affect the results. Quantities Table 2 are identified as a form or roundness descriptor as Mitchell and Soga (2005) sub-quantities classification (according to authors). The classification is not showing any relation with the attrition results, shape changes are seems not to be related with the shape sub-quantities classification (Figure 1). Choose of the best quantity or shape descriptor is of interest and attempts can be found in literature (Bouwman, et al., 2004) Even if there are some standards e.g. ASTM D3398, D4751 they are only there to avoid unfavorable shape (e.g. elongated particles) but still there is no general agreement on which shape descriptor, or descriptors, should be considered universally. In the scientific literature some quantities appear more than others e.g. Wadell (1932) comparison chart (recently computerized by Zheng & Hryciw, 2015), aspect ratio (Hawkins, 1993), circularity (Cox, 1927) among others. Soil strength is shape dependent as Santamarina & Cho (2004) conclude, in perspective the gain in angularity for the smaller fractions (0.125 and 0.063mm) could increase the soil strength but reduce it in the coarser fractions. Any way it is unknown on which size has the major influence thus, more studies should be necessary. Ulusoy (2008) study shows that ball milling produces more angular and irregular particles opposite of what this study concluded. Furthermore his quantity values were always lower indicating that produced particles are more angular and irregular in form compared with this research. There are two main

7


differences that could affect the results; the mineralogy and the milling process itself. Talk mineral was used by Ulusoy (2008); talk has the lower value in the Mohr hardness scale. The Ulusoy (2008) milling intention was to break the particles while this study the attrition was the objective; it is known that the final shape of the particles is considered to be the result of several factors among them the rigor of the transport (Wentworth, 1922).

6 CONCLUSIONS

Breakage occurs in all states even in sample deposits (bags, trays) Particles with ball attrition eventually change their initial shape to be more rounded/circular in sizes 0.5 and 0.25mm but fine particles 0.125 and 0.063mm come back to be angular/irregular. Attrition is more intensive when erosional agents as iron balls are included. It is possible to use any of the statistical methods Two samples t-test or Man-Withney since there is practically no difference in the results. There is no skewed shape change difference in between the sub-quantities morphology and roundness.

7 REFERENCES

Arasan, S., Hasiloglu, A. S. & Akbulut, S. (2010). Shape particle of natural and crushed aggregate using image analysis. International Journal of Civil and Structural Engineering, 1(2), pp. 221-233.

Blott, S. & Pye, K. (2008) Particle shape: a review and new methods of characterization and classification. Sedimentology 55, 31-63.

Bouwman, A. M., Bosma, Jaap C.; Vonk, Pieter, Wesselingh, J (Hans) A. & Frijlink, Henderik W. (2004). Which shape factor(s) best describe granules?. Powder Technology, Volume 146, pp. 66-72.

Cabalar, A. F. (2011). The effects of fines on the behaviour of a sand mixture. Geotech. Geolog. Eng. 29, 91-100.

Cho, G., Dodds, J. & Santamarina, J., (2006). Particle shape effects on packing density, stiffness and strength: Natural and crushed sands. Journal of geotechnical and geoenvironmental engineering, 132(5), pp. 591-602.

Clayton, C., Theron, M. & Vermeulen, N. (2004). The effect of particle shape on the bahaviour of gold tailings. London, Advances in Geotechnical Engineering: The Skepton Conference, Thomas Telford.

Cox, E. P. (1927). A method of assigning numerical and percentage values to the degree of roundness of sand grains. 1(3), pp. 179-183.

Garga, V., ASCE, M. & McKay, L. (1984). Cyclic triaxial strength of tailings. Journal of Geotechnical Engineering, 110(8), pp. 1091-1105.

Hawkins, A. (1993). The shape of powder-particle outlines. New York: Wiley.

Image Pro Plus v. 7.0, 2011. http://www.mediacy.com/. [Online].

ImageJ, 1., 2013. version 1.47v. Wayn Rasband, National Institutes of Health, USA.. [Online] Available at: http://imagej.nih.gov/ij/

Islam, M., Hossain, M., Rahman, A. & Asad, M. (2011). Effect of particle size on the shear strength behaviour of sands. Australian Geomechanics, 46(3), pp. 75-85.

Janoo, V. C. (1998). Quantification of shape, angularity, and surface texture of base course materials. Cold Region Research and Engineering Laboratory. US Army Corps of Engineers. Special report 98-1.

Kossoff, D., Hudson-Eduards, K.A.; Dubbin, W.E., Alfredsson, M. & Geraki, T. (2012). Cycling of As, P, Pb and Sb during weathering of mine tailings: implications for fluvial environments. Mineralogical Magazine, 76(5), pp. 1209-1228.

Lade, P.V., Yamamuro, J.A. & Bopp, P.A. (1996). Significance of particle crushing in granular materials. Journal of Geotechnical Engineering, 122(4), pp. 309-316.

Lindvall, M. and Eriksson, N. (2003). Investigation of the weathering properties of tailings sand from Boliden´s Aitik copper mine – a summary of twelve years of investigations. Proceedings 6th International Conference on Acid Rock Drainage.Cairns.

Mitchell and Soga (2005). Fundamentals of soil behavior. 3rd ed. Wiley.

Mora, C.F. & Kwan, A.K.H. (2000). Sphericity, shape factor, and convexity measurement of coarse aggregate for concrete using digital image processing. Cement and Concrete Research, 30(3), pp. 351-358. Moses, L. E. (2014). Wilcoxon-Mann-Whitney test: Definition and example. John Wiley and Sons.

Ormann, L., Zardari, M., Mattsson, H. & Knutsson, S. (2013). Numerical Analysis of strengthening by rockfill embankments on an upstream tailings dam. Canadian Geotechnical Journal, Volume 50, pp. 391-399.

Powers, M. (1953). A new roundness scale for sedimentary particles. Journal of Sedimentary Petrology, 23(2), pp. 117-119.

Riley, N. A. (1941). Projection Sphericity. Journal of Sedimentary Petrology, 11(2), pp. 94-97.

Rodriguez, J. M. (2012). Particle shape quantities and influence on geotechnical properties: a review, Lulea, Sweden: Lulea University of Technology.

Santamarina, J. & Cho, G. (2004). Soil behaviour: The role of particle shape. London, Skempton Conf. London.

8


Snedecor, G.W. & Cochran, W.G. (1989). Statistical methods. 8th ed. Iowa: State University Press.

Tickell, F.G. (1938). Effect of the angularity of grain on porosity and permeability. bulletin of the American Association of Petroleum Geologist, Volume 22, pp. 1272-1274.

Ulusoy, U. (2008). Application of ANOVA to image analysis results of talc particles produced by different milling. Powder Technology, 188(2), pp. 133-138.

Valdes, J.R. & Koprulu, E. (2007). Characterization of fines produced by sand crushing. Journal of Geotechnical and Geoenvironmental Engineering, 113(12), 1626-1630.

Wadell, H. (1932). Volume, shape and roundness of rock particles. Journal of Geology, Volume 40, pp. 443-451.

Wadell, H. (1935). Volume, shape, and roundness of quartz particles. Journal of Geology, 43, 250-279.

Wentworth, W. (1922). A method of measuring and plotting the shape of pebbles. U.S. Geological Survey Bulletin, Volume 730C, pp. 91-114.

Yamamuro, J. & Lade, P. (1993). Effects of strain rate on instability of granular soils. Geotechnical Testing Journal, 16(3), pp. 304-313.

Zheng, J. & Hryciw, R. (2015). Ttraditional soil particle sphericity, roundness and surface roughness by computational geometry. Geotechnique, 65(6), pp. 494-506.

APENDIX

Q Description Graphic description 2

2

4

9


.

7 2

11

10

Paper VI Rodriguez, J.M.; Bhanbhro, R.; Edeskär, T. and Knutsson, S. Shear strength in uniformed sized tailings particles. To be submitted to Journal of Earth Sciences

Shear strength in uniformed sized tailing particles Juan Manuel Rodriguez, Ph.D. Student Department of Civil, Environmental and Natural Resources Engineering, Luleå University of Technology, Luleå, 97187, Sweden. Email: [email protected]. Tel:+46 703406247 Civil and Mining Depatment., University of Sonora, 83000, Hermosillo, Mexico. Email: [email protected]. Tel:+52 662 259 21 83 and 84 Riaz Bhanbhro, Ph.D Student Department of Civil, Environmental and Natural Resources Engineering, Luleå University of Technology, Luleå, 97187, Sweden. Email: [email protected] Tommy Edeskär, Senior Lecturer Department of Civil, Environmental and Natural Resources Engineering, Luleå University of Technology, Luleå, 97187, Sweden. Email: [email protected] Sven Knutsson, Professor Department of Civil, Environmental and Natural Resources Engineering, Luleå University of Technology, Luleå, 97187, Sweden. Email: [email protected] Abstract Mining industry provides mineral to the modern society. Minerals are indispensable raw materials for commodities. A by-product of the mineral extraction is the mine waste also called tailings. Tailings are safety storage in tailing dams. Tailings dams troughs the history had had incidents and failures. Economic, environmental and social consequences of a tailing dam failure could be devastating. Soil strength is given by the consolidation, particle shape, stresses path, water content, hydraulic conductivity among other factors. Change on these factors produces changes in the soil strength. The development of economical and fast tests could improve the safety of the tailings deposits. Drained direct shear tests using uniformed graded tailing particles were performed. Three different size ranges 0.25, 0.125 and 0.063mm were used. Effect of particle size on shear strength and the effect of shearing on the tailing particles were studied. Normal consolidation pressure, void ratio, particle size and particle shape were monitored properties. Strength of the tailings was related with the monitored properties to suggest four empirical relations, two of them base in the morphology of the particle and two bases in the angularity. Results have shown that particle elongation diminishes the tailings strength but the angularity increases the strength. Particle size results are ambiguous and seem to be more related with the shape descriptor.

Introduction The first mining vestiges can be found as early as in the paleolithic era and it is known that mining appears and develop as the civilization does (Armengot, et al., 2006). From the industrial revolution to the modern world the technology had advance enabling to the mining industry to increase maybe exponentially the amount of ore extracted, even at lower grades than before. The mined ore rock is crushed and milled to liberate and concentrate the precious mineral from the host rock; the valueless rock debris left after wet concentration is the so called tailings and since ore extraction increases also tailings do. Typical amount of tailing materials produced by the ore extraction is 43% for iron ore (EPA, 1994) and 99% for copper ore (Northey, et al., 2014). With this amount of left overs tailings dam structures around the world should contain hundreds, thousand, or even millions of tons of tailings e.g. 2 million metric tons tailing deposit in Lokken mine, Norway (Wolkersdorfer & Bowell, 2005). Historically incidents as Aznalcollar (Spain) in 1998, Baia Mare (Romania) and Aitik (Sweden) both in year 2000 and the public opinion had encouraged to the representatives involved to work in the prevention instead of react after an incident. Tailings dams are usually considered as walk-away solutions and needs to be designed and constructed to be safe in a long time perspective. The consequences of the failures may be fatal to the local society and harm the surrounding environment. As a consequence of the operation and raising procedures of tailings dams, the conditions in the tailings dams could be considered to be dynamic in a longer perspective. Grain size distribution, formation of layers, pore pressures and stress states are continuously changing during the operation (Ormann, et al., 2013). Tailings may be susceptible to weathering in the deposit environment. The safety of the impoundment, structural and environmental, in a long time perspective should be addressed. Relatively few studies had been developed in tailings properties considering the dam failure consequences. Design of safe tailing dams should not only account for actual tailing properties but also by future changes e.g. mechanical weathering. Sophisticated equipment to monitor the tailing dams and laboratory tests could be expensive thus; economical and reliable monitoring methods development could help to overcome this constrain and improve the safety. Empirical relations found in literature (Cho et. al., 2006; Rousé et. al., 2008) have shown that they underestimate the internal friction angle (Rodriguez & Edeskär, 2013) when they are applied to tailings. The inclusion of an empirical relation that could predict qualitatively the tailings strength would enhance the tools available to avoid incidents as mentioned before. In this study direct shear test pre-consolidated in the range of 50 to 500 kPa were performed using uniformed sized tailing particles. Internal friction angles obtained were in the same range of other copper tailings (Guangzhi, et al., 2011 and Liu, et al., 2012). According to Vick (1990) effective stress level is the most important parameter controlling the internal friction angle with the strength envelope curved at high stress levels as a result of particle crushing. Breakage at low stress levels can also break the tailing particles due to stress concentration in its angularities (Lade, et al., 1996). During the research mechanical weathering was monitored using regular sieving to detect the fines generated and image analysis to identify the particle shape change. Results have shown that particle size is the driving parameter that determines the fine generation and void ratio since both increases as the particle size increases in agreement with Craig (2002) and Meade (1968).

Furthermore the empirical relations suggested were not able to predict the particle size effect. Form and roundness show an opposite effect on the internal friction angle maybe result of the alignment of the particles during shearing. The increase of the angularity strengthens the tailings but the elongation of the particles weakens it. The limited amount of data shrinks the action area of the empirical relations but state an initial relation for further research and new data acquisition. The particle shape In this study the word shape is used to describe a grain’s overall geometry. Furthermore, in order to describe the particle shape in more detail, there are a number of terms, quantities and definitions used in the literature. Some authors (Mitchell and Soga, 2005; Arasan, et al., 2010) are using three sub-quantities describing the shape but at different scales. The sub-quantities are morphology/form, roundness and surface texture. In Figure 1 is shown how the scale terms are defined.

At large scale a particle’s diameters in different directions are considered. At this scale, describing terms as spherical, circular, platy, elongated etc., are used. An often seen quantity for shape description at large scale is sphericity. Graphically the considered type of shape is marked with the dashed line in Figure 1. At intermediate scale is focused on description of the presence of irregularities. Depending on at what scale an analysis is done; corners and edges of different sizes are identified. By doing analysis inside circles defined along the particle’s boundary, deviations are found and valuated. The mentioned circles are shown in Figure 1; a generally accepted quantity for this scale is roundness or the antonym: angularity. Regarding the smallest scale, terms like rough or smooth are used. The descriptor is considering the same kind of analysis as the one described above, but is applied within smaller circles, i.e. at a smaller scale. Surface texture is often used to name the actual quantity. Materials and Methods In this study samples from the Aitik tailings dam has been used. The Aitik tailing dam is located about 100 km North of the Arctic Circle in the boreal parts of Northern Sweden (Figure 2) about 15 kms from the community of Gällivare. The value mineral is chalcopyrite (CuFeS2). Main sulphides are pyrite, chalcopyrite and sphalerite. Main gangue minerals are quartz, feldspar, plagioclase and mica (Lindvall and Eriksson, 2003).

Figure 1 Particle describing the shape scale attributes (Mitchell and Soga, 2005)

Disturbed samples were taken from a depth of 0 to 1 meters in a trial pit. The trial pit was located (Figure 2, right) in the north east part of the dam. The sample from the trial pit was splitted into three different test sizes as it is shown in table 1. Further and for simplicity size ranges are called by its lower limit.

Table 1 Particle size ranges used during direct shear tests Particle size (mm)

Upper limit Lower limit 0.5 0.25 0.25 0.125

0.125 0.063

Wet sieving was used with Sodium Diphosphate decahydratate (Na4P2O7·H2O) as a dispersant to enhance the particle separation. After sieving specimens were dried for 24 hours at 105° Celsius. Remolded samples in 50mm diameter and 170mm length sampling tubes were casted with one particle size (e.g. 0.25mm) by using the methodology describes by Dorby (1991). Dorby’s procedure includes the filling of the tube specimen by steps, usually 5 to 6 in total, where each step comprises 2-3cm of the tube height. Water is added until the step is reached followed by the addition of the tailings sample and posterior self-settlement for at least 6 hours. Same procedure is followed for every step until the tube is filled up. This methodology imitates the natural sedimentation process leading the tailings settle in beds with natural segregation. Since the test specimens are uniformly graded the segregation should be based on the grain density differences and not in the particle size. Image acquisition was performed through a microscope (Motic B1) lightening sources from below and from the side of sample. The magnification lenses 4x was used for 0.25 mm sample

Figure 2 Mark shows the location of the Aitik tailing dam in the north of Sweden (left) and the tailings dam (right). (Google map, 2015)

and 10x for 0.125 and 0.063mm samples. The magnification used for each sample was chosen based on the results obtained by Rodriguez et. al. (2012). The camera mounted on top of the microscope ( Infinity 2) has 2 megapixel resolution. The quantities in the Table 2 were used to determine the shape of the particles. The two dimensional image analyses to determine the shape quantity was done using (ImageJ, 2013) software. Direct shear tests were performed on the remolded disturbed samples. Samples were mounted by surrounding reinforced latex membrane and porous filter spikes were placed on top and bottom. Rubber tape at the end of membrane edges was used to avoid any leakage from membrane edges (see Figure 3). NGI (Norwegian Geotechnical Institute) direct shear apparatus was used for this study. The apparatus has been rebuilt and modified with electronic sensors which enable to record applied load, specimen height and pore pressure continuously during shearing. Shearing velocity was 0.012 mm/min. The logged data is then transferred to computer program which helps with the monitoring of stresses and deformations during the test. Tests were performed in saturated and drained conditions. Uniform sample size as table 1 shows were used. Consolidation loads for the samples were 50. 100. 150, 300 and 500 kPa. According with the Swedish criteria (Swedish Geotechnical Society, 2014) the shear strength is evaluated at 0.15 radians shear angle if no shear peak stress has been observed.

Table 2 Quantities and definition After shearing the samples were dried for 24 hr at 105° C and studied by image analysis to recognize the effects on the tailings by regular sieving and shape change. Sieving in the same size as the original sample was used to determine the amount of fines generated by weight (see

QUANTITY NUMBER DEFINITION REFERENCE QUANTITY

NUMBER DEFINITION REFERENCE

Q1

ImageJ (2013) Q2

ImageJ (2013)

Q3

Cox (1927) Q4

Mora and Kwan (2000)

A, area AC, area convex P, perimeter Major, major axis based on fitting ellipse Minor, minor axis based on fitting ellipse

Figure 3 Direct shear apparatus and sample mounting (Bhanbhro, 2014)

MinorMajor

2Major 4A

2PA4

CAA

Table 4). Shape recognition was determined under same conditions mentioned before. Original samples were compared with the subset of tests (see Table 5). Results Consolidation In order to achieve the required consolidation, the normal loads were applied with an increment of 50-100kPa. Since all the samples were constructed with uniformed particle sizes, the rapid consolidation process was observed. The pore pressures suddenly increased on application of normal load followed by rapid decrease to its original state (Figure 4) due to fast drainage. The Figure 4 shows the typical consolidation curves for the sample prepared with 0.063mm particle size and normal stresses of 150kPa. Table 3 and Figure 5 show the initial void ratio for the test specimens higher void ratios are recognized for bigger fraction size. Void ratios in all samples are diminishing while consolidation pressures increase.

Table 3 Initial void ratio after pre-consolidation before shearing

0.25mm 0.125mm 0.063mm load initial load initial load initial

(kPa) void ratio (e) (kPa) void ratio

(e) (kPa) void ratio (e)

0 1.096 0 0.978 0 0.767 50 1.061 50 0.948 50 0.744

100 1.034 100 0.943 100 0.730 150 1.015 150 0.924 150 0.709 300 0.989 300 0.884 300 0.677 500 0.917 500 0.831 500 0.679

Figure 4 Typical consolidation behavior for sample size 0.063mm for normal stress of 150 kPa

Shearing behavior and results

Figure 6 shows the typical behavior of the samples during shearing. Lines show the height reduction and shear force during the shearing process. Shearing resistance shows a unique behavior at all normal load tests; hardening follow by a slight softening. Primary and secondary height reductions can be seen in the samples behavior during shearing (see circles in Figure 6). Dilatancy can be found in samples 50,100,150 and 300kPa in the middle of the primary and secondary height reduction (see Figure 6, left). Samples at 500kPa present only height reduction with height stabilization in the middle (see Figure 6, right). It could suggest a primary and secondary rearrangement of the particles. Effect of shearing on particles Table 4 shows the percentages of fines generated after the shearing tests depending on the normal load applied to consolidate the sample. The amount of fines particles in size 0.25mm seems to increase. For size 0.063mm is more stable for all loads and for 0.125mm size there is a peak at 150kpa with high breakage compare with the rest loads. Graphical display can be seen

Figure 5 Initial void ratios as a function of normal consolidation

Figure 6 Typical sample height and shear resistance evolution during the test for 50 (left) and 500 kPa (right) consolidation loads (sample 0.063mm). Same height reduction behavior is

present in samples 50, 100, 150 and 300 kPa consolidations.

in Figure 7 where fraction 0.25mm has a more clear increase of breakage in relation with the specimens 0.125 and 0.063mm anyway there is a slight increase in the fines generated by the tests as the consolidation load increases for all samples.

Table 4 Particle fines content by sample and test load Size Load Fine Size Load Fine Size Load Fine (mm) (kPa) content (%) (mm) (kPa) content (%) (mm) (kPa) Content (%) 0.25 50 26.6 0.125 50 21.0 0.063 50 10.3

100 32.8 100 24.1 100 9.2 150 29.5 150 29.6 150 10.9 300 34.8 300 22.3 300 12.1 500 39.6 500 23.9 500 12.7

Shape change in the different studied fractions is shown in Table 5. In relation with the original shape only the fraction size 0.063mm have change the shape for all quantities while fraction 0.125mm has change in half of them but in both cases for the highest consolidation load 500kpa.

Table 5 Quantity values for particles by size and load state Size

(mm) Quantity Initial value 50kPa 500kPa

0.063 1 1.513 1.494 1.422 2 0.698 0.698 0.729 3 0.707 0.703 0.730 4 0.934 0.932 0.943

0.125 1 1.472 1.434 1.386 2 0.714 0.733 0.750 3 0.701 0.698 0.703 4 0.933 0.932 0.934

0.25 1 1.452 1.397 1.395 2 0.718 0.743 0.739 3 0.693 0.682 0.685 4 0.928 0.922 0.924

Figure 7 Amount of breakage by weight by shearing at different consolidation loads

Suggested empirical relation Table 6 contains laboratory data collected from the samples. This data has been used to develop an empirical relation to estimate the internal friction angle as a function of the consolidation load, void ratio, particle size and particle shape with the help of a regression analysis.

Table 6 Quantity values for the samples Suggested empirical relations equations are listed below (table 7, 8, 9 and 10); they were obtained from a regression analysis. Additionally to the regression analysis a subset regression has been carried out to identify the factors affecting the empirical model. Data undergoing the empirical relation represents the R2 when using the available parameters. For all the models the best fit includes all the suggested parameters.

Table 7 Empirical relation for Q1 φ = 289 – 0.134 c – 76.3 e + 63.6 s – 134 Q1 [1]

R2 Consolidation (c)


Size (s)

Q1

56.4 X X 60.5 X X X 56.7 X X X 73.9 X X X X

residual ±2

Table 8 Empirical relation for Q2 φ = -5 -0.060 c – 61.1 e + 76.4 s + 109 Q2 [2]



Size (s)

Q2

56.4 X X 60.5 X X X 56.6 X X X 65 X X X X

residual ±2

Table 9 Empirical relation for Q3 φ = 801 -0.1526 c – 43.6 e – 27.9 s – 1036 Q3 [3]



Size (s)

Q3

56.4 X X 40.0 X X 94.6 X X X 60.5 X X X 96.7 X X X X

residual ±0.7

Internal Consolidation (c) Initial Size (s) Quantity values (Q) friction angle (φ) kPa void ratio (e) mm 1 2 3 4

25.9 50 0.75 0.063 1.513 0.707 0.698 0.934 25.2 100 0.73 0.063 1.494 0.702 0.698 0.932 23.3 50 0.95 0.125 1.472 0.701 0.714 0.933 17.9 100 0.94 0.125 1.434 0.698 0.733 0.932 22.1 50 1.06 0.25 1.452 0.693 0.718 0.928 26.7 100 1.03 0.25 1.397 0.682 0.743 0.922

Table 10 Empirical relation for Q4

φ = 2026 – 0.1431 c +2 e – 95.3 s – 2129 Q4 [4] R2 Consolidation

(c) Initial void

ratio (e) Size (s)

Q4

56.4 X X 50.0 X X 93.1 X X X 83.6 X X X 93.1 X X X X

residual ±0.9 Discussion Drained direct shear test were performed in homogenous tailings particle sizes 0.25, 0.125 and 0.063mm. Internal friction angle was calculated using the data obtained from the tests. Normal consolidation pressure, initial void ratio, particle shape and size (Table 6) were monitored. Shear test results Results indicate that the void ratio is governed by the consolidation pressure and particle size as it is indicated in Table 3 and Figure 5. The consolidation pressure reduces the void ratio and the void ratio is higher in in bigger particles. The consolidation pressure has a direct effect on the void ratio and the compression index represents this relation (Craig, 2002). Void ratio is dependent on the grain-size characteristics of the soil, primarily mean particle size and sorting (Meade, 1968). Figure 6 height sample changes during test can be notice; height changes present primary and secondary settlements recognized by the continuous height reduction probably result of the rearrangement of the particles. In those samples having dilatant behavior the particles rotate and rearrange overlapping the rest of the particles increasing the sample height. Samples subject to 500 kpa present a flat surface (Figure 6, right) here the normal load is enough to not let the sample to dilate. The deformation of soil mass can required a low energy if particles are free to rotate; this is the case of loose packing (Santamarina & Cho, 2004). The lack of particle rotation suggests a higher breakage corroborated by the Figure 7 that shows the highest breakage for all samples at 500 kpa. Internal friction angles results range between 17.9˚ and 26.7˚ this values are in agreement with those observed by Guangzhi, et al. (2011) and Liu, et al. (2012) in direct shear tests for copper tailings (range between 24.5˚ and 28.1˚). Effect of shearing on particles Void ratios (Figure 5) and particle size reduction (Figure 7) happened during the increase of the normal consolidation load. This particle size reduction have a strong jump from 0 to 50kpa resulting in a 10 to 25% of fines generation (depending on the particle size). In this study size

reduction has been detected even before tests start in stored ready to use samples. Ocular inspection on coarse size sample before test (0.25mm) show a fines generation (reduction in size) probably result of the handling and preparation. Lade, et al. (1996) suggest that angular particles are very susceptible to break even at low stresses because they concentrate in the angular contact points. Furthermore the particle breakage is also a function of time even under constant stresses (Yamamuro & Lade, 1993). Time and load could explain this breakage detected during storage and handling of the tailing samples. Peak in Figure 7 sample 0.125mm at 150 kpa of consolidation suggest this behavior, sample probably broke before test and that’s why result is an outstanding difference with respect the neighbor loads (100 and 300 kpa). An initial high breakage will be produced even at low consolidation loads with a slight increase during each step increased load. Breakage of the coarse fraction (0.25mm) show no shape change (see Table 5) and only the elongation of the particles were affected in size 0.125mm at high consolidation load (500 kPa), roundness was not affected. Fraction 0.063mm show shape change in all shape descriptors at high loads (500 kPa). In perspective this changes could influence the final tailings strength and this relations are discuss within the empirical relations. Empirical relations Empirical relations suggested in this study are applicable under defined conditions. Empirical relations equations 1, 2, 3 and 4 were obtain from the laboratory data showed in Table 6. Empirical relations were suggested using the common parameters initial void ratio, particle size, consolidation pressure but differing in the quantity used. The intention to introduction diverse quantities is to evaluate its relationship or contribution. Results from Cho et. al. (2006) to estimate the internal friction angle has a correlation coefficient of R2 = 0.84 and Rouse et. al. (2008) varies from ± 3.1 to 5.3 degrees with 95% of confidence. Results in this papers has shown an R2 = 0.95 in the best of the cases (See equations 1, 2, 3 and 4) with ±2 and ±0.8 residual values depending in the used quantity. Data in Table 6 was used to asses a regression analysis. Shape and void ratio are the driving parameters in the empirical relations (in this order), this is recognized by the multipliers (see equations 1, 2, 3 and 4. The exception is in equation 4 where void ratio absolute value is smaller. Empirical relations are suggesting that friction angle would decreases if consolidation increase and void ratio decreases. It is evident that the physical changes in void ratio and consolidation are not in agreement with the known, if consolidation (relative density) increases and void ratio decreases friction angle should increase (Holubec and D’Apolonia, 1973). This can be explained if we consider the entire equations. Shape is the main driving parameter in the equations and probably the rest of the parameters are adjusting the model with no physical mining. From this conclusion it is not possible to determine the real physical effect of the consolidation, size and void ratio but shape as a main driving parameter could.

Quantities 1 and 2 are related with the form (first order scale of the Mitchel and Soga classification) and quantities 3 and 4 describe the roundness (second order scale). Empirical relations results show that when particles become more regular in form (quantities 1 and 2) the friction angle (φ) increases but when they become more rounded (quantities 3 and 4) friction angle (φ) decreases. The methodology used to cast the direct test samples is based in the natural sedimentation process (Dorby, 1991) and it is possible that this could have an effect on the results. Results could suggest that elongated particles like mica minerals are deposited in horizontal layers creating preferential sliding horizons that have reduced strength compared with regular in form particles. Harris, et al., (1984) found decrees in the shear resistance with increase the mica content in regular sands. Furthermore the void ratio increases with mica content (and flat-elongated particles) resulting in a looser and weaker soil (Santamarina & Cho, 2004; Chen, et al., 2005). Deposition methods in tailing dams as spigotting could create this preferential layering deposition. The mica minerals presence during further experimentation should be account to understand the effects in tailings. Tailings dam research should consider deposition method effects over the particle settlements and layering in the tailing impounds that could drive the strength of the materials. Tailings strength decrease when roundness increases (equations 3 and 4) in agreement with (Santamarina & Cho, 2004). In this case angular tailing particles are providing more interlocking strength been necessary to break the corners to be able to rearrange the soil structure. Figure 7 shows the fine generation increase while increasing stresses and Figure 6 is the behavior of the height sample during shearing. Considering both (Figure 6 andFigure 7) it seems like samples have a primary and secondary consolidation separated for a height sample increase; this height increase could represent a major rearrangement of the particles (overlapping). The increases of stresses in the sample avoid the rearrangement generating more fines. The breakage of the particles is not necessary to produce more angular or rounded material, corners can always break and keep its angularity as direct shear results showed it (Table 5 sizes 0.25 and 0.125mm). Particle size increase has been identified in the literature as a possible soil strength parameter modifier inherent to the material strengthening the soil in some studies (Lewis, 1956; Kolbuszewski & Frederick, 1963) but weakening in others (Kirkpatric, 1965; Marschi, et al., 1972) or even having no effect (Bishop, 1948; Vallerga, 1957). Particle size influence in the friction angle is ambiguous for this research results. Quantities seem to have the main influence on the size. For form quantities, if the size increases the strength increases but for roundness quantities if the size increases the strength decreases. The limited amount of data shrinks the action area of the empirical relations but state an initial relation for further research and new data acquisition. In perspective if particles become more rounded the strength of the tailings could be reduced. If particles are more regular in form the strength should not be compromised. However the reduction of the sizes (fine soil production) could have an effect that it is still unknown. The validity of the empirical relations is limited to the minimum and maximum values show in Table 6.

Conclusions

Effect of particle size on shear strength was ambiguous and not clearly identified. Friction angle increases when angularity increases. Friction angle increases when particle form is more regular or symmetric. Shearing of the particles is producing fragmentation with no change in shape for coarse

samples but roundness and form symmetry for finer fraction 0.063mm. Symmetry form of the tailing particles was increased for samples size 0.125. Empirical relations were suggested

References

Arasan, S., Hasiloglu, A. S. & Akbulut, S., (2010). Shape particle of natural and crushed aggregate using image analysis. International Journal of Civil and Structural Engineering, 1(2), pp. 221-233.

Armengot, J., Espi, J. & Vazquez, F., (2006). Origenes y desarrollo de la mineria. Industria y mineria, Volym 365, pp. 17-28.

Bhanbhro, Riaz., Rodriguez, Juan., Edeskär, Tommy and Knutsson, Sven (2014). Electronical Journal of Geotechnical Engineering, vol. 19, bund Z, pp. 9023-9039.

Bishop, A., (1948). A large shaer box for testing sand and gravels. Proceedings of the 2nd international conference of soil mechanics and foundation engineering, Volym 1, pp. 207-211.

Chen, Chang & Lin, (2005). Influence of coarse aggregate shape on the strength of asphalt concrete mixtures. Journal of Eastern Asia Society for transportation studies, vol. 6, pp. 1062-1075.

Cho, G., Dodds, J., and Santamarina, (2006). Particle shape effects on packing density, stiffness and strength: natural and crushed sands. Journal of Geotechnical and Geoenvironmental Engineering, 132 (5), pp. 591-602

Cox, E. P. A (1927). method of assigning numerical and percentage values to the degree of roundness of sand grains. Journal of Paleontology, 1(3), pp.179-183.

Craig R.F. (2002). Soil mechanics. Spon press, Taylor and Francis group. Sixth edition. London and New York.

Dorby, R., (1991). Soil properties and earthquake ground response. Proceedings of the 10th European Conference on Soil Mechanics and Foundation Engineering. Florence, Italy.

EPA, (1994). Extraction and beneficiation of ores and minerals, Volume 3: IRON, Washington: U.S. Environmental Protection Agency, EPA 530-R-94-030, NTIS PB94-195203.

Guangzhi, Yin., Guangzhi, Li., Zuoan, Wei., Ling, Wan., Guohong, Shui and Xiaofei, Jing. (2011). Stability analysis of a copper tailings dam via laboratory model tests: A Chinese case of study. Mineral Engineering, vol.24, pp. 122-130.

Harris, W., Parker, J. & Zelazny, L., (1984). Effects of mica content on engineering properties of sand. Soil Science Society of America, 48(3), pp. 501-505.

Holubec and D’Apolonia, (1973). Effect of particle shape on the engineering properties of granular soils. ASTM STP, vol. 523, pp. 304-318.

ImageJ, (2013). version 1.47v. Wayn Rasband, National Institutes of Health, USA.. [Online]. Available at: http://imagej.nih.gov/ij/

Kirkpatric, W., (1965). Effects of grain size and grading on the shearing behavior of granular materials. Proceedings of the 6th International Conference on Soil Mechanics and Fundation Engineering, vol. 1, pp. 273-277.

Kolbuszewski, J. & Frederick, M., (1963). The significance of particle shape and size on the mechanical behavior of granular materials. European Conference of soil mechanics and foundation engineering, vol. sec. 4 paper 9, pp. 253-263.

Lade, P. V., Yamamuro, J. A. & Bopp, P. A., (1996). Significance of particle crushing in granular materials. Journal of Geotechnical Engineering, vol. 122(4), pp. 309-316.

Lewis, J., (1956). Shear strength of rockfill. Proceedings of the 2nd Australia-New Zeland conference of soil mechanics and foundation engineering, pp. 50-52. Lindvall, M. and Eriksson, N. (2003) Investigation of the weathering properties of tailings sand from Boliden´s Aitik copper mine – a summary of twelve years of investigations. Proceedings 6th International Conference on Acid Rock Drainage. Cairns.

Liu, Hai-ming., Yang, Chun-he., Zhang, Chao and Mao, Hai-jun. (2012) Study of static and dynamic strength characteristics of tailings silty sand and its engineering application. Safety Science, vol. 50, pp. 828-834.

Marschi, N., C.K., C. & Seed, H., (1972). Evaluation of properties of rock fill materials. journal of the soil mechanics and foundation division, vol. 98(1), pp. 95-114.

Meade, R.H. (1968). Compaction of sediments underlying areas of land subsidence in central California. U.S. Geological Survey Professional Paper, vol. 497-D, pp. 39

Mitchell and Soga, (2005). Fundamentals of soil behavior. 3rd ed. :Wiley. Mora, C. F. and Kwan, A. K. H. (2000). Sphericity, shape factor, and convexity

measurement of coarse aggregate for concrete using digital image processing. Cement and Concrete Research, 30, (3), pp.351-358.

Northey, S., (2014). Modelling future copper ore grade decline based on a detailed assessment of copper resources and mining. Resources, Conservation and Recycling, vol. 83, pp. 190-201.

Ormann, L., Zardari, M., Mattsson, H. & Knutsson, S., (2013). Numerical Analysis of strengthening by rockfill embankments on an upstream tailings dam. Canadian Geotechnical Journal, vol. 50, pp. 391-399.

Rousé, P.C.; Fennin, R.J.; Shuttle, D.A. (2008). Influence of roundness on the vid ratio and strength of uniform sand. Geotechnique, vol. 58 (3), pp. 227-231.

Rodriguez, J. M.; Johansson, J. M. A. and Edeskär, T. (2012). Particle shape determination by two-dimensional image analysis in geotechnical engineering. Danish Geotechnical society bulletin, 27, pp. 207-218.

Rodriguez, Juan & Edeskär, Tommy (2013). Case of study on particle shape and friction angle on tailings. Journal of Advanced Science and Engineering Research, vol. 3 (4), pp. 373-387.

Santamarina, J. & Cho, G., (2004). Soil behaviour: The role of particle shape. London, Skempton Conf. London.

Swedish Geotechnical Society (2014). Direkta skjuvförsök – en vägledning (in Swedish), Linköping: Swedish Geotechnical Society, SGF Notat 2:2004, SGF:s Laboratoriekommitté.

Vallerga, B. (1957). Effect of shape, size and surface roughness of aggregate particles on the strength of granular materials. ASTM Special technical report No. 212.

Vick, S.G. (1990). Planing, Design and Analisys of Tailing Dams. BiTech Publisher Ltd, Canada. Wolkersdorfer & Bowell (2005). Contemporary reviews of mine water studies in Europe. Mine Water Environ. Vol. 24, pp. 1-76.

Yamamuro, J. & Lade, P., (1993). Effects of strain rate on instability of granular soils. Geotechnical Testing Journal, vol. 16(3), pp. 304-313.

Paper VII Bhanbhro, R.; Rodriguez, J.M.; Edeskär, T. and Knutsson, S. (2013). Basic description of tailings from Aitik focusing on mechanical behavior. International Journal of Emerging Technology and Advanced Engineering. Vol. 3, No. 12, pp. 65-69. ISSN: 2250-2459.

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified Journal, Volume 3, Issue 12, December 2013)

65

Basic Description of Tailings from Aitik Focusing on Mechanical Behavior

Riaz Bhanbhro1, Roger Knutsson2, Juan M. Rodriguez3, Tommy Edeskar4, Sven Knutsson5 1, 2,3,4,5Division of Geotechnology, Department of Civil Environmental and Natural Resources Engineering, Luleå University of

Technology, SE-97187, Sweden

Abstract— Tailings are artificial granular materials that behave different as compared to natural soil of equal grain sizes. Tailings particle sizes, shapes, gradation and mechanical behavior may influence the performance of tailings dams. Hence it is essential to understand the tailings materials in depth. This article describes present studies being carried out on Aitik tailings. Basic tailings characteristics including specific gravity, phase relationships, particle sizes, particle shapes and direct shear behavior are presented in this article. The results showed that particles size decreases along depth from surface for collected sample locations. The angularity of the particles increases as the grain size decreases. Vertical height reduction was observed during shearing of samples by direct shear tests.

Keywords—Mechanical Properties of tailings, Tailings Particles, Tailings, Mechanical behavior of tailings

I. INTRODUCTION Tailings dams are geotechnical structures which are

raised with time [1] as the impoundments are increased depending upon production rate of mining activity. Generally tailings itself are used in some extent for construction of tailings dams. Hence the mechanical properties of tailings material have important role in construction of tailings dams. Tailings are artificial granular materials and not like natural soils [1, 2]. Thus tailings might behave differently e.g. in anisotropic shear strength, permeability properties [3] and particle shapes which possibly might affect the performance of dam. Tailings dams are supposed to withstand for long times, i.e. in general as walk away solutions. For safe existence of the tailings dams it is important to know mechanical behavior of tailings being used in construction of dams.

This paper presents the initial stage of laboratory work being carried out on Aitik tailings materials. Preliminary results from specific gravity test, phase relationships, particle size analysis, particle shapes analysis and vertical height behaviors during direct shear tests are discussed.

II. EXPERIMENTAL WORK Samples were collected by the consulting company

Sweco during spring 2013 at the depths 12-47m from the sections of dam E-F and G-H of the Aitik Tailings dam. Figure 1 shows the locations of samples A, B, C and D from the dam sections. The samples were taken from weak zones previously determined by CPT tests. Both disturbed and undisturbed samples were collected.

Figure 1 Location of samples, DAM E-F and DAM G-H from Aitik

Tailings Dam

The laboratory work was carried out on the collected samples for determination of basic characterization; Particle size, hydrometer analysis, particle shapes, and strength parameters by direct shear. Undisturbed samples were used for determination of shear strength parameters by direct shear tests. Drained and undrained tests were performed from each sample tube for direct shear test.

III. BASIC DESCRIPTION OF TAILINGS The body of tailings is composed of solid particles,

water and air. Table 2 shows the summary of tests performed for basic geotechnical characterization on Aitik Tailings in this study. These tests were performed on the undisturbed samples. Water content (w) for the tested samples were in range of 15.2-47%.



66

The bulk density ( ) was determined to be in range of 1.66 - 2.06 t/m3. Similarly the saturated density (ρsat) and dry density (ρd) were found to be in range of 1.76 – 2.065 t/m3 and 1.18 – 1.65 t/m3 respectively. The average particle density ( for collected samples was 2.833 t/m3. The void ratio (e) and porosity (n) were calculated to be in range of 0.72 – 1.41 and 41.9 – 58.5% respectively.

A. Particle Size Sieve analysis was conducted for undisturbed and

disturbed samples. Similar grain size distribution curves were observed for disturbed and undisturbed samples. It was observed that the particle size decreases with depth from the surface of dam section for the locations under investigation. This decrease is might be due to breakage of particles because of higher stresses, particle decaying over time, chemical reactions or as a consequence of changes in ore quality or process technology. The other possibility of size reduction with depth is that the deposition methods and locations of depositions in earlier years were different.

And may be because of the samples were taken from different distances from the dam. The particles size distribution curves are shown in figure 2 and the values of D30, D50, and D60 are read from particle distribution curves and presented in Table 1.

TABLE 1 SIEVE CURVE CHARACTERISTICS, D30, D50, AND D60

Sample Depth (m)

D30

(mm)

D50

(mm)

D60

(mm)

GH56+450 inkl 12-15 0.078 0.14 0.21

DGH56+450E 12-15 0.11 0.16 0.2

Temp62+315 18 0.025 0.05 0.062

DEF62+315D 20 0.018 0.028 0.035 VFT 349

undisturbed 38 0.0032 0.006 0.0077

Temp 56+450D 38 0.0032 0.006 0.008

DEF62+315D 43 0.0035 0.007 0.0092 GEO29B

undisturbed 47 0.0039 0.008 0.011

TABLE 2 SUMMARY OF BASIC GEOTECHNICAL CHARACTERIZATION ON AITIK TAILINGS

Sample Description Water content

(%)

Bulk Density

( ) t/m3

Saturated Density

(ρsat) t/m3

Dry Density (ρd) t/m3

Void Ratio [e]

Porosity [n] (%) Tube

Elevation /Depth

BKAB125 387.1/7.6 15.2 1.681 1.944 1.46 0.941 48.5 ORRJE4786 384.6/10.1 23.9 1.787 1.933 1.44 0.964 49.1

KK1822 365.0/18.6 37.22 1.915 1.903 1.40 1.030 50.7 CTH546 365.0/18.6 43.7 1.997 1.899 1.39 1.039 51.0 AIB839 363.6/20 39 1.950 1.908 1.40 1.020 50.5 VFK438 363.6/20 37.2 1.856 1.875 1.35 1.094 52.2

KLBF784 371.5/21.1 47.1 1.831 1.806 1.24 1.276 56.1 GL41 371.5/21.1 45.7 1.859 1.826 1.28 1.220 55.0

BBK93 370.4/22.2 43.3 2.03 1.92 1.42 1.0 50 VPLANB150 370.4/22.2 43.9 1.887 1.848 1.31 1.161 53.7

VFT349 343.16/38 41.4 1.66 1.76 1.18 1.411 58.5 AIB852 343.16/38 39.5 1.914 1.887 1.37 1.066 51.6

5580 344.7/47.4 28.6 2.06 2.065 1.65 0.721 41.9 GEOB29 344.7/47.4 38.8 1.99 1.93 1.43 0.977 49.4

HSRB1016 344.7/47.4 37.5 1.933 1.909 1.41 1.016 50.4



67

Roundness (Intermediate Scale)

Surface Texture (Small Scale)

Morphology (Large Scale)

Figure 2: The particle size distribution curves at various depths relative to the surface level of Dam section

B. Particle Shapes Particle shapes are known to affect various engineering

properties of soils including friction angle and permeability [5]. One of the general differences between natural soils and tailings is that tailings are more angular compared to soil particles. As a step to understand the difference of mechanical behavior between soil and tailings it is of interest to quantify this difference. Particle analysis can be described as quantitative and qualitative; qualitative description is subject to shape of particles whereas quantitative refers to measuring of dimensions [6].

Figure 3: Particle shape scale factors illustrated on tailings from this study. After Mitchell & Soga (2005) [7]

Particle shapes and properties are categorized in different scales, number of terms and their details. Mitchell & Soga (2005)[7], Rodriguez & Edeskär (2013)[11] and Rodriguez J,M.(2013)[12] described the particle shape in three terms; which are morphology, roundness and surface texture, presented in figure 3. Morphology is described as a particles’ diameter at large scale. At this scale terms are described as spherical, platy, elongated or elongation etc. The intermediate scale presents the explanation of irregularities i.e. corners, edges of different sizes.

This scale is generally accepted as roundness or angularity; and smaller scale defines the roughness or smoothness and surface texture that can be whole particle surface including corners.

Figure 4: Particle shapes of different sizes from Aitik Tailings

Powers (1953) [8] introduced the roundness qualitative scale for particle shapes which depends upon shapes of particles. Using the Powers roundness qualitative scale (Intermediate scale, Roundness), it is initially concluded that particle shapes of larger to smaller sizes from Aitik’s samples varied from sub angular to very angular. The particles having bigger diameter (1 mm) are less angular as compared to smaller diameters (0.063 mm). Figure 4, shows images of particle shapes of different sizes, which were collected for analysis from Aitik Tailings.

IV. VERTICAL HEIGHT REDUCTION DURING DIRECT SHEAR TESTS

Direct shear tests (18 drained and 12 undrained) were performed on undisturbed samples. Samples were mounted with minimum disturbance surrounded by reinforced latex membrane and porous filter spikes were placed on top and bottom. Rubber tape at the end of membrane edges was used to avoid any leakage from membrane edges especially when test was performed as undrained.

1mm 0.5mm 0.125mm 0.063mm



68

Figure 5: Direct Shear apparatus and sample mounting

Normal stresses were applied at rate of 20 kPa (in steps) per hour or by monitoring dissipation of pore pressures. Shearing rate was kept as 0.018 mm per minute. Samples were sheared up to 5mm horizontally with shearing angle up to 0.40 radians. Figure 5 shows the direct shear testing machine, membrane, filters, mounting of samples.

Figure 6: Typical vertical height changes and shear stress behaviors

during direct shear test

During direct shear test, reduction in vertical height was observed for all the conducted tests i.e. drained and undrained. The reason for this behavior can be rearrangement of particles or breakage of particles. Figure 6 shows the shear stress and vertical height changes along shearing angle. Slight increase in pore pressure was also observed with vertical height reduction while shearing (Figure 7); this indicates the decrease in voids. Stresses on particle edges may lead to additional crushing; creates particles that are very angular which might offer higher resistance to shear [9].

Figure 7: Typical vertical height changes and pore pressure behaviors

during direct shear test

V. DISCUSSIONS The results from this study shows that the grain sizes

were reduced by depth relative to surface. The fine content in the samples increased from 5% in the upper layer (above 20m) to about 20% in the subsequent layers 20-47m below ground surface. The increase in finer particles along depth might reduce the permeability and may lead to higher pore pressures. The void ratio of investigated tailings was found to be about 20% to 40% higher as compared to natural soils (silt)[4], which indicates loose condition. The observed vertical height reduction during shearing might lead to increased pore pressures because of reduction in voids. Increased pore pressure might transit the stress supporting grain system to fluid-grain slurry resulting loss of strength [10]. If the vertical height reduction is due to breaking of particles then bigger particles may break into smaller particles. From initial study on particles it was observed that smaller particles are very angular as compared to bigger particles. This means if there is breakage during shearing; the bigger particles get smaller and hence more angular, which may offer more resistance to shear. However, this needs to be investigated further.



69

VI. FUTURE WORK The height reduction during shearing indicates

deformations and this is important in monitoring and modeling of tailings dams. The reasons for height changes (figure 6) should be studied in detail in order to investigate whether this change is due to breaking of particles or rearrangement of particles. This can be done by analyzing particle sizes and shapes before and after shear, then to develop certain empirical relationships by finding percentage and sizes of particles that are being broken during shearing process. Rearrangement of particles should also be taken into consideration and studied before and after shearing the samples. These are important steps towards deep understanding of typical tailings behaviors and different aspects towards constitutive behavior of tailings.

VII. CONCLUSIONS The grain sizes reduced along with depth from surface of

dam for the tested locations. Initial particles analysis showed that smaller particles of size 0.063mm were very angular, whereas the larger particles of size 1 mm were sub angular. Water content (w) was in range of 15.2-47%. The average particle density (ρs) of collected tailings samples was 2.833 t/m3. The bulk density (ρ) was varying from 1.66–2.06 t/m3. Similarly the saturated density (ρsat) and dry density (ρd) were found to be in range of 1.76 – 2.065 t/m3 and 1.18–1.65 t/m3 respectively. Void ratio (e) and porosity (n) were in range of 0.72–1.41 and 41.9–58.5%. Reductions in vertical height of samples were observed during direct shear tests with slight increase in pore pressures. Future studies should be carried out to understand tailings behavior more in detail.

Acknowledgements This work was carried out at Luleå University of

Technology as a part of a study of the mechanical properties of tailings materials.

The study was initiated and financed by Boliden AB whose support is highly acknowledged. Support was also provided from persons at Aitik mine, Sweden. The support from Luleå University of Technology is highly acknowledged as most of the laboratory studies were carried out in LTU laboratories.

REFERENCES [1] Vanden Berghe J, Ballard J, Wintgens J, List B. Geotechnical risks

related to tailings dam operations. Proceedings Tailings and Mine Waste, Vancouver, BC, November 6 to 9 2011. 2011.

[2] Lars J. Geomechanical properties of tailings: A study of backfill materials for mines. Licentiate Thesis, Luleå University of Technology, Sweden 1990

[3] Vanden Berghe J, Ballard J, Jewell R, Pirson M, Uwe R. Importance of shear stress anisotropy and bottom drainage on tailings dam stability: A case history. Proceedings of the 17th ICSMGE 2009

[4] Lambe, T. William, and Robert V. Whitman. Soil mechanics SI version, John Wiley & Sons, 2008

[5] Rodriguez J, Johansson J, Edeskär T. Particle shape determination by two-dimensional image analysis in geotechnical engineering. Proceedings of Nordic Conference on Soil Mechanics and Geotechnical NGM 2012

[6] Rodriguez JM, Edeskär T, Knutsson S. Particle shape quantities and measurement Techniques–A review, Electronic Journal of Geotechnical Engineering 2013; 18]

[7] Mitchell JK, Soga K. Fundamentals of soil behavior. Third Edition ed. Wiley; 2005

[8] Powers MC. A new roundness scale for sedimentary particles, Journal of Sedimentary Research 1953; 23 (2):117-119.

[9] Jantzer I, Bjelkevik A, Pousette K. Material properties of tailings from swedish mines. Lulea: Norsk Geoteknisk Forening.ICOLD and UNEP. 2001.

[10] Goren L, Aharonov E, Sparks D, Toussaint R. Pore pressure evolution in deforming granular material: A general formulation and the infinitely stiff approximation. Journal of Geophysical Research: Solid Earth (1978–2012). 2010; 115(B9)

[11] Rodriguez JM, Edeskär T. Case of Study on particle shape and friction angle on tailings– Journal of Advanced Science and Engineering Vol 3, No 4 December 2013.373-387

[12] Rodriguez Juan M. "Importance of the Particle Shape on Mechanical Properties of Soil Materials" Licentiate Thesis, Luleå University of Technology, 2013

Paper VIII Bhanbhro, R.; Rodriguez, J.M.; Edeskär, T. and Knutsson, S. (2015). Evaluation of primary and secondary deformations and particle breakage of tailings. Pan-American Conference of Soil Mechanics and Geotechnical Engineering. 15-18 November, 2015. Buenos Aires, Argentina. Ed. Diego Manzanal; Alejo O. Sfriso, IOS Press, 2015, pp. 2481-2488.

Evaluation Of Primary And Secondary Deformations and Particle Breakage

of Tailings Riaz BHANBHROa,1 Juan RODRIGUEZa, Tommy EDESKÄRa and Sven KNUTSSONa

a Division of Mining and Geotechnical Engineering Luleå University of Technology, SE- 971 87, Sweden

Abstract. Tailings are the waste product of mining which is left over after extraction of materials of interest. Tailings material may possess different material properties depending upon type of ore and method of concentration. Sometimes the tailings material itself is used in construction of tailings dams and tailings dams are constructed to withstand for long times. A tailing dam can be exposed to settlements due to incremental load as these dams are raised in stages. Increasing load with time may also lead to particle breakage. This article presents the results from oedometer tests conducted on tailings materials. The study includes the stress-deformation behavior and particle breakage of tailings material of different gradations upon application of incremental loads in oedometer tests. The samples were collected from different sections of tailings dam from Sweden. Remolded samples were manufactured in laboratory as four batches of particle sizes i.e. 1-0.5 mm, 0.5-0.25mm, 0.25-0.125mm and 0.125-0.063mm. The results are analyzed from tested samples at different stress levels and compared with different particle sizes. The breakage of particles of each batch is analyzed by sieving the specimens after oedometer tests. The results are evaluated in terms of primary and secondary deformations. The primary and secondary deformations are also compared with different particle sized specimens.

Keywords. Deformation, Oedometer Test, Particle Breakage, Tailings, Tailings Dams

1. Introduction

Tailings is the waste material from the mining industry. The mining industry produces huge amount of waste material i.e. up to extent of 70-99% of ore as waste material [1]. Tailings dams are built to store the tailings material. Tailings materials itself sometimes are used in construction of tailings dams.

Tailings dams failures demand towards the deep understanding of behaviors of tailings materials under application of loads. Particularly, when upstream construction method is used where tailings material are subject to incremental loads. Several studies (see e.g. [2]-[9]) on tailings material have been conducted and are reported in literature. These studies described the mechanical properties of tailings with focus on strength of tailings.

1 Corresponding Author: PhD Student. Division of Mining and Geotechnical Engineering, Luleå

University of Technology, Sweden; E-Mail: [email protected] , [email protected]

From Fundamentals to Applications in GeotechnicsD. Manzanal and A.O. Sfriso (Eds.)IOS Press, 2015© 2015 The authors and IOS Press. All rights reserved.doi:10.3233/978-1-61499-603-3-2481

2481

In recent studies ([10]-[12]) conducted on material from a tailings dam, unexpected vertical height reductions were observed during shear tests. With these reductions slight increase in pore pressures was also observed during shearing. If pore pressures prevail for long term they may lead to stability issues. A tailings dam might be exposed to deformations/creep behavior due to incremental load when dam is operational and due to constant load upon closure of dam. These deformations can lead to shearing and changes in granular skeleton [13].

Effects of creep are also important towards safety of tailings dams in long-term perspective. The creep in tailings can be defined in terms of secondary compression. The creep can increase on increasing confining pressures and mainly after particle crushing becomes important [14]. Tailings are very angular materials [15] and it is defined by [1] that higher stresses at edges of particles may cause them to break. Change in external stress directly results in change of effective stress, so time-dependent creep can be immediately observed upon application of stress [14]. Sudden effect of effective stress on particles may also cause them to break, and at higher stress it is assumed that granular material will get crushed [16]. It is further defined by [16] that granular material have rapid spreading of strains, the driving force, which results in strains, in these materials gradually reduce along loading axis due to energy dissipation. Several studies by [14], [17], [18] were performed to study creep effects on sand governed by crushing of particles. However, in these studies particle size was not analyzed after the test to prove direct connection to particle breakage.

This article presents results and analysis of stress–strain, secondary compression tests on uniformly graded specimens of tailings material. Sieve analysis was used to study breakage of particles after test. The tailings materials are separated into different range of particles sized specimens and tested in oedometer under different vertical stresses. An attempt has been made to develop a relation of breakage of particles with the size of particles and to analyze stress-strain and secondary compression characteristics of tailings material of different particle sizes.

2. Materials & Methods

The materials used in this research were collected from a copper tailings dam in northern Sweden. The materials were collected undisturbed from the dam. In this study disturbed samples were used. The samples were constructed from uniformly graded material by sieving. Tailings material was sieved and separated to four different particle size ranges i.e. 1-0.5 mm, 0.5-0.25mm, 0.25-0.125mm and 0.125- 0.063mm.

The samples were prepared according to method developed by [19]: Dry tailings material was poured with 5mm nozzle from just above water surface and was left to settle in 2-3 cm under water. Depending on the grain size the particles were allowed to settle in the time range from 30 min to 24 hours. The process was repeated 5-6 times till the full sample tube was obtained. The tubes of 17 cm height and 5 cm diameter were used for preparation of samples. The bottom of tube was closed and tube was placed vertically.

Table 1 shows the description of materials, moisture content, specific gravity, bulk density, initial void ratio and degree of saturation for the tested materials. This description is recorded after preparation of samples in laboratory.

R. Bhanbhro et al. / Evaluation of Primary and Secondary Deformations2482

Table 1. Description of Tailings material used in this study

Material (Particle size-mm)

Moisture Content

average %

Specific GravityAverage

t/m3

Bulk Density average

t/m3

Initial Void Ratio

(e)

Degree of Saturation

(%) 1-0.5 mm 7 % 2.88 1.51 0.98 – 1.12 20%

0.5-0.25 mm 21% 2.90 1.91 0.83 – 0.87 73% 0.25-0.125mm 26% 2.87 2.0 0.76 – 0.85 93%

0.125-0.063mm 29% 2.94 2.04 0.80 – 0.87 99%

Oedometer tests were performed on the prepared samples according to ASTM D2435. Series of loads were applied in incremental stages of 10, 20, 40, 80, 160, 320 and 640 kPa. Each stage of load was applied and then specimen was allowed to consolidate for 24 hours.

3. Results

3.1. Stress-Strain Behavior

The plotted vertical strains are shown in Figure 1 where the results are plotted in form of �� . The strains are plotted for stress interval of 40 – 640 kPa by considering linear portion of line in plot �� . When comparing the vertical strains under same applied vertical effective stresses, it was observed that the specimens of particles size 1-0.5mm showed higher strains. Specimens of particle size 0.25-0.125 mm showed lower strains. Specimens with particles size 0.5-0.25 attained higher strains than particle of size 0.25-0.125 and lower strains than particles of sizes 1-0.5mm. The strains observed in all tests were in the range between 1 and 10%.

Figure 1. Example of results plotted as �� at stress interval of 40 – 640 kPa for the materials of

different particle sizes.

The strains can be well described in form of Eq.1 as it appears straight line in plot�� . Table 2 shows the summary of tests represented in the form of Eq.1 and reductions in void ratios in percentage and final void ratios. The Eq.1 is written as;

�� (1)

log �'v (kPa)10 100 1000

Verti

cal S

train

��

�

1

10

100

1-0.5 mm 0.5-0.25 mm 0.25-0.125 mm 0.125-0.063 mm

40

R. Bhanbhro et al. / Evaluation of Primary and Secondary Deformations 2483

In Eq.1, ��is vertical strain in (%) and �� is vertical effective stress in (kPa). These values are taken from best fit straight line between stress range between 40 kPa and 640 kPa; therefore, the values of�� and�� are valid for this stress ranges only. Table 2. Summary of tests performed in terms of��,�� reduction in void ratio (e) and final void ratios

Material (Particle size range-mm)

� � Reduction in (%) void ratio e

Final Void ratio e

1-0.5 mm 1.795 1.387 1.124 0.496

0.279 0.305 0.632 0.450

23.1% 21.0% 19.0% 15.0%

0.79-0.96

0.5-0.25 mm 1.240 0.957 0.716 0.343

0.227 0.310 0.362 0.480

18.0% 16.9%

0.68-0.72

0.25-0.125mm 0.496 1.922 0.238

0.353 0.212 0.458

18.7% 12.4% 10.1%

0.66-0.78

0.125-0.063mm 1.335 2.815 0.297 0.235

0.226 0.181 0.505 0.433

13.3% 8.8%

0.67-0.77

3.2. Void ratios during compression

The void ratios plotted against effective vertical stress are shown in Figure 2. It was observed that specimens constructed with particles of size 1-0.5mm possessed higher void ratio reductions in relation to initial void ratios as compared to particles with smaller sizes particles i.e. (0.5-0.25, 0.25-0.125 and 0.125-0.063 mm). Also, the specimens of size 1-0.5 mm showed higher reduction in void ratios while application of effective vertical stresses as compared to specimens with smaller particle size. The percentage of reduction of void ratios corresponding to different particle size specimens is shown in Table 2. High value of reduction in void ratio as 23.1% was seen in specimens with particle size 1-0.5mm, whereas, lowest value as 8.8% was seen in specimens with size 0.125-0.063 mm.

Figure 2. Results plotted as�� for the materials of different particle sizes.

log �'v (kPa)

10 100 1000

Void

Rat

io e

0.7

0.8

0.9

1.01-0.5 mm 0.5-0.25 mm0.25-0.125 mm0.125-0.063 mm


3.3. Compressibility and Compression Index

The compressibility can be defined as coefficient of volume compressibility ��and is defined as volume change per unit volume per unit increase in effective stress [20]. The compression index ��is the slope of linear portion of normal consolidation line in the plot �� (see e.g. Figure 2). The coefficient of volume compressibility and compression index can be written as Eq.2 and Eq.3 respectively [20],

� � ��

� !"#!"$!%

&%$&"' (�()*+) (2)

�� !"$!%

,-.#&/%)&/"'

(3)

Where, � is void ratio, � is effective stress and subscripts 0 and 1 represent arbitrary points on the normal consolidation line (i.e. two stress points on consolidation line). The calculated values of coefficient of volume compressibility and compression index are shown in Table 3. These values presented here are calculated for stress range of ��0� � 123�456 and�� 783�456. It is further defined by [20] that � is stress dependent and is valid for that stress range only.

It was observed that ��is proportional to particle size on which specimens are manufactured i.e. specimens, with large particle size and higher initial void ratio, possessed higher values of � and vice versa. Similarly the slope of linear portion in the graph � � �� was also proportional to particle size of specimens i.e. specimens constructed with large particle sizes attained steep slope (see e.g. Figure 2) and vice versa. Table 3. Evaluated parameters for coefficient of volume compressibility and compression index (at ��9� �:;9�<=> and��? � @A9�<=>), coefficient of secondary compression B��

Material Particle size (mm)

C��(C;)DE) BF B�� B��)BF

1-0.5 mm 0.0705 – 0.0863 0.174 – 0.138 8G8 H I3$J 0.025 0.5-0.25 mm 0.0548 – 0.0648 0.101 – 0.121 2G3K H I3$J 0.019

0.25-0.125mm 0.0438 – 0.0564 0.080 – 0.103 1G3 H I3$J 0.036 0.125-0.063mm 0.0288 – 0.0358 0.054 – 0.060 IGL H I3$J 0.032

According to [21], the coefficient of secondary compression (MN!) is defined as the relationship of void ratio and log of time, which is usually linear during secondary compression, and is written as�#�N! � O�/O �� P�). It is further defined by [21] that �N!�are generally related to compression index�� aQ��N!)��.

The secondary compression curves are shown in Figure 3 as void ratio (e) and log time. These typical curves are plotted from consolidation under effective vertical stress of 320 kPa. The coefficient of secondary compression for tested material is represented in Table 3 along with the values of��N!)��.

3.4. Particle Breakage

Possibility of particles breakage was determined by sieving the materials after finishing the test. Results are shown in Table 4. Here particles passing through corresponding sieve are considered as particles that are breaking down. It was observed that larger


particles (1-0.5 mm) showed higher breakage (14.2%) as compared to small sized particles (0.125-0.063 mm) which showed 0.8% particle breakage. Table 4. Particle breaking in percentage determined by sieving after finishing each test

Material (Particle size range-mm)

Percentage of particles passing after test (%)

1-0.5 mm 14.2% 0.5-0.25 mm 10.1%

0.25-0.125mm 12.5% 0.125-0.063mm 0.8%

Figure 3. Typical secondary compression curves for different particle sized specimens plotted as void ratio

vs. log time (min) corresponding to effective stresses of 320 kPa

4. Discussion

In this study, the particles that pass through sieve (e.g. 1-0.5mm particles when pass through 0.5mm sieve after test) are taken as the particles that are broken or may have changed their shape due to high stresses. This is evident that particles are reduced in size from its original as high as 14% in case of specimen constructed with 1-0.5 mm and 0.8% for the specimen of size 0.125-0.063 mm. Particles in skeleton when breakdown, result in a skeleton with more fine contents [22]. As seen in this study that finer particle has less ability to breakdown (i.e. breaking less than 1%). So, it might be possible that breakage due to creep in coarser particles continues till the coarser particles reduce to the size which is less susceptible to breakage i.e. finer particles. Having more fractions of finer particles may reduce long term creep due to particle breakage as finer particles showed less breakage, on the other hand finer particles can give raise to pore pressures that can reduce effective stresses and may lead to failure.

It is reported by [22] that coarser particles are more susceptible to particle breakage and that can be a cause of higher compression in coarser particles. The other reason for higher compression in coarser particles is that specimens were of uniform particle size range i.e. 1-0.5 mm. So, more the coarser particles with higher void ratio,

0.1 1 10 100 10000.697

0.698

0.699

0.700

0.701

0.702

0.703

0.5-0.25 mm

0.1 1 10 100 1000

Void

Rat

io e

0.872

0.874

0.876

0.878

0.880

0.882

0.884

1-0.5 mm

Log t (min)0.1 1 10 100 1000

Void

Rat

io e

0.795

0.796

0.797

0.798

0.799

0.800

0.801

0.802

0.803

0.25-0.125 mm

Log t (min)0.1 1 10 100 1000

0.773

0.774

0.775

0.776

0.777

0.778

0.779

0.125-0.63 mm

Log t (min) Log t (min)


more chances to break and result in compression [21]. The particle breakage is a progressive process that starts at low stress levels due to wide spreading of the amount of interparticle contact forces [21]. There is a possibility that soil grains can break or diminish while creep, resistance of grain contact may reduce and structural bond may also get destroyed [16].

The particle crushing can be one reason for large strains on compression curve [21] as seen in Figure 2 for 1-0.5mm particles. According to [21], field compression for many sands and gravels is as high as 6.5% at 700 kPa; however, field results can vary from laboratory results. In this study, the specimens with coarse particles attained higher compression as 23.1% whereas the compression for specimens with finer particles was 8.8%. The percentage of compression for different particle sizes is shown in Figure 4. Higher compressions are probably due to breakage and use of uniformly graded material. The compressions, mentioned here, are taken in terms of reductions in void ratios.

Figure 4. Compression in percentage (of void ratio) vs. different particle sizes

It is defined by [23] that primary consolidation in the sand tailings is almost impossible to measure at laboratory because it happens very fast. Tailings are more compressible as compared to equivalent natural grain soils due to grading characteristics, method of deposition and high angularity [23]. The secondary compression parameter ��N!)��studied in this study was 0.019 for coarser particles and 0.032 for finer particles. The ��N!)�� value for clean sands reported in literature falls in range of 0.015-0.03 [21]. The secondary compression parameters were observed to be in agreement with what is available in literature.

Results, presented herein, were from tests performed on tailings material of uniformly graded specimens to see effects of breaking of particles for each size. It would be a great addition to perform more tests by constructing specimens with different known percentages of particle sizes and then see effect of breakage of particles. Results can be then optimized and compared with in situ conditions to predict particle breakage.

5. Conclusions

Based on the results from this study it can be concluded that; � The larger vertical strains in specimens made of coarse particles were

observed as compared to specimens with relatively finer particles. � The maximum void ratio reduction (%) after consolidation at 640 kPa for

materials constructed with coarse particles (1 – 0.5 mm) was 23%. Similarly, maximum void ratio reductions for specimens with particle sizes (0.5 – 0.25 mm, 0.25 – 0.125 mm and 0.125 – 0.063 mm) were 18%, 18.7 and 13%

0

5

10

15

20

25

30

Compressio

n�in�%� 1�0.5�mm

0.5�0.25�mm

0.25�0.125mm

0.125�0.063mm

1�0.5��0.5�0.25��0.25�0.125 0.125�0.063(mm)��(mm)��(mm)��(mm)�


respectively. This means higher void ratio reductions were observed in coarser particles specimens.

� The maximum coefficient of volume compressibility ��for (1 – 0.5 mm, 0.5 – 0.25 mm, 0.25 – 0.125 mm and 0.125 – 0.063 mm) specimens was 0.086, 0.064. 0.056 and 0.035 respectively. Whereas ��N!)��was found to be 0.025, 0.019, 0.036 and 0.032 respectively.

� Particle breakage in coarser particles was high (14%) as compared to finer particles (0.8%).

6. References

[1] I. Jantzer, A. Bjelkevik and K. Pousette, "Material properties of tailings from swedish mines," Lulea: Norsk Geoteknisk Forening.ICOLD and UNEP 2001.

[2] G. Blight and G. Bentel, "The behaviour of mine tailings during hydraulic deposition," J.S.Afr.Inst.Min.Metall., vol. 83, no. 4, pp. 73-86 1983.

[3] G. Blight and O. Steffen, "Geotechnics of gold mining waste disposal," Current geotechnical practice in mine waste disposal, pp. 1-53 1979.

[4] H.K. Mittal and N.R. Morgenstern, "Parameters for the design of tailings dams," Canadian Geotechnical Journal, vol. 12, no. 2, pp. 235-261 1975.

[5] Y. Qiu and D. Sego, "Laboratory properties of mine tailings," Canadian Geotechnical Journal, vol. 38, no. 1, pp. 183-190 2001.

[6] P. Guo and X. Su, "Shear strength, interparticle locking, and dilatancy of granular materials," Canadian Geotechnical Journal, vol. 44, no. 5, pp. 579-591 2007.

[7] A. Shamsai, A. Pak, S.M. Bateni and S.A.H. Ayatollahi, "Geotechnical characteristics of copper mine tailings: A case study," Geotech.Geol.Eng., vol. 25, no. 5, pp. 591-602 2007.

[8] R. Volpe, "Physical and engineering properties of copper tailings," Proceedings of Current Geotechnical Practice in Mine Waste Disposal.Edited by The Committee on Embankment Dams and Slopes of the Geotechnical Engineering Division.ASCE, New York, pp. 242-260 1979.

[9] H. Chen and D. Van Zyl, "Shear strength and volume-change behavior of copper tailings under saturated conditions," Geotech Spec Publ, pp. 430-430 1988.

[10] R. Bhanbhro, "Mechanical properties of tailings basic description of a tailings material from sweden," Licentiate thesis, Luleå tekniska universitet. Licentiate thesis / Luleå University of Technology 2014.

[11] R. Bhanbhro, R. Knutsson, T. Edeskär and S. Knutsson, "Mechanical properties of soft tailings from a swedish tailings impoundment: Results from direct shear tests," Electronic Journal of Geotechnical Engineering (EJGE), vol. 19, no. Z, pp. 9023-9039 2014.

[12] R. Bhanbhro, R. Knutsson, J.M. Rodriguez, T. Edeskar and S. Knutsson, "Basic description of tailings from aitik focusing on mechanical behavior," International Journal of Emerging Technology and Advanced Engineering, vol. 3, no. 12, pp. 65-69 2013.

[13] R. Knutsson, A. Bjelkevik and S. Knutsson, "Importance of tailings properties for closure," in Mine Closure Solutions 2014 Conference, apr. 26 2014 - apr. 30 2014, Ouro Preto, Minas Gerais, Brasilien.

[14] P.V. Lade and C. Liu, "Experimental study of drained creep behavior of sand," J.Eng.Mech., vol. 124, no. 8, pp. 912-920 1998.

[15] J.M. Rodriguez and T. Edeskär, "Case of study on particle shape and friction angle on tailings," Journal of Advanced Science and Engineering Research Vol, vol. 3, no. 4, pp. 373-387 2013.

[16] J. Feda, "Irregular creep in granular materials," Acta Technica CSAV, vol. 48, no. 4, pp. 395-410 2003. [17] P.V. Lade, C.D. Liggio Jr and J. Nam, "Strain rate, creep, and stress drop-creep experiments on crushed

coral sand," J.Geotech.Geoenviron.Eng., vol. 135, no. 7, pp. 941-953 2009. [18] P.V. Lade, J. Nam and C.D. Liggio Jr, "Effects of particle crushing in stress drop-relaxation

experiments on crushed coral sand," J.Geotech.Geoenviron.Eng., vol. 136, no. 3, pp. 500-509 2010. [19] R. Dobry, "Soil properties and earthquake ground response," in Proceedings of the10th European

Conference on Soil Mechanics and Foundation Engineering, Vol. 4., Florence, Italy, 1991. [20] R.F. Craig, Craig's Soil Mechanics (7th Edition). London, GBR, CRC Press, 2004. [21] J.K. Mitchell and K. Soga, Fundamentals of soil behavior, Third Edition ed., Wiley, 2005. [22] K.L. Lee and I. Farhoomand, "Compressibility and crushing of granular soil in anisotropic triaxial

compression," Canadian Geotechnical Journal, vol. 4, no. 1, pp. 68-86 1967. [23] S. Vick, "Planning, design, and analysis of tailings dams, BiTech," Publishers, Vancouver 1990.


Effect of Physical Weathering on - DiVA...

Documents

Transcript of Effect of Physical Weathering on - DiVA...