Triaxial shear strength characteristics of some sands ...

For Reference

NOT TO BE TAKEN FROM THIS ROOM

0X IIBEiS

awsinsraJis HiMOSTOMlS

University

Digitized by the Internet Archive in 2018 with funding from

University of Alberta Libraries

https://archive.org/details/triaxialshearstrOOalbe

UNIVERSITY OF ALBERTA

TRIAXIAL SHEAR STRENGTH CHARACTERISTICS OF SOME SANDS

STABILIZED WITH FOAMED AND EMULSIFIED ASPHALTS

BY

ALBERT DANIEL LAPLANTE

A THESIS

SUBMITTED TO THE FACULTY OF GRADUATE

STUDIES IN PARTIAL FULFILMENT OF THE

REQUIREMENTS FOR THE DEGREE OF

MASTER OF SCIENCE

DEPARTMENT OF CIVIL ENGINEERING

EDMONTON, ALBERTA

JUNE, 1963

i

ABSTRACT

One aspect of the thesis dealt with the triaxial shear strength

characteristics of a poorly-graded sand stabilized with either a foamed

or an emulsified asphalt. A comparison of shear strengths obtained

shows the superiority of the cured foamed asphalt specimens; however

immersion results in very low and similar cohesive strengths to the

emulsified asphalt specimens.

A second aspect of this thesis was to investigate the effect of

varying gradations on shear strengths at the same estimated bituminous

film thickness. This aspect shows the beneficial effect of adding

fines to a poorly graded sand. The increased cohesive strengths with

increasing fines is believed to be caused by increasing ratios of

contact area to surface area and by increasing viscosities of the

filler-asphalt binder.

A third aspect dealt with the effect of alternate cycles of curing

and wetting on shear strengths and on volume changes of foamed asphalt

specimens. This aspect shows that alternate cycles of curing and

wetting do not significantly affect the final cured shear strength of

foamed asphalt specimens; also measured volume changes are inconclusive

and are believed inadequate to substantiate the postulation of

capillary tension in the asphalt phase of foamed asphalt specimens.

Finally, the altered University of Alberta foamed asphalt laboratory

installation was proved to be adequate to foam asphalt.

,i ;:i ■-: m a-' 1 '•

■

■ ' •: ' . "■ ■ -• , jyv btiv

ii

AC KN OV'LEIXTEMENTS

The author wishes to extend his appreciation to the following for

their assistance in preparation of this thesis:

K.O. Anderson, Associate Professor of Civil Engineering, for his

constructive criticism in carrying out the testing program and in

preparation of the report.

R.C.G. Haas, fellow graduate student, for his generous help and

advice in carrying out the testing program. The author is also

thankful to R.C.G. Haas for the lending of his computer solution of

Palmer's determination of the Mohr envelope based on least squares.

• . - . . > ■ ■.A .0. -f

rid? .

iii

TABLE Qi‘‘ CONTENTS

PAGE

ABSTRACT . i

ACKNOWLEDGEMENTS. ii

TABLE OF CONTENTS. iii

LIST OF TABLES. vii

LIST OF FIGURES. viii

LIST OF PLATES. x

GLOSSARY OF TERMS AND SYMBOLS. xi

CHAPTER I INTRODUCTION. 1

General . 1

Purpose and Limitations of the Investigation ... 2

Organization of the Thesis . 4

CHAPTER II A LITERATURE REVIEW OF SOME FACTORS AFFECTING THE

SHEARING STRENGTHS OF BITUMINOUS-SOIL MIXTURES. . . 5

General . 5

Influence of Film Thickness on Shear Strengths... 6

Influence of Type of Aggregate and Filler on

Shear Strengths. 8

Influence of Type of Asphalt on Shear Strengths . . 11

Mechanism of Foamed Asphalt . 13

Theory of Foamed Asphalt . lU

CHAPTER III CLASSIFICATION AND ANALYSIS OF MATERIALS . 17

Sand. 17

Asphalt. 18

iv

TABLE OF CONTENTS (Continued)

CHAPTER IV THE TESTING PROGRAM . 22

Outline of Testing Program. 22

Preparation of Batches and Specimens . 26

Apparatus and Testing Procedures for Triaxial Tests . 26

Discussion of Testing Procedures and Apparatus ... 26

General. 26

Density gradients and compaction technique ... 26

Size of specimen. 27

Mixing time. 27

Grain size distribution. 33

Gradation effects on shear strengths compared at

the same estimated film thickness. 33

Mixing and molding water content for emulsion

asphalts. 3^

Curing of specimens. 35

Immersion of specimens. 36

Length and diameter measurements of specimens . . 36

Triaxial test procedures . 36

CHAPTER V SUMMARY AND DISCUSSION OF RESULTS.. . 39

General. 39

Statistical Evaluation of Test Results . 39

Strength of Foamed Asphalt Stabilized Sand ..... ^0

Series ia.. ho

Series IB. 69

V

TABLE OF CONTENTS (continued)

Series 1C. 71

Gradation Effects on Shear Strengths Compared at the

Same Estimated Film Thickness. 72

General. 72

Series 2A Gradation effects on shear strengths . . 73

compared at an estimated film thickness of 5-4 microns

Series 2B Gradation effects on shear strengths . . Jb

compared at an estimated film thickness of 2.7 microns

Comparison of Gradation Effects on Shear Strengths at

Estimated Film Thicknesses of 2.7 and 5*4 Microns . . 75

Strength of Emulsion Asphalt Stabilized Sand . 77

Series 3. 77

Strength Comparisons . 78

Variation of Cohesion with Percent Asphalt, Filler to

Asphalt Ratio, Percent Passing No. 200 Sieve, and

with Total Dry Unit Weight.8l

General.8l

Variation with percent asphalt . 8l

Variation with filler to asphalt ratio . 82

Variation with percent passing No. 200 sieve .... 82

Variation with total dry unit weight.83

Variation of Angle of Internal Friction with Percent

Asphalt, Filler to Asphalt Ratio, Percent Passing

No. 200 Sieve, and with Total Dry Unit Weight ... 84

Stress-Strain Relationships.*85

TABLE OF CONTENTS (Continued)

CHAPTER VI CONCLUSIONS AND RECOMMENDATIONS . 86

Conclusions. 86

Recommendations. 88

BIBLIOGRAPHY . 90

APPENDIX A SAND CLASSIFICATION DATA. A-l

APPENDIX B PROCEDURE USED TO OBTAIN DIFFERENT GRAIN SIZE

DISTRIBUTIONS . B-l

APPENDIX C MEASUREMENT OF SURFACE AREAS OF SOIL PARTICLES .... c_1

APPENDIX D DETAILS OF PREPARATION OF BATCHES, MOLDING OF

SPECIMENS AND OF APPARATUS AND TESTING PROCEDURES

FOR TRIAXIAL TESTS. D-l

APPENDIX E SAMPLE SHEETS. E-l

vii

LIST OF TABLES

TABLE PAGE

I Sand Properties 19

II Calculated Grain Size Distribution of Sands 20

III Analysis of 150-200 Penetration Asphalt Cement 21

IV Analysis of SS-1 Emulsified Asphalt 21

V Outline of Testing Program 23

VI Summary of Test Results ^3

viii

LIST OF FIGURES

FIGURE PAGE

1(a) Cross Section of Foamed Asphalt Nozzle 42

1(b) Diagram of Foamed Asphalt Equipment used by Haas 42

2 Strength Envelopes for Foamed Asphalt Series with Varying Molding Water Content 44

3 Variation of Cohesion with $ Passing No. 200 Sieve or Total Dry Unit Weight 45

Water Absorption Curves for Foamed Asphalt Mixtures with Varying Molding Water Contents 46

5 Strength Envelopes to Illustrate Effect of Slight Gradation Variations 47

6 Strength Envelopes to Illustrate Effect of Slight Gradation Variations 48

7 Strength Envelopes to Illustrate Effect of the Steam Jet Setting 49

8 Strength Envelopes for Varying Number of Wetting and Curing Cycles 50

9 Volume of Foamed Asphalt Specimens After Alternate Cycles of Curing and Soaking 51

10 Cured 4.2% Foamed Asphalt Strength Envelopes for BCH and U. of A. Pugmills 52

n Strength Envelopes to Illustrate Effect of Gradation on Cured Strengths Compared at same Film Thickness 53

12 Variation of Total Dry Unit Weight with 'jo

Passing No. 200 Sieve or jo Asphalt 54

13 Variation of Cohesion with Jo Asphalt or

Filler to Asphalt Ratio 55

14 Variation of Angle of Interna] Friction with Jo

Asphalt or Filler to Asphalt Ratio 56

15 Variation of Angle of Internal Friction with Jo

Passing No. 200 Sieve or Total Dry Unit Weight 57

.'"fl' ' ; • -W • • ", ,11:

•

ix

LIST OF FIGURES (Continued)

FIGURE PAGE

16 Strength Envelopes to Illustrate Effect of Gradation on Cured Strengths Compared at same Film Thickness 58

IT Strength Envelopes for Varying Film Thickness at same Gradation 59

18 Strength Envelopes for Varying Film Thickness at same Gradation 60

19 Strength Envelopes for Varying Film Thickness at same Gradation 6l

20 Strength Envelopes for Emulsion Asphalt Series with Varying Molding Water Content 62

21 Strength Envelopes for Varying Type of Aspahlt and Molding Water Content 63

22 Strength Envelopes for Varying Asphalt Type at Approximately Same Molding Water Content 6k

23 Strength Envelopes for Varying Asphalt Type at Same Unit Dry Weight 65

2k Strength Envelopes for Varying Asphalt Type at Approximately Same Molding Volatile Content 66

25 Variation of Cohesion with Total Dry Unit Weight or with $ Asphalt 67

LIST OF FLATES

PLATE PAGE

1(A) U. of A. Pugmill 28

l(B) Compaction Equipment 28

2 Extrusion of 2 Inch by 1+ Inch Sand-Asphalt Specimens 29

3(A) Measuring Specimen 30

3(B) Setting Up Specimen 30

!+ Triaxial Testing of 2 Inch Diameter by k Inch Sand-Asphalt Specimens 31

5(A) Sta-Warm Asphalt Heater and Foamed Asphalt Container 32

5(B) U of A Foamed Asphalt Laboratory Installation 32

GLOSSARY OF TERMS

Cohesive strength - is the shearing strength at zero normal load on the Mohr diagram. In the thesis, this is sometimes referred to as cohesion.

Filler - material finer than the No. 200 sieve.

Curing - is a drying out of volatiles by heat. In this investigation curing refers to 7 days of drying in a 100°F oven.

Soaking - in this investigation soaking refers to lb days of immersion in room temperature water at 70 to 75°F*

Asphalt content - in this investigation, this is referred to on the basis of oven dry weight of the sand.

Foamed asphalt - is the combination of steam and asphalt in a special nozzle which results in an expanded asphalt.

Grain size distribution - is the distribution of grain sizes; in this thesis this is sometimes referred to as gradation.

X ~ shearing stress in pounds per square inch.

cr ~ normal stress in pounds per square inch.

m.~ is the estimated thickness of asphalt films in microns. This estimated film thickness is equal to the volume of asphalt per pound of sand divided by the estimated surface area per pound of sand in square feet.

Total dry unit weight - is the dry unit weight based on both mineral and asphalt residue.

CHAPTER I

INTRODUCTION

General

Because of the ever-increasing need for primary highways; high

quality gravel sources in many parts of North America have become

scarce. Many highway agencies in the United States were faced with

this problem 20 to 30 years ago, while Alberta road builders became

acutely aware of the situation in the early 1950's. Because of these

shortages, the Alberta Department of Highways has stabilized local

materials of inferior quality so as to improve their ability to

resist deformation under loads and to resist the adverse effects of

the weather.

To date, Portland Cement has been mainly used to stabilize these

inferior soils. The current practice of the City of Edmonton and the

Department of Highways has been to use soil cement predominantly as a

base course with 2 to 4 inches of wearing surface to protect the soil

cement from the abrasive action of traffic. Besides the cementing agents

such as Portland cement, lime and lime-pozzolan, other possible stabilizing

agents include wTater emulsion asphalts and cutback asphalts. Still

another and probably the most promising agent is foamed asphalt in

which the viscosity of the asphalt cement is altered by the introduction

of steam into the asphalt.

Foamed asphalt though still quite new locally and proven only

to a certain degree in the United States is becoming more and more

the object of research because it can be added to unheated in-situ

soils. Thus it is a promising method as compared to conventional

hot-mix stabilization.

Preliminary laboratory testing (Csanyi (1957)), the originator

and foremost authority on foamed asphalt, has indicated that various

sands stabilized with a 150-200 penetration foamed asphalt cement,

had superior cured strengths and were more resistant to the destructive

action of freezing and thawing and to stripping due to water than the

same mixes prepared with the binder in an atomized form.

Purpose and Limitations of the Investigation.

This investigation is part of a continuing series of investigations

on stabilization of highway materials sponsored by the Cooperative

Highway Research Program in Alberta. Most of the past work within this

program dealt with stabilization using Portland cement, lime, lime-

flyash and lime-pozzdLan stabilization. Recent work however, has dealt

with stabilization of inferior soils with asphalts and asphalt-lime

mixtures. Recent work by Haas (1963) dealt with the strength relation¬

ships of a uniform, fine-grained sand stabilized with a cutback, asphalt

and as asphalt cement. In conjunction with Haas's work and immediately

before it, other work on sand-asphalt stabilization has been conducted

by Knowles (1962) on strength relationships of emulsified asphalt

stabilized sands with lime additives, by Jones (1962) on the durability

characteristics of cutback asphalt stabilized sand, by Pennell (1962)

on the comparison of structural strengths of a Portland cement stabilized

sand with a cutback stabilized sand, and by White (1962) on aeration of

cutback asphalt stabilized sand.

After a close examination of previous work, it was decided to

stabilize the same sand used by White, Jones, Pennell and Haas with

still another type of asphalt; that is, SSI water emulsion asphalt.

Also, due to the limited investigation by Haas (1963) on foamed

asphalt, research in the use of foamed asphalt seemed warranted.

Because of the many variables affecting the strength of sand-

bituminous mixtures, it was possible to investigate only a limited

number of variables. Variables such as quantity and type of asphalt,

film thickness, type of aggregate, the size, shape and surface texture

of particles, grain size distribution, degree of mixing and finally

the amount of mixing and molding water illustrate that it is not

possible within the scope of a single thesis to examine the effect of

most of these variables on shear strength. Thus, because of these

many variables affecting shear strength, it was decided to examine

only the effects of mixing water content, molding water content, film

thickness and grain size distribution on shear strengths of sand-foamed

asphalt mixtures; also the effects of various molding water contents

were examined on water emulsion sand mixtures and the strengths obtained

were compared to those obtained for the foamed asphalt mixtures.

As previously mentioned, because of the many variables affecting

shear strength of sand-bituminous mixtures, it was decided to make

constant as many variables as possible that were not directly involved

in comparison of the various variables. In accordance with this,

curing conditions and water immersion conditions were the same for all

specimens of the various mixtures; also compactive effort was kept

constant throughout. Various other conditions of procedure and testing

were kept as consistent to those used by previous investigators at the

University of Alberta.

Organization of the Thesis.

The second chapter is a brief review and commentary on literature

concerning asphalt stabilization. Particular emphasis has been placed

on the effect of grain size distribution on shear strengths and on

foamed asphalt stabilization.

Chapter III is the classification and the analysis of the materials.

Chapter IV describes and outlines the testing program carried out.

A detailed procedure of specimen preparation and tests is given; also

a discussion of the testing program procedures is included.

Chapter V contains the results of the laboratory testing. Results

are presented both in tabular and graphical form and a detailed

discussion is included.

Chapter VI contains a brief summary of conclusions and recommendations

made from the actual test results and from the literature review.

The appendices follows a section on Bibliography and generally

contains detailed test procedures, routine information on classification

of the sand and other minor information of insufficient important to

incorporate into the main body of the thesis.

CHAPTER II

A LITERATURE REVIEW OF SOME FACTORS AFFECTING

THE SHEARING STRENGTHS OF BITUMINOUS-SOIL MIXTURES

General

Many stabilizing agents are used today either to increase the

bearing capacity of soils by cementing or to protect the existing

bearing capacity of soils by waterproofing. Among these the chief

stabilizing agents are asphalts, cements and limes. Mainly because

of their excellent waterproofing and cementing abilities, asphalts

are widely used as stabilizing agents in surface courses, base

courses and in sub-bases.

According to Endersby (19^2)* asphalt can either plug the

capillaries in the soil mass thus preventing the water from entering

or leaving the mass or it can coat the particles with a thin film of

asphalt, thus preventing water from attacking the individual soil grains.

In the case of cohesionless soil as used in this thesis, it is believed

that a thin film of asphalt coats the individual particles and adds

cohesion to the compacted mix. The magnitude of cohesion depends on

the viscosity and the thickness of asphalt film . It has been shown

(Csanyi (19^8)) that the shearing resistance of a viscous film between

two solid soil particles depends on the thickness of the film. Further¬

more, it has been shown by Csanyi and by Sommer (1957) that the type of

process used materially affects the asphalt distribution and thus the

*References are cited by indicating the author and the year of publication The references are contained in the bibliography at the conclusion of the thesis.

* '-HO'/'j: mr \ni iq £& ft q;.*rfc ■

' '

6

uniformity and thickness of films; also these same two authors showed that

the type of process materially affects the continuity of films, the

theory being that as the film becomes more continuous the strengths are

increased. Both showed that the impact or atomized process forms .

thinner, more uniform and continuous films than the conventional process

of spreading asphalt directly into the mixer.

Influence of Film Thickness on Shear Strengths.

A bituminous-sand mixture should have a film as thin as possible

without materially affecting durability and hardening effects. It has

been shown (Csanyi (19^8)) that of all the factors affecting film

thickness, the most significant ones are the adhesion between the

aggregate and the binder, the temperature of both the binder and the

aggregate, the viscosity, the cohesion and the surface tension of the

binder at the temperature of the mixture , the quantity of binder used

and the degree of mixing.

Many more sources indicate that the cohesion or the shearing resistarn e

of an asphalt film depends on the thickness of the film. Goetz (195l)

showed that as the asphalt content (film thickness) increased,the stability

of the mix increased to a maximum and then decreased. Using the triaxial

test as a measure of shear strength of bituminous aggregate mixes, he

found the cohesion to increase to a maximum and then to decrease; on the

other hand, however, he found the angle of internal friction to decrease

with an increase in asphalt content. Douglas and Tons (1961), using the

Smith Triaxial test as a measure of shear strengths of bituminous mastic

concretes, found cohesive strengths of 103 and 25 psi for estimated film

7

thicknesses of 5 end 7 microns respectively. For the same estimated film

thicknesses., friction angles of l6 and 10 degrees respectively were obtained.

Mack (1957) demonstrated optimum film thickness using a tension test

with metal plates cemented together with bitumen. His work showed optimum

film thicknesses varying from k.5 to 7*0 microns for different binders;

also, he found the optimum film thickness to increase with increasing

viscosity of the binder. When these same asphalt cements were used in

a sand-filler mixture, the calculated film thickness between sand

particles corresponded very closely at optimum stability with the optimum

film thickness determined by the tension tests. The calculated film

thickness between sand particles included both filler and bitumen in the

film, and led Mack to indicate that filler becomes part of the film.

Generally, it can be said that a film that is too thick merely

lubricates the aggregate particles; on the other hand, too little asphalt

will not coat all of the particles and thus may decrease the stability

and durability.

Work done by other authors has shown stability and durability of

mixes to be related to the voids in the mixture. McLeod (1959) has shown

asphalt aggregate mixture to be durable and stable with an air void

content of 3 to 5 percent. Still other authors have correlated film

thickness to this air void requirement. Work done by Griffith and

Kallas (1958) on minimum aggregate air voids versus surface area of

aggregates showed film thickness requirements to be 5*3 to 13-5 microns

for available surface areas of l8.0 to 72.0 square feet respectively per

pound of aggregate. Lefebvre (1957) showed the film thickness requirements

'

:'>v bi. • )Cj s >o .3

: f 1: lI Jt -V >j : 'Jv :• ;■ t ::j; “■ J .

; - '

...

8

for ^4 percent air voids to be 3*0 to 18-5 microns for available surface

areas of 15 and 57 square feet respectively per pound of aggregate.

Influence of Type of Agrrenate and Filler on Shear Strengths.

Besides the effects of film thicknesses on stability of mixes it

has been known for a long time that the type of aggregate used affects

the shearing strength of mixes. Herrin and Goetz (195^0 showed that the

shape of the aggregates used affected the compressive strengths of asphalt-

aggregate mixtures. They showed that at all lateral pressure used in the

triaxial test, the compressive strength becomes greater with an increase

in the angularity of the aggregate. Holmes (19^7) using the Hubbard-

Field test as a measure of stability and using sand, crushed granite,

crushed traprock, crushed limestone and varying amount of Venezuelan

base RC-1 cutback asphalt as materials, found that there appeared to be

a critical range of the mean particle size at which changes in the

trends of density, stability, water absorption and volumetric swell

occurred. For each property it was found that the critical range was

different but that a relatively small range covered all of the four

properties. Holmes also found that the mineralogical nature of the

aggregate exerts a pronounced effect upon the strengths of compacted

asphalt-aggregate mixtures. In his investigations, using the Hubbard-

field test as a measure of stability, he found sand containing impure

Silica led to the least strength and that the limestone (calcium carbonate)

had the highest strength. Granite containing silica-feldspar and the

crushed traprock containing Meta-Morphic Basalt were found to be better

than the sand but worst than the limestone in contributing to shearing

• ' 5‘ .Jf.

9

strength of the asphalt-aggregate mixtures.

Rice and Goetz (1949) in their investigation concerning the suitability

of Indiana dune, lake and gravel-pit sands as aggregates for low-cost

sand-bituminous pavement mixtures found that particle size, shape, texture

and grain size distribution affected both the cured and immersed compressive

strengths. Also they found that fillers affected strengths to varying

degrees depending on the amount and types used. In Rice and Goetz's

investigation 4 types of sands were used to determine the influence of

the characteristics of the sands on the properties of asphalt mixtures

made from them. To isolate the variable of asphalt content, a constant

amount of RC-2, 3 percent by weight, was used. The after-cured residual

asphalt content was 2.4 percent.

This investigation showed that surface texture of grains to be

quite important. It showed that even when particle size distribution

and particle shape are the same, compressive strengths can vary consider¬

ably. It showed that surface texture factors such as cleanliness and

smoothness, influence the strength either by promoting better adhesion

or more friction or both. This investigation also showed that the

most significant characteristic of the sand was the amount passing the

No. 200 sieve. Unconfined compressive strengths were found to increase

as the percentages passing the No. 200 sieve increased. It was also

found that the sands containing the greatest amount of fine material

had the highest absorption and percent loss in strength upon immersion

in water; however, the final immersed strengths were found to increase

with increasing fines.

. v-e-,7 hna drx ;0#ts Bti'S ;o as>s ;...

‘ y ‘ • yj ■ '••• . - .v ■' : A

10

Furthermore, this investigation showed that increases in asphalt

content were accompanied by greater density of the sand in the mixture,

and less water absorption and percentage loss in strength upon

immersion. The water adsorption was found to increase with increasing

void ratios, the percent voids filled with water remained more or less

constant for each sand. Finally this very excellent investigation

showed that adding fine material generally increased the compressive

strengths of mixes. It was found that the optimum asphalt content

varied for various filler to binder ratios. It was also found that as

the percentage of filler used increased, the optimum asphalt content

increased; also as percentage of filler increased the percentage

strength loss due to water immersion increased. Finally, it was found

that the type of filler used affected the strength. It was found that

the mixtures made with "active" fillers resisted the action of water the

best, though the clay mixtures evidenced some swell.

Pauls and Goode (19^7) found that:

"Mixtures containing clay as the filler are similarly unsatisfactory when subjected to moisture although some mixtures, when dry, have higher strength than mixtures containing limestone dust."

Invesgitations on the effects of fillers on strengths have been

carried out by many others besides the above-mentioned authors; however,

for the purpose of this thesis a brief and pertinent summary follows.

Endersby and Vallerga (1952) found that an increased surface area

can result in increased contact areas between particles thus

influencing the strength characteristics of bituminous concrete.

‘ -ii "J • • ti .

McLeod (195M found the adhesion and the selective adsorption of

some components from the asphalt to be determined to a great degree by

the surface chemistry of the filler.

Griffith and Kallas (1958) found the effect of filler particle

shape on the stability of mixes to be greatest when the fine fraction

(minus No. 8 sieve size) is between kO and 60 percent of the total

weight of the mix. Kallas, Puzinauskas and Krieger (1962) found

stability of mixes to increase as the amount of filler used increased.

Also they found the stability of paving mixtures to increase with

increasing viscosity of the filler asphalt binder; viscosity, in turn,

was found to increase with increasing filler to binder ratios.

Stewart (1962) in his investigations at Queen's University, with

various untreated sands, found stability of sand-emulsion mixtures

to increase with percent fines and to be dependent on the Bagnold Peak

diameter. The Bagnold relationship (peak diameter included) is a method

of expressing the grading of sands. Besides the peak diameter, the

slopes of the frequency curves for sizes finer and coarser are character¬

istics of the particular sands used. Because this peak diameter and the

slopes of the frequency curves define the complete grading of a sand,

it is considered advantageous to those values that define the sand

characteristics at only several isolated points. Stewart (1962) in his

investigation with various untreated sands, found the angle of internal

friction to increase as the Bagnold peak diameter increased-

Influence of the Type of Asphalt on Shear Strengths.

Besides film thickness, aggregate shape, texture and grain size

; -■ V ■

12

distribution, it has been shown by Borgfeldt and Ferm (1962) that the

use of cationic asphalt emission can result in higher strengths than

aggregates stabilized with an anionic asphalt emulsion. They showed the

importance of emulsion electrical charge for the adhesion of emulsified

asphalt to mineral aggregates- Also they reported that the use of cationic

(positively charged) emulsions is superior to anionic (negatively charged)

emulsions on mineral aggregates containing silicate-type minerals which

become predominantly negatively charged v/hen moist.

They also showed the superiority of cationic emulsion over anionic-

emulsion for stabilization of a well-graded 3A inch crushed gravel

containing a considerable amount of clay. Stabilization of a sandy silt

using a cationic emulsion and rapid and medium curing cutback asphalt shoved the

cured strengths of the cationic emulsion mixtures to be superior to the

cutback mixtures.

Goetz (1951) showed that as the penetration of asphalt cements

increased the cohesion of asphalt-aggregate mixtures, as measured by the

triaxial test- decreased. He also showed that the viscosity of the

asphalt had little or no effect on the angle of internal friction.

Work done by Knowles (1962) on the stabilization of a poorly-graded

sand■ with a SSI water emulsion asphalt and 150-200 penetration asphalt

showed that there was no significant difference on either the cured,

the freeze-thaw or the water-immersed compressive strengths. He

showed, however that the SSI mixture was more sensitive to the addition

of 3 percent lime than was the mixture with 150-200 penetration asphalt.

Work done by Haas (1963) showed that the triaxial shear strengths

depended on the type of asphalt used and on the process. He showed

that the stabilization of a poorly-graded sand using an MC3 cutback

asphalt was superior to stabilization using a hot mix 150-200 asphalt

cement. He also showed that stabilization with the foamed asphalt

process was superior to the conventional hot mix process; and that

cured strengths of foamed asphalt-sand mixtures were superior to MC3

cutback asphalt-sand mixtures. Haas postulated that the superiority

of cured foamed asphalt strengths was due either to more uniform

dispersion of the asphalt or to capillary tension in the asphalt phase.

Haas's second postulation of capillary tension in the asphalt phase

was based partly on measured volume increases of foamed asphalt

specimens upon immersion. He postulated these volume increases to be

caused by the release of the capillary tension in the asphalt phase.

Because of the apparent significance of volume changes, more

information on volume changes with alternate cycles of curing and

soaking seemed warranted.

Mechanism of Foamed Asphalt.

According to Csanyi (1957) the physical properties of viscosity

and surface tension of bitumen are altered by steam introduced through

a special nozzle (see Figure l). The steam reduces the viscosity and

increases the "wetting" action of the asphalt on the aggregate during

the mixing process. According to Csanyi (1957):

"Adhesion between the film of binder and an aggregate particle depends largely upon the surface tensions of the two materials and the interfacial tension developed. The surface tension of the binder must therefore be such, in relation to that of the aggregate,

that surface moisture on the aggregate can be displaced and a strong physical bond generated between the binder and the aggregate."

The author of this investigation believes that there is no doubt

about the "increased wetting power" because of the ready-made thin films

of asphalt cement with stored energy available to assist in coating

particles as the bubble breaks. However, more information on surface and

interfacial tensions of the binder and the aggregate seems warranted to

substantiate Csanyi's theory of displacement of surface moisture on the

aggregate. Csanyi showed foamed asphalt to have higher penetration

(lower viscosity). Thus distribution of a binder throughout an aggregate

can be accomplished at a much lower temperature than if the binder is

added as a liquid. A foamed asphalt has a rubbery nature and is

extremely sticky, with both high cohesive and high adhesive properties.

When a foamed asphalt is used as a binder in a bituminous mix, it is

reasonable to expect improved adhesion between binder and aggregate.

Theory of Foamed Asphalt

It is believed that foamed asphalt when added in proper quantities

and added to proper materials can lead to the formation of bituminous

mastic concrete.

According to Douglas and Tons (1961):

"A bituminous mastic concrete is an aggregate asphalt mixture in which the binder is a mastic holding the aggregate together. This mastic consists of a high percentage of mineral filler dispersed uniformly throughout the asphalt cement."

Upon microscopic comparison of a conventional asphalt concrete to a

bituminous mastic concrete, Douglas and Tons found that the conventional

0 ■' V S'-.

■

■

15

concrete was open and showed voids while the mastic concrete was dense,

smooth and next to voidless. The asphalt binder in the conventional mix

appeared shiny and plentiful whereas the mastic binder was dull in

appearance.

As postulated by Douglas and Tons shear strength of bituminous

mastic concretes come from four separate things. These are:

(l) From the ability of the asphalt binder to shear or flow.

The author believes that Sommer*s (1957) theory on the flow of thin

bituminous films applies. This theory by Sommer points out that the

flow of thin bituminous films follows the Law of Poisseuille related

to capillary flow where:

vr ir.p>cr . ~&Lrru

Where: Vs Volume of fluid flowing through the capillary per second

p= pressure head over length L

L - length of capillary

7^* absolute viscosity

ascapillary radius

This equation shows that the tendency to flow is inversely

proportional to the absolute viscosity and proportional to the fourth

power of the capillary radius.

(2) From the very thin adsorbed layer which adds rigidity to the

mix. It is thought here that when films are of the same thickness as

adsorbed films, it is believed that these films behave more rigidly

than thicker films and thus an apparent internal friction could be

developed which would not be apparent in conventional mixes where the

i 7.r;r, v;. rlS{2

-

,

'

16

films are thicker.

(3) From the ratio of contact area to the surface area of the

particle assumed to be spherical for calculation purposes.

(M From stresses induced by cooling. On cooling, the asphalt

cement shrinks more than the aggregate and tends to draw the particles

closer together. During the cooling period the total effective contact

area will be developed. Besides this, the cooling process causes the

asphalt to go into a state of tension which could be made analogous

to prestress due to consolidation. Calculations shown by the authors,

using Troutons equation, indicate that very high prestress forces can

be developed due to contraction of asphalt. Stresses up to 5^0 kg/cm^

have been calculated by the authors using Trouton's equation which

states that 3where:

— absolute viscosity of asphalt

Cr=r tensile stress in asphalt

tensile strain in asphalt

t,= time of loading that is cooling.

From the preceding discussion by Douglas and Tons, it is believed

that the sand-foamed asphalt specimens in this investigation, are

approaching the condition of a semi-mastic, where the films are

approaching minute dimensions due to the excellent dispersion by the

foamed process. It is further postulated that the high strengths are

due partly because of the thinner more uniform and more continuous film l

developed. It is also possible that the asphalt exists in capillary

tension thus increasing the strengths.

V.-V-' ‘V

' ■ "■ O . ■■ . ■-1 . : ‘ , ,■ .

CHAPTER III

CLASSIFICATION AND ANALYSIS OF MATERIALS

Sand

Since this was a continuation of previous investigations by

Pennell (1962), Jones (1962), White (1962) and Haas (1963)* only

one type of soil was used in this investigation. This soil was a fine,

rounded, poorly-graded quartz sand obtained from the Department of

Highways McGinn pit No. 2 near Stony Plain, Alberta. This was part of

the same sand obtained by White (1962). A microscopic examination done

by Haas (1963) showed the sand to have ellipsoidal and spherical quartz

sand with a small amount of blocky grains. Table I shows the gradation

of the sand which is an average of three sieve analysis as obtained by

the author; other significant properties are included in the Table.

Appendix A shows some of the preliminary testing done by White (1962)

and by Pennell (1962). It was felt that repetition of such routine

classification tests was not necessary.

Although only one type of sand was used, the gradation of this

sand was altered by recombining mineral filler* with the original sand

(See Appendix B for details). Different amounts of mineral filler were

added to obtain various gradations and percentages passing the No. 200

sieve. Table II shows the calculated gradations of the sands used in

this particular series.

The moisture content of the bagged sand was determined to be very

close to 0.1 percent. The results of the classification tests showed the

*Filler material was obtained by sieving the original material and retaining the minus No. 200 sieve material.

- ■

■

original sand had a relatively low uniformity coefficient (Cu) of 2.16,

thus indicating a uniformly graded sand. White (1962) obtained a

uniformity coefficient of 2.00 for the same sand. This difference is

believed due partly to differences in drawing the grain size distribution

curves and to slight differences in gradations of the bagged sand. These

slight differences were probably due to poor sampling and quartering

methods. The Standard AASHO maximum dry density and optimum water

content, obtained by White (1962), were determined to be 103.2 pounds

per cubic foot and 13.6 percent respectively.

Asnhal •’

A 150-200 penetration asphalt supplied by Husky Oil and Refining

Company Limited was used in Series 1 and 2. The analysis of this 150-200

penetration asphalt was performed by the supplier and is included in

Table III.

A SSI water emulsion asphalt, supplied by T.J. Pounder and

Company Limited was used in Series 3* Table IV lists the pertinent

information relative to this emulsion asphalt.

■- V f

'

19

TABLE I

Original Sand Properties

Unified Classification SP

AASHO Classification A~3

Standard AASHO Dry Density 103-2 pcf

Optimum Water Content (Standard AASHO) 13.6$

Specific Gravity 2.66

Uniformity Coefficient (Cu) 2.16

Grain Size Distribution

U.S. Standard Sieve Size Percent Passing

10 100.0

20 100.0

UO 99.6

60 6l.O

100 15.5

200 5-7

v;J-j : .X

'

20

TABLE II

CaJcniated Grain Size Distribution of the Sands

Series No. Mix No. Percentages Passing the Screen Nos. Uniformity Coefficient

. 50 .. 100 200 Cu

2A 8 87.0 1+5.0 20.2 b.3

2A 9 85.7 39.7 1^.5 1+ .02

2A 10 83.7 30.5 12.3 3-19

2A 11 82.5 26.2 9.1+ 2.59

2A 12 81.5 21.5 7-2 2.1+1+

2A 13 87.I 7-5 2.7 1.90

2B IT 85.7 39-7 11+.5 1+ .02

2B 18 82.5 26.2 9.1+ 2.59

2B 19 87.I 7«5 2.7 1.90

2B 3 87.0 15.5 5-7 2.16

21

TABLE III

Analysis of 1S0-200 Penetration Asphalt Cement

Flash Point (Cleveland Open Cup . Degrees F)

Penetration at 77 Degrees F (100 grams, 5 seconds)

Ductility at 60 Degrees F (5 cms. per minute)

Loss on Heating (50 grams, 5 hours at 325 Degrees F)

Penetration at 77 Degrees F (100 grams, 5 seconds) after

heating to 325 Degrees F, percent of original penetration

Solubility in CCl^ (percent)

Specific Gravity at 60°F

TABLE IV

Analysis of SS-1 Emulsified Asphalt

Percent residual asphalt hy evaporation

Saybolt furol viscosity at 77 degrees F

)425°F

153

100 +

1 -°jo

70 +-

99.8

1.0291

62.0$

1*5

Penetration of residue 150-200

CHAPTER IV

THE TESTING PROGRAM

Outline of Testing Program

All specimens were made with a poorly-graded fine quartz sand

stabilized with either a foamed or an emulsion asphalt. Cured and

immersed shear strengths, of the 2 inch diameter by 1+ inch specimens,

were determined using a triaxial compression test with constant

positive lateral pressure. A constant rate of strain of 0.01 inches

per minute was used throughout the investigation. In all cases the

curing period consisted of 7 days in a 100°F oven and the soaking

period consisted of immersing the samples in room temperature water

(70 to 75°F) for Ik days.

Preliminary testing done by White (1962) ascertained what

compactive effort was necessary to give densities, in the 2 inch

diameter by k inch high mold, comparable to Standard AASHO densities.

He found the necessary compactive effort to be 30,300 foot pounds per

cubic foot.

The main objectives of the testing program were to compare shear

strengths of a foamed asphalt sand mixture with the same sand stabilized

with an emulsified asphalt. Also the effects of varying gradations

on triaxial shear strengths were determined. More specific details of

program objectives are given on the next two pages.

The first part of the testing program, Series 1A, was set up so as to

■

'

.'•'j&qm'oo J-. g..crf i >ni t! ^cf • ipsajiifi

■

23

TABLE V

OUTLINE OF TESTING PROGRAM

Test Series

-.----—-—--—

Asphalt Used

. Part No.

Mix No.

1 Conditions

1 Foamed 150-200 Fenetration 1A 1,2,3,! Cured and Soaked

Asphalt cement. IB 3,5,6 Cured and Soaked 1C 7 Cured

2 Foamed 150-200 Penetration 2A 8,9,10,11 Asphalt cement. 12,13 Cured

2B 17,18,19 Cured

3 SS-1 water emulsion asphalt 3A 11,15,16 Cured and Soaked

0 B J E C T S

Series 1 General: To ascertain the behaviour of the altered University of Alberta

foamed asphalt laboratory installation and to develop technique in

the new method of developing foamed asphalt.

Part 1A Same as general; also to check the effect of mixing and molding water contents, and to check the effect of water immersion on strengths of foamed asphalt specimens.

Part 13 Same as general; also to check the effect of foamed asphalt setting, to check volume changes and finally to check the effect of alternate cycles of immersion and curing.

Part 1C Same as general.

Series 2 Part 2A To check the effect of gradation at an estimated film thickness

of 5.1 microns.

Part 2B To check the effect of gradation at an estimated film thickness

of 2.7 microns.

Series 3 Part 3A To check the effect of curing back to various molding water

contents; also to compare results to those obtained for foamed asphalt mixtures.

develop technique in the new method of producing foamed asphalt , to check

the effects of mixing water on shear strengths as measured by the triaxial

test and to check the adequacy of the slightly altered foamed asphalt

pugmill at the University of Alberta. The adequacy of the equipment was to

be determined by comparing the results to those obtained by Haas (1963),

who used the facilities of a proven foamed asphalt laboratory installatiortf

At the same time, it was decided to determine whether the reduction in the

shear strength due to water immersion, was permanent or not.

Because preliminary results showed significant differences with Haas's

results, it was necessary to carry out Series IB and 1C. Also because

these significant differences were attributed either to differences in the

foaming characteristics of the nozzle settings used or to differences in the

gradation of the sands, it was necessary in these series, to look at both

of these variables. After this investigation, it was finally determined

that the University of Alberta foamed asphalt laboratory installation was

adequate to foam asphalt.

Series 2 of the testing program was set up to check the effect of

grain size distribution on the cohesion and the angle of internal friction

of cured specimens as measured by the triaxial test. Because shearing

strengths of the sand bituminous mixtures were known to vary with film

thicknesses, it was decided that a logical method of comparing variations

in strengths due to variations in grain size distribution would be to

compare them at approximately the same film thickness. Because film

thicknesses were known to vary with surface area and quantity of asphalt

used (Campen, Smith, et.al. (1959))> it was necessary for a given film

*BCH commercial pugmill used by Haas (1963) (See Figure IB)

25

thickness, to increase asphalt content as the total surface area of the

mixes increased.

Because this was a preliminary investigation of the effect of film

thickness with varying gradations, it was decided that more direct methods

of determining surface areas were not warranted. Instead, an approximate

method of determining surface areas using Hveem's(1942) theoretical

procedure was used . (See Appendix C for more details). This method

assumes spherical particles and the surface area is then determined

on this "basis. Based on the assumption that asphalt exists in the

form of uniform films, film thicknesses were determined by dividing the

total surface area into volume of asphalt. In this series, it was hoped

to measure the effect of the uniformity coefficient and the percentage

passing the No. 200 sieve on shearing strengths. See Table II for the

different materials employed in Series 2A and 2B. In Series 2A an estimated

film thickness of 5-^ microns was used and in Series 2B an estimated film

thickness of 2.7 microns was used. The thickness of 2.7 microns

corresponded to the optimum foamed asphalt content of 4.2 percent found

by Haas (19&3) •

Series 3 "was run to determine the effect of drying back emulsion

asphalt mixtures to various molding water contents and to compare

these strengths to those using foamed 150-200 penetration asphalt cement.

Because a comparison of shear strengths with the 4.2 percent foamed

asphalt mixtures was required, a residual emulsion asphalt content of

4.2 percent was used. Also because various degrees of curing back were

required, molding water contents of 3,5 and 8 percent were used.

'

f Ot :

: •'\0'.

26

Preparation of’ Batches and Specimens

Generally, the preparation of batches involved the mixing of the

sands with the proper mixing water contents and asphalt contents. Molding

or preparation of specimens resulted in 2 inch diameter by k inch high

specimens. The specific details of the preparation of batches and of

specimens are included in Appendix D.

Apparatus and Testing Procedures for Triaxial Tests.

Apparatus and testing procedures were the same as those used by Haas

(1961) and Pennell (1962). Details of both the apparatus and testing

procedures used can be found in Appendix D.

Discussion of Testing procedures and Apparatus.

General. Because it was desired to compare results to previous

work done by investigators such as Pennell (1962) and Haas (1963) testing

procedures and apparatus were dictated by those used by the above-mentioned

Reasons for the original choices are included in the following paragraphs.

Density gradients and compaction technique. Pennell's investigation

showed that a double plunger-type mold with dynamic compaction was not

practical, although ideally this procedure should minimize longtitudinal

density gradients. Pennell found that when compacting wet samples, the

plunger base moved a considerable distance into the mold; conversely for

dry samples, it would not move up into the mold. Thus, it was necessary to

use a solid base which prevented compaction from the bottom and thereby

produced some density gradient in the sample. To minimize these density

gradients, he increased his blow count as each layer was added. However,

the procedure of compaction used in the investigation seems to be justified

27

because extensive testing by the Highways Division of the Research Council

of Alberta showed negligible density variations in 2 inch diameter by

4 inch specimens.

The compactive effort of 30>300 foot pounds per cubic foot was

used because it resulted in densities comparable to Standard AASHO

densities for the natural sand. The procedure of obtaining comparable

Standard AASHO densities is logical because field densities are usually

between 95 to 100 percent of Standard AASHO densities. Thus strengths

measured in the laboratory can be expected to be comparable to those

existing in the field.

Size of specimen. The 2.0 inch diameter by 4.0 inch sample was

used throughout this investigation. Thus any variation in strength due

to different length to diameter ratios was excluded. The length to

diameter ratio of two compares with ratios used by many authors.

Evidence presented by Lambe (1951) shows that ratios of 1-5 to 3*0

are satisfactory and that restraining effects and column buckling effects

are insignificant for this range. Also evidence by Lambe showed the

length to diameter ratios to increase with increasing angles of internal

friction.

Mixing time. In Series 1 arbitrarily chosen mixing times of 2\, 2-^,

2 and 13/4 minutes for mixing water contents of 3> 7> 11 and 15 percent

respectively represents an attempt to produce the same degree of

distribution of asphalt for each mix. Visual inspection indicated that

this worked well for molding or mixing water contents of 7 to 15 percent.

However, for 3 percent molding water, the mixture was extremely "spotty"

with a very poor distribution of asphalt.

EX

TR

US

ION O

F 2

IN

. B

Y 4 I

N.

SA

ND

-AS

PH

AL

T S

PE

CIM

EN

S

30

[A] MEASURING SPECIMEN

[B] SETTING UP SPECIMEN

PLATE 3

S.F

/

32

 CD a.

CL <

 z o o

<

c I— D <

In Series 2 n constnnt mixing time of 2 minutes was used so as to

"be consistent with Series lj that is, for 11 percent mixing water content

a mixing time of 2 minutes was used. Another reason for the constant

mixing time was to eliminate possible variations in results within a

particular test group due to varying degrees of distribution of the

asphalt binder.

In Series 3 a mixing time of 1 3A minutes was used to be consistent

with Series 1; that is, for 15 percent mixing water content a mixing time

of 1 3A minutes was used.

Grain-size distribution. The procedure of using filler material

obtained from the original sand and then adding it in different

proportions to obtain different gradations is quite logical. The use

of such a method excludes possible variations in strengths due to

variations in the shape, surface texture and surface chemistry of the

filler. Thus strengths obtained can be compared to mixtures using the

original sand without the possibility of variations due to the above-

mentioned reasons.

Gradation effects on shear strengths compared at the same estimated

film thickness. The procedure of comparing shear strengths of sands using

the same estimated film thickness throughout has advantages and disadvantage

The first disadvantage is a practical one in that no consideration to

air voids is given. As shown on page 7 the smaller the surface area per

pound of aggregate, the greater the required film thickness to give the

required air voids. Another disadvantage of comparing shear strengths of

sands, with varying gradations, at the same film thickness is that no

consideration is given to the ratio of contact area to surface area. It

;

■

is realized that this ratio is dependent on the ratio of film thickness to

particle diameter and that variations in this ratio exist for any one

particular mix. However, due to the small range of particle sizes used,

it is believed that this ratio does not vary significantly for any one

particular mix. Because of the relatively large size of particles used

and the very thin films used, the ratio of conta.ct area to surface area

is believed to be more an inherent property of the sand itself and thus

is affected only slightly by varying film thickness.

Even though the procedure of comparing shear strengths of sands

at the same estimated film thickness, has its limitations, it is believed

that because of previous work done by investigators (See Pages 6 and 7)

this method is justified. As shown on pages 6 and 7? varying film

thickness has a very significant effect on shear strengths. Thus, in

accordance to this, it was decided to compare shear strengths of sands

at the same estimated film thickness. The procedure of trying to achieve

the same film thickness was determined using Hveem's theoretical surface

area constants. As mentioned previously on page 25, because this was

a preliminary investigation of the effect of film thickness, more direct

methods of determining surface areas were not employed. For the

sand investigated it is considered that because of the rounded, non¬

absorbing, quartz particles, Hveem's theoretical method of assuming

perfectly round particles seems reasonable.

Mixing and molding water content for emulsion asphalts. The procedure

of using 16 percent mixing water content was based on work by Knowles (19$•

He reported that the use of mixing water contents less than the optimum

’

. .-.I .i

35

AASHO water content resulted in poor distribution of asphalt; this, he

reported to be caused by the breaking of the emulsion before mixing was

completed. Thus it appears that the use of l6 percent water made it

possible for the asphalt emulsion to be distributed before breaking.

The procedure of drying back to some molding water content is a

common procedure in the field. This enables the emulsion to break and

the asphalt to coalesce. The degree of coalescing depends on the

extent of curing. Thus the curing back to 8, 5 and 3 percent molding

water content was a laboratory procedure in which the degree of

coalescing was determined by strength tests and by visual inspection.

Curing of specimens. The curing period of 7 days in a 100°F oven,

was chosen because previous work done by Jones (1962) showed that cured

strengths levelled off after approximately 5 days. Thus, by choosing

a curing period of 7 days differences due to insufficient curing were

negligible. Also by choosing a period of 7 days, a delay in testing

of 1 day would not change the moisture conditions of the sample and thus

not significantly affect the strength. Generally, it was found that

the water content after 7 days curing was 0.^ to 1.5 percent.

Although all the sand-asphalt specimens were cured under the

same conditions, it must be emphasized that this curing procedure is

extreme and would not easily be obtained in the field. The severe

curing exists partly because of the high surface areas that are

exnosed. It is believed that a more realistic approach would be to

simulate field conditions by allowing curing only from the top of the

specimen instead of all sides.

36

Immersion of specimens. The immersion period of 1*4 days was selected

once again so as to compare results to work done by Haas (1963). As

reported by Pennell (1962) and Haas (1963) water continued to be

absorbed after 14 days immersion. It is believed, however, that 1*4

days is sufficient to show the effectiveness of asphalt as a water¬

proofing agent.

Length and diameter measurements of specimens. Volume changes were

measured using both the calipers and the extensometer method as reported

by Jones (1962). In some cases the calipers were used and in other cases

the extensometer method was used. For any one specimen both methods gave

approximately the same readings. However, the accuracy of both methods

is questioned because of rough surfaces and of loose sand grains on

these surfaces.

Triaxial test procedures. Test conditions were dictated by other

research investigations, so that tesi results could be compared without

introducing variables due to variations in testing procedures.

Testing at room temperature is a reasonable procedure, mainly from

the point of convenience; also it is the same testing temperature used

by Haas (1963) and Pennell (1962).

The no drainage condition during the triaxial test once again is

coincident with work done by Haas (1963) and Pennell (1962). As reported

by Haas (1963) and also verified by the author's work:

"The fact that the Mohr rupture envelopes are essentially linear, within the range of lateral pressures employed, suggests that no neutral pressures were set up during the test. Certainly, the range of lateral pressures used seems insufficient

-

■ .. . ; _ ^ ,

to cause consolidation of the sand-asphalt structure within the relatively short testing periods."

The use of membranes is considered to provide negligible lateral

support to the specimens. Work done by Bishop and Henkel (1962) shows

the effect of lateral support of 0.008 inch thick membranes to be

negligible.

The air-over-water pressure device used to maintain cell pressures

can be controlled to within 0.1 to 0.2 psi. Thus, this method of

applying pressure is reliable^ accurate and extremely practical from the

point of view of time.

The rate of strain of 0.01 inches per minute was used because result

were to be compared with Haas's (1963) work in which he used a rate of

strain of 0.01 inches per minute. The reason behind the original use

of this rate was primarily due to convenience. That is, it was then

possible to test many specimens in a reasonably short time. There is

evidence Knowles (1962) that for sands similar to the one employed in

this investigation rates of strain of 0.011 inches per minute to 0.055

inches per minute do not significantly affect compressive strengths as

measured by the positive lateral pressure triaxial test. Thus any rate

in this range could have been used. There is other evidence, Goetz and

Schaub (1959)> that a very wide range in the rate of deformation (.02

to 2 inches per minute) significantly affects the compressive strength;

however, they showed that a small range of .02 to .05 inches per minute

had a negligible effect on compressive strengths. Thus, the rate of

strain used is consistent with rates of strain used by other authors

and is convenient from the point of view of time.

'

'

'

38

As reported by Haas (1963) as the number of tests per Mohr envelope

increased the 95 percent confidence limits decreased. He showed that

for 8 specimens or more, the 95 percent confidence limits are sufficiently

small so as to reasonably define the Mohr rupture envelope. The 95

percent confidence limits are the accuracy or confidence to which the

shearing strength can be defined.

As reported by Haas (1963) and recommended by Pennell (1962)

t he procedure of a number of tests at 9 different lateral pressures

(0 to 80 psi with 10 psi increments) is considered to define the shape

of the Mohr rupture envelope more accurately than the procedure of 9

tests at only 3 different lateral pressures.

CHAPTER V

SUMMARY AND DISCUSSION OF RESULTS

General

This chapter contains a summary of results presented both in

tabular and graphical form, with detailed discussion. All figures

are grouped together and are numbered consecutively from 1 to 26.

Statistical Evaluation of Test Results

As illustrated by Haas (1963)* 8 or more specimens were required

to determine with reasonable accuracy the Mohr envelope based on

95 percent confidence limits.

The actual values of 95 percent confidence limit in shearing

strength are tabulated in Table VI. These values are tabulated for

each Mohr envelope at normal stress values of zero, mean normal stress

and double the mean normal stress. The values shown were obtained from

Haas's computer solution of Balmer's method (1952).

The largest confidence limits obtained were for the cured k.2 percent

foamed asphalt mixture using the McGinn pit sand used by Haas (1963).

These limits ranged from 6.9 to 3*6 to 6.9 psi for respective normal

stresses of zero of the mean and of double mean values.

In series 1A, IB and 1C, confidence limits ranged from 1.0 to 5-2

psi at zero normal stress, from 0.6 to 3*^ at the mean normal stress

and from 1.0 to 5.2 psi for double the mean normal stress.

Confidence limits in Series 2A and 2B ranged from 1.9 to 5-2 psi

bo

for zero and double the mean normal stress; smaller limits ranging

from 1.0 to 2.6 psi were obtained for the mean normal stress.

Confidence limits in Series 3A ranged from 1.0 to 3*1 psi for

zero and double the mean normal stress; smaller limits ranging

from 0.7 to 2.1 were obtained for the mean normal stress. This means

that in all of the work done, Mohr envelopes shown are defined with

greater accuracy at the mean normal stress than at either zero or

double the mean normal stress.

Confidence limits generally increased with increasing shearing

strengths. For example, the 95 percent confidence limits in Mixes Nos. 1 to

b at zero normal stress increased from 1.0 to 3-8 for respective

cohesion values of 12.0 to 31.2 psi.

In other words, the calculated unit cohesion values can be

considered to be defined with 95 percent confidence in its accuracy,

with plus or minus 1.0 to 3*8 psi, for respective unit cohesion values

of 12.0 and 31-2 psi. This represents respective 95 percent confidence

limits of 8.3 and 12.0 percent.

Some of the 95 percent confidence limits were as high as plus or

minus 23 percent of the calculated cohesion values; however, the

majority of the limits, are in the range of plus or minus 10 to 15

percent of the calculated cohesion values. Thus, it is considered

that the majority of the confidence limits show that the Mohr strength

values of this investigation are defined with reasonable accuracy.

Strength of Foamed Asphalt Stabilized Sand

Series 1A. As indicated in the previous chapter, Series 1A was

-.V, .... o ■ >,

' • • - ... V i;K -

■ r. ■ .• O ' -• ■> ■

O, ' ' : JO:' .....

run to determine the adequacy of the altered University of Alberta

laboratory pugmill, to determine the effects of mixing and molding

water contents on strengths of foamed asphalt mixtures and to determine

the effects of water immersion.

Figure 2 and Table VI show strength envelopes for the cured and

the soaked series of the McGinn pit sand stabilized with 4.2 percent

asphalt and varying water contents. This asphalt content was used

because it was necessary to determine the adequacy of the equipment

by comparing results to Haas's (1963) work in which he used 4.2

percent asphalt. As shown cured cohesive strengths ranged from 12.0

to 31-2 psi for respective molding water contents of 3 and 15 percent.

The other cohesion values were 19-9 and 25.2 for respective molding

water contents of 7 and 11 percent. This illustrates that as the

molding water increases the cured cohesion increases. Referring to

Figure 3> Series No. 1, it can be seen that the changes in cohesion were

largely the result of increased density. This increase in density was

caused by increasing molding water contents. It is also believed that

the better asphalt distribution with increasing mixing water contents

caused the mixtures to become more cohesive.

As shown in Figure 2, the cohesion intercepts decreased considerably

die to water immersion. This Figure shows cohesion values of 0.5 to 5-0

psi for respective molding water contents of 7 and 11 percent. The low

cohesion intercept of the mixture with 7 percent water as compared to 11

percent water was caused by a higher amount of absorbed water and a

higher degree of saturation after immersion. (See Figure 4). This was

caused by a lesser degree of asphalt distribution and by a less dense mix.

42

f CO UR Tt

STEAM

STEAM-_

LINE

•: ■ v W „ F „ C U R T I ' |

OPERATION CONTROL

STEAM VOLUME PRESSURE REGULATOR REGULATOR VALVE

PUG

MILL

MIXER

ASPHALT SUPPLY

DIAGRAM OF FOAMED

ASPHALT SYSTEM USED BY HAAS[ 19631

FI EURE i l ]

NOZZLES

PRESSURE

GAGE

PUMP

- ASPHALT LINE

SPRAY

BARS

CONTROL

PADDLES

TA

BL

E

VI

IY

OF

TE

ST

RE

SU

LT

S

A „ P

'sR- to 5 «H

b <

ft ft O O

EH Pf

CD > ft < t3

P> •H a

LfN P ,11 ON +1

TO" —ft]— p> ft • • o P p ft ft W -q *P qj q> •H 0) o H a • • o ^1 rH ft ft > p ft a ft ft p •H

P c— rc cd ft cd ft P p ft a> u> d) as • •P a p ft •P r3 a) o ft o cj • P p ft P LfN Cti CM Pi •rl ft

V CQ a; • *\ ft ft 0) o o cO «P cti ft to d P P •H p ft ft ft ft 10 1 o -P O nd

•H 0) rH rH 0) to p o 1 ft ft rH ad cu p QJ P p P •H otf p o ft cu QJ (/) oj CO N o N O H d ft rM 0) QJ o p Cl) ft P CO h 1 •H •H •H »H QJ > ro ft p p o P a a 0) Pi CO QJ rH CO H CQ

p> p IQ •H p p d) P o Me p ft fl •H rH •H rH QJ to tfl d CQ <v ft CJ cq g ft Ctf

a ft ft a ft P ft R H ft ft • 1 o CJ o o c\3 a d a

•H ft ft <D P as d) • 'C ' ft o •H ( 1 a •H p QJ Qj O ft p 0) P d t* •tfi. d) d) ft •P ft CO CO P 1—1 cti H 0) <U d) rH p P p p p p rH H •

o CO CO ft ft to p i—1 pi o o ad d • 'H co tJ P a o cq P •H o O •H a <1) •H a p P ft P P QJ R p <p O •H a H 1 1 p H rH p ft d ft QJ d ft ft P p •H P P ft p • cu ■d ft d •H cd •H CQ •H P) i (0 *H d

o ft p O i—1 Cl) to ft d) d) ft ft ft cq a CQ o *H ?S rt cO P ft P ft •H p •H rH p p CJ P O ft p o o O OJ • •£ O QJ • ft ft •

•H •H P •H p to CO cq P cq A-l P CQ p p p p CO x o qj cq CO • o cu p • 1 1 • • p DU • P d p p p p p qj •H o ft 0) p ft p cq •H ft ft ft ft ft M o ft H o 3 QJ QJ P QJ QJ

•H a •H 5sJ o o O o CJ QJ a O qj a S ft CJ S Pi o

p QJ • O VO H ft •H H Pi p • •H CQ d 1—1

ft <! ft

O t— on LfN CO VO CM ft CM ltn CM i—I On E CO ft ON 1N-

i—I r—I i—I CM

O

V>

W XTW

LfN LT\ cn-d- CM ft LfN ft ft

LTN o ft CM ft

CVj ON ft CM ON i—1

CM

CM CM CM CM CM CM CM CM OJ CM CM CM CM CM LTN O ON On CM VO o o 00 CM CM CM CM CM CM

-d- -d" ft- -zr ft ft ft ft ft ft ft O ON ON [>- i—1

VO LfN rn ft ft ft ft ft ft

O ft -d- ft ft cn ft LTN CM

co co ON ON

co co co o On On On O

i—1

VO ft ON ON ON ON

OO CO CO CM co oo co vo ft co CO CO ft tc- co o LTNCO ft vo rH O cn ft ft O ft ft o

ON CO ON oo ON On O

ft

00 ON ON ON

ON O ft On o o

1—1 1—1

On On On i—1 On On on o

H

on O rH

coco o o ft ft

LCN CM CM ft O O O ON i—1 i—1 i—1

CM ft o o 1—1 1—1

vo ON

ft ON o On On o

ft

ft ON O ON On O

1—1

oo 00 CO CM CO CO co vo ft co co <o ft ft CO O LT\C0 ft V0 rH O on ft ft O ft ft o

co ON

CO ON on ON On O

ft

CO ON ON On

ON O ft ON o o

ft ft

ON On ON ft ON On On O

i—1

on O i—1

co co o o ft ft

LTN CM CM ft O O O On ft ft ft

CM ft O O ft ft

VO ON

ft CN o On ON O

ft

ft ON o ON ON o

ft

o VO VO co O 00 VO co c— CO VO CM ft On ONCO LTNCO CO VO CM CO o ft ft cC' O O ft

rH CM CM cn m m CM cn CM cn cm lt\ CM VO ft cn mft ft cn LTN on on * 1 C‘J ft CM ft rn

VO LTNP- O ft LTN -d- ON LTN LTN -d" ft ft VO O «H OnV0 ft O 1—1 ft on cn cm -H- t i—1

o. ft ft CM CM CM ft ft ft oj <—1 m i—l on ft CM ft CM CM CM CM CM rH O ft ft ft O CM

O VO VO CO O CO VO co 1ft CO vo CM ft ON On CO LOCO CO VO CM 00 o ft ft co O O ft

1—1 CM CM CO cn cn OJ cn CM cn cm LfN CM V0 ft cn enft ft cn LfN O'-' m ft CM ft cm ft cn

O ft ft VO rH rH ft 1ft o ft ft t- 1ft o -d- O O O cn ft on oo ON CO ftvo ft O LfN

H on

ON CO CO CM CM CM

O ON on CM

CO tft ON CM CM CM

On CO ON ft CM CM CM on m

VO ONCO VO LTN LTN CM CM CM CM CM CM

ft O CM cn

CO CM

co o o CM cn on

ON ON ON CM CM CM

O On CM CM O O CM ft ON O CM cn ft ft ft LTN VO LTN ft CM O LfN LfN ftCO ON co ft ft

% ON LTN ft ft CM cn

O LTN LTN ON (ft CM CM CM

ltn m CM CM CM CM CM

i—l oo cn ft cn ft ft CM CM CM CM CM ft

cn o ft on

ft CM

ft O ft ft ft

O ft CM

p p p p QJ QJ Q) QJ P P P P d d d d ts is is Js

ro rl in 1—1 1—1

P P QJ 0) P P d d

Tft-'eR.

P QJ

CO = -

P QJ ft = - V

ary

ing

perio

ds

of

cu

rin

g

an

d

so

ak

ing

.

i—1 i—1 •H

p Ph G

radati

on

IT

Grad

ati

on

»T

p p p QJ QJ QJ P P P d d d ts ts ^

pt-Nsa-tR CO LfN CO

P P P QJ QJ QJ P P1 P d d d ft IS ft

'on Sn do

rH CM enft OJ cn m LTN vo (Vi cn mvo tft co on o H oj m i—1 i—1 rH i—1 17

18

ON i—1

-r-t LfN VO i—1 i—1 i—1

ft LTN vo ft ft ft

3 < <: FQ m o 1 < PQ

i < 1

rH rH rH rH rH CM CM cn cn

*»3

*

K*

• 1

.

1

■

IMM

ER

SIO

N T

IME

IN

HO

UR

S

o r-> lo cj

U"'

Lil Ll. _i

il

iSd - X- SS3d_LS 9N!dV3: -

NO

RM

AL S

TR

ES

S - r

— P

SI

NO

RM

AL S

TR

ES

S -

0"

—

PS

i

VO

LU

ME

OF S

PE

CIM

EN

S

IN C

UB

IC C

MS

.

21 3.0 0

212.00

2 11.0 0

210.00

2 0 9,00

2 08 .0 0

VOLUME OF FOAMED asphalt SPECIMENS AF^EP ALTERNATE

CYCLES OF CURIN'' AND SOAKING

SHAKE D RE CURE n R ESHAKE n R E CI.'Rb D

MIXING IN L . OTA . p UG MILL

MOLDING WATER CONTENT OF

FOAMED ASPHALT CONrtN - 4 2'J7Ji

I CYCLE - 7 DAY'S OF CORING AT

100 DEGREES F. AND 14 D A Y 5 OF

IMMERSION IN ROOM TEMP. WATER

EACH POINT REPRESEN S A N AVERAGE,

VA LUE OH II SPECIMENS

NO

RM

AL S

TR

ES

S -

<T -

PS

!

0

2-4

28

TIO

B

Y W

EIG

HT

FIG

UR

E

14

o

0

o 0 <

O 0

° < r-s. 00 cm

0 0 o II z 7 0 (ft Ut w Ul

5 " a

•

) 2

4

6

8

10

12

14

16

0

.4

.8

1.2

1.6

2

%

AS

PH

AL

T

FIL

LE

R T

O A

SP

HA

LT

RA

VA

RIA

TIO

N O

F A

NG

LE O

F

INT

ER

NA

L F

RIC

TIO

N

WIT

H %

A

SP

HA

LT

OR F

ILL

ER T

O B

IND

ER

RA

TIO

( 5 •

o u ■ r

< >

c >

0 o

o

C )

p

O

o

•

0

... ^

1

— O -Ji oa r-> <£> lo CO CO CM CNJ CM C\J CM

S33a03Q - (j) - NOli-OIdd “lVNa3J.NI 30 300NV

r\

O a ^

O # h <

•

a « m C\J ^

N

K 0 0 0 z z

J" (/) Ifl o L) w

£ 5 S

I-

C ' •

I I O

t El

5. £

c T~

Z p O 2

A =>

•

o >

CL CL Ll □

J -1 •

o

0. O UJ

i n h a. ■ z o

Ll LL o > .

Ill

3 W O o Z o . < "

3 O O z o z y- 0 z

h cr < T

C

Ci- Cl < _ > 3

c i

* c

O

•

A W

•

O

t-

o CO

cn c\j

CO CNJ

rv CM

o CM

Lf>

CM

S33d33Q -0 -NOIJ-OIdJ 3VNd3±NI 30 339NV

OT

AL D

RY

UN

IT W

EIG

HT -

PC

F

% P

AS

SIN

G

NO

. 2

00 S

IEV

E

69

As shown in Figure 2 and in Table VI, friction angles decreased

sLightlv with increasing mixing water contents up to 11 percent. However

a slight increase was noted for 15 percent mixing water. This is believed

to have been caused by experimental error. The general decrease of friction

angles is believed to have been caused by better asphalt distribution

\hich is believed to decrease the grain to grain contact.

A comparison with Haas's (1963) results (Figures 5 and 6) showed the

cohesive strengths and the angle of internal friction to be larger for

his investigations. As shown in Figure 5> respective cured cohesion and

angle of internal friction values of hk. k psi and 33*5 degrees were

obtained by Haas as compared to values of 25.2 psi and 28.1 degrees

obtained in this investigation. Also as shown in Figure 6, respective

soaked cohesion and angle of internal friction values of 5-8 psi and 32.0

degrees were obtained by Haas as compared to values of 5*0 psi and 29*1

degrees obtained in this investigation. Thus the comparison showed that

the results varied considerably and were not within experimental error.

It was then decided that perhaps the inconsistencies in the results were

c aused by either a difference in the foaming characteristics of the nozzles

used in the two investigations or by the difference in the grain size

distribution curves.

Series IB. Because it was known that the setting of the foamed

asphalt nozzle was very critical in controlling its foaming characteristics,

it was decided to investigate this aspect before investigating the

relatively small differences in the grain size distribution curves of the

two sands.

70

Also because work done by Haas showed high strength losses in

f oamed asphalt specimens upon immersion, it was decided to check the

effect of alternate curing and soaking. Finally, because Haas postulated

that the asphalt in foamed asphalt mixtures was in capillary tension,

it was decided to check the volume changes so as to correlate these

changes to alternate releasing and regaining of the capillary tension in the asphalt

phase. This postulation by Haas was based partly on measured volume

changes of foamed asphalt-sand specimens. These measurements showed

no volume changes due to curing, however, increased volumes were noted

after immersion. Haas then postulated that the increased volumes were

caused by the releasing oi capillary forces in the asphalt binder

upon water immersion. It was believed that only the release of these

high capillary forces in the asphalt changed the volume of the sand-foamed

asphalt specimens. Thus, it was postulated that if the specimens

regained their high cohesive strengths after recuring, that these

specimens would shrink back to their original volumes.

As shown in Figure 7> the nozzle settings (See Figure l) used

showed that within experimental error, there was no significant difference

in the Mohr strength envelopes. The nozzle setting is the distance the

steam jet is set in from the bottom outside edge of the nozzle. A visual

inspection test showed that there was a slight difference in the

percent expansion due to foaming of the asphalt and therefore a slight

difference in the foaming ability of the nozzle at the three settings.

(See Figure l). The greatest percent expansion occurred at a setting

of 0.625 inches. Settings of 0-575 inches and O.705 inches showed

71

considerable foaming, but not as much as at 0.625 inches. In all cases,

the asphalt was estimated to expand from 7 to 10 times its original volume.

Thus, for the relatively small range of the settings and steam and asphalt

pressures used, there was no significant difference in the ability of

the nozzle to foam asphalt.

As mentioned on the previous page, because Haas postulated that

volume expansion destroyed the supposed capillary forces in the binder,

it was decided to check volume changes with alternate cycles of wetting

and drying; also, it was decided to check whether or not the low

immersed shear strengths were permanent or whether recuring would restore

the specimens to their original cured strengths.

As shown in Figure 8, for the sand and percent asphalt used,

alternate cycles of wetting and curing do not materially affect the

final cured strength. Figure 9 shows the measured volume changes with

alternate cycles of wetting and drying. As shown volume changes are not

dependent on the supposed capillary tension in the asphalt phase, that is,

the volume of the specimens did not return to original values upon

recuring.

One point worthy of mentioning is that measured volume changes

were very small. Because of the innaccuracies in the length and diameter

measurements, it is felt that not too much emphasis can be attached to

measured volume changes. Thus, because of this, it is felt that measured

volume changes are inconclusive to substantiate the postulation of

capillary tension in the asphalt phase.

Series 1C. Because Series 1A and IB did not determine the adequacy

of the altered University of Alberta laboratory pugmill, it was necessary

72

to investigate the differences in the grain size distribution curves of

the two McGinn pit sands used. The main difference in the McGinn pit sands

used was that the sand used by Haas was finer and slightly less uniform.

A comparison of the grain size distribution curves is shown as follows:

Percent Passing Various Sieve Sizes No, 20 No. J40- No. 60 No. 100 No. 200

McGinn pit sand used by Haas 100.0 99-3 76. k 2U.7 8.1

McGinn pit sand used by author 100.0 99-6 6l.O 15.5 5-7

A comparison of the uniformity coefficients showed respective values

of 2.58 and 2.l6 for the McGinn pit sand used by Haas and the author.

It was then decided to use the remainder of the McGinn pit sand originally

used by Haas so as to determine whether or not the different Mohr

envelopes obtained in Series 1A, was caused by the difference in the

gradation of the sands used. As shown in Figure 10, the adequacy of the

foamed asphalt equipment was finally determined. This Figure showed,

within experimental error, no significant difference in the Mohr

envelopes obtained using either the BCH or the University of Alberta

pugmill.

Gradation Effects on Shear Strengths Compared at the Same Estimated Film

Thickness.

General. Because Series 1 showed the grain size distribution to be

critical, Series 2A and 2B were set up to investigate the effect on shear

strength of varying gradations. Because of previous work done by Mack (1957)?

Goetz (1951) and Douglas and Tons (1961) (Refer to pages 6 and 7); it was

decided to compare the effect of gradation at the same estimated film

: 7 • c' ■ : 7 3g

thickness. As explained in Chapter IV, and in Appendix B, the different

gradations were artificially produced by adding filler material to the

original sand; also the amount of asphalt required to produce the same

estimated film thickness for the different gradation was based on Hveem's

theoretical procedure, using surface area constants.

Series 2A. Gradation effects on shear strengths compared at an

estimated film thickness of S.t microns. The Mohr envelopes obtained

for the six gradations used are shown in Figure 11 and in Table VI.

Referring to Figure 3, Series 2A it can be seen that cohesive strengths

ranged from 17-2 to 28.7 psi for gradations with respective percentages

of 2.7 and 20.2 passing the No. 200 sieve. As shown in Figure 3? this

increase in cohesion was partly caused by an increase in total dry

unit weight. This increase in total dry unit weight (Figure 12) was

partly caused by increasing percentages of asphalt used and by increasing

percentages passing the No. 200 sieve.

Referring to Figure 13, Series 2A, for the sands used, cohesion

increased with increasing percentages of asphalt for the estimated film

thickness of 5-^ microns; also cohesion increased with an increasing

ratio by weight of filler to asphalt. It is postulated that the increase

in cohesion was caused partly by an increase in the viscosity of the fillo-

asphalt binder. That is the viscosity of the filler-asphalt binder is

believed to increase with increasing filler to asphalt ratios. This

postulation is substantiated in work done by Ka.llas, Puzinauskas and

Krieger (1962) in which they found shear strengths of mixtures to increase

with increasing filler to asphalt ratios; they also found these increasing

T1!

filler to asphalt ratios to cause increases in the viscosity of the filler

asphalt binder.

As shown in Figure 11 and in Table VI, friction angles varied from

25o "to 29.O degrees. This shows that for the gradations and percentages

of asphalt used, the friction angle is relatively insensitive to changes

in gradations compared at an estimated film thickness of 5*^ microns. As

shown in Figures ll and 15, Series 2A, no definite trends in friction angles

with filler to asphalt ratio,percent asphalt, total dry unit weight 01

percent passing the No. 200 sieve were observed.

Series 2B. Gradation effects on shear strengths compared at an

estimated film thickness of 2.7 microns. The Mohr envelopes obtained

for the four gradations used are shown in Figure 16. Referring to

Figure 3; Series 2B, cohesive strengths ranged from 21-5 psi to ^3-3 psi

for gradations with respective percentages of 2.7 and lU.5 passing the

No. 200 sieve. As shown in Figures 3 and 12, cohesion is dependent on

dry unit weight which is in turn dependent on the percentages passing

the No. 200 sieve and on the percentages of asphalt used. Referring to

Figure 3; Series 2B, cohesive strengths increased from 20.1 psi to

43.3 psi for respective total unit dry weights of 96.3 and 102.1 pcf;

when compared to 21-5 psi, this represents a relatively large increase

of 100 percent. Figure 12 shows total dry unit weights to increase from

96.3 to 102.1 pcf for respective percentages of 2.7 and 1^.5 passing the

No. 200 sieve; also total dry unit weights increased from 96.3 to 102.1 pcf

for respective percentages of 3-8 and 6.0 of asphalt used.

Referring to Figure 13, Series 2B, it can be seen that cohesive

strengths increased with increasing percentages of asphalt and increasing

filler to asphalt ratios for the estimated film thickness of 2.7 microns.

Cohesive strengths increased from 21.5 to 43-0 psi for respective asphalt

contents of 3.8 and 6.0 percent; also cohesive strengths ranged from

21.5 to 1*3.0 psi for respective filler to asphalt ratios of 0.72 and

2.1+2. It is postulated that the increase in cohesive strengths was caused

partly by an increase in the viscosity of the filler-asphalt binder. That

is, the viscosity of the filler-asphalt binder increased with increasing

filler to asphalt ratios.

Figure l6 and Table VI shows the friction angles to vary from 27-3

to 30.9 degrees for the sands investigated. This shows that for the

gradations and percentages of asphalt used, the friction angle is

relatively non-sensitive to changes in gradations compared at

an estimated film thickness of 2.7 microns. As shown in Figures 14 and

15, Series 2B, no definite trends in friction angles with filler to

asphalt ratio, percent asphalt, total dry unit weight, or percent passing

the No. 200 sieve were observed.

Comparison of Gradation Effects on Shear Strengths at Estimated Film

Thicknesses of 2.7 and 5.1+ Microns.

For the gradations and percentages of asphalt used, Figures 17,

l8 and 19 show higher shear strengths for the estimated film thickness

of 2.7 microns as compared to 5^ microns. Figure 17 and Table VI show

that, for the gradation having 14.5 percent passing the No. 200 sieve,

cohesive strengths of 23*5 and 43.0 psi were obtained for respective

film thicknesses of 5*4 and 2.7 microns. When compared to cohesive

76

strengths of 23.5 psi, this shows an increase of 83 percent. Figure l8

and Table VI show that for the gradation having 9-^ percent passing the

No. 200 sieve, cohesive strengths of 23*5 and 30.5 psi were obtained for

respective film thicknesses of 5-4 and 2.7 microns. This shows an

increase of 30 percent. Figure 1.9 and Table VI show that for the

gradation having 2.7 percent passing the No. 200 sieve, cohesive strengths

of 17*2 and 21.5 psi were obtained for respective film thicknesses of

5.4 and 2.7 microns. This shows an increase of 25 percent. Thus, for

the gradations used, higher cohesive strengths were obtained with an estimated

film thickness of 2.7 microns as compared to 5*4 microns. However, this

increase in the cohesive strengths increased with increasing fines

(percentages passing No. 200 sieve.) Increases in cohesive strengths,

for a film thickness of 2.7 microns as compared to 5*4 microns, of 83,

30 and 25 percent were obtained for respective gradations with 14.5, 9«^

and 2.7 percent passing the No. 200 sieve.

As previously mentioned, the 2.7 microns corresponded to Haas's

(1963) optimum asphalt content of k.2 percent and thus it is reasonable

to expect higher cohesive strengths for a film thickness of 2.7 microns

as compared to 5-4 microns.

Friction angles obtained varied from 25.7 to 29.0 degrees for the

series with a film thickness of 5*4 microns and varied from 27-3 to 3°*8

degrees for the series with a film thickness of 2.7 microns. Figure 17

and Table VI show that, for the gradation having 14.5 percent passing the

No. 200 sieve, friction angles of 27*3 and 29-0 degrees were obtained for

respective film thicknesses of 2.7 and 5-4 microns. Figure l8 and Table YE

show that, for the gradation having 9-^ percent passing the No. 200 sieve,

friction angles of 30-8 and 26.0 degrees were obtained for respective

film thicknesses of 2.7 and 5-^ microns. Finally, Figure 19 shows that,

for the gradation having 2.7 percent passing the No. 200 sieve, friction

angles of 28.9 and 25-7 degrees were obtained for respective film

thicknesses of 2.7 and 5-^ microns. This shows that for the film

thicknesses and gradations used, the friction angle is relatively insensitive

to changes in the film thickness.

Strength of Emulsion Asphalt Stabilized Sand

Series 3. As indicated in Chaper IV, Series 3 was run to determine

the effect of drying back to various molding water contents and to compare

strengths obtained to those of the foamed asphalt mixtures with b,2 percent

asphalt and with varying molding water contents (Series LA).

Figure 20 and Table VI show the strength envelopes obtained for the

cured and soaked series of the McGinn pit sand stabilized with k.2 percent

residual asphalt. As shown, cured cohesive strengths of 7*7, 10.8 and 11.9

psi were obtained for respective molding water contents of 3*0; 5-0 and 8.0

percent (molding water content includes emulsion water) and respective

total unit dry densities of 91-b, 99•b and 100 pcf. Figure 20 shows that

the cured strengths obtained using molding water contents of 5*0 and 8.0

percent are the same within experimental error ; also this Figure shows

that the cured cohesive strength obtained using a molding water content

of 3*0 percent is smaller than those obtained using molding water contents

of 5 and 8 percent. Finally, this Figure shows that the strengths obtained

for the immersed specimens with various molding water contents are the

78

are the same within experimental error.

The increase in the cohesive strengths obtained for molding water

contents of 3j 5 and 8 percent is believed to have resulted primarily from

higher density at higher molding water contents.

Strength Comparisons

Previous discussions in this Chapter have been concerned with a

detailed outline of the shear strength results of sand-asphalt mixtures

produced by stabilization with foamed asphalt and with emulsion asphalt.

Also included is a detailed explanation of the effect of gradations on

diear strengths compared at estimated film thicknesses of 2.7 microns and

5 .*4 microns. Because this thesis is a continuation of work done by Haas

(l963);a comparison with his results was necessary. Also,, a comparison

of results of the stabilization of the McGinn pit sand used with a foamed

asphalt and an emulsion asphalt was necessary to determine the superiority

or inferiority of the foamed asphalt process. To sum this, Figures 21 to

2k graphically compare these results.

Figure 21, showing cured strengths, illustrates the difference in

c ohesive strength between the foamed asphalt mixtures and the emulsion

asphalt mixtures. For the sand used cured cohesive strengths are shown

to be higher for the foamed asphalt process where adequate molding water

conten-tsOf 7, 11 and 15 percent were used.

A comparison of the cured Mohr envelopes obtained using approximately

the same molding water content (Figure 22) showed the superiority of

the foamed asphalt mixtures when compared to emulsion asphalt mixtures.

For the sand used cured cohesive strengths of 19*9 and 11.9 psi were

obtained for the foamed and the emulsion asphalt mixtures. When compared

79

to the lower cohesive strength of 11.9 psi, this represents an increa.se

of 76 percent. However, a comparison of shearing strengths at 50 pounds

per square inch normal stress shows the foamed asphalt mixture to have

a shearing strength of only l8 percent higher than the emulsion asphalt

mixture. As noted above, this compares to an increase of 76 percent at

zero normal stress. This is in agreement with previous work done by

Haas (1963) and Pennell (1962). Haas noted that the shearing strength

of -ured foamed asphalt mixtures was at least k times that of any other

cured sand-asphalt of his investigations at zero normal stress; on the

other hand at 50 pounds per square inch normal stress the foamed asphalt

mixture possessed only approximately double the shearing strength of the

other mixtures. Pennell (1962) showed that some of his soil-cement

specimens had times the unconfined compressive strength of some of

the immersed sand-asphalt specimens, yet, at 50 pounds per square inch

the same soil cement possessed only 4 times the shearing strength of

the sand-asphalt. Thus, it can be concluded that the exclusive use of

the unconfined compression test has shortcomings because it compares

shear strength at zero normal stress. Hence internal frictional effects

are omitted and are not accounted for suitably.

Figure 22 also shows the Mohr envelopes obtained for the soaked

f oamed asphalt and emulsion asphalt mixtures stabilized with k.2

percent asphalt and molded at approximately the same molding water

content. As shown, within experiments,1 error, there is no significant

difference in the Mohr envelopes obtained.

Figure 23 illustrates the Mohr envelopes obtained for the foamed

and emulsion asphalt mixtures compared at the same total dry unit weight.

8o

It shows cured cohesive strengths of 25.2 and 11.9 psi for the foamed

and emulsion asphalt mixtures. When compared to 11.9 psi, the foamed

a sphalt mixture represents an increase of 110 percent over the emulsion

asphalt mixture. At a normal stress of 50 pounds per square inch, the

some increase is only 43 percent. Thus a comparison of the foamed

and emulsion asphalt mixtures shows the superiority of the foamed asphalt

mixtures. This superiority, however, is not as great as that found by

Haas (1963).

Figure 24 shows a comparison of the Mohr envelopes obtained by

stabilizing the McGinn pit sands used with 4.2 percent emulsion asphalt

aid with 5 percent MC3 cutback asphalt. Cohesive strengths of 11.0

and 12.1 psi for the emulsion and the MC3 cutback show no significant

difference in the shearing strengths at zero normal stress. However,

friction angles of 34.4 and 30.6 degrees show a slight difference of

3.8 degrees. This difference is of the same magnitude of differences

recorded in Series 1A where the McGinn pit sands used by Haas and by

the author showed a friction angle difference of 5-4 and 2.9 degrees.

This difference was probably caused by a difference in the grain size

distribution of the two McGinn pit sands used. Because the McGinn pit

sand used by Haas was less uniform and contained more fines than that

used by the author, it is believed that his sand exhibited more

interlocking of particles and thus a greater friction angle. Finally,

for the McGinn pit sand used in this investigation, the above evidence

shows no significant differences in the cured shear strength of this

sand, stabilized with either an MC3 cutback asphalt or an SSI water

emulsion asphalt.

8l

Variation of Cohesion with Percent Asuhn.lt, Filler to Asphalt Raticy

Percent. Pass Inc: No. 200 Sieve, and with Total Dry Unit Weight.

General. The above-mentioned factors affecting cohesion are

illustrated in Figure 3; 13 and 25. In the following paragraphs,

discussion of results is based on the apparent trends and is not based

on the individual points obtained.

Variation with percent asphalt. For the gradations of the McGinn

pit sand investigated and for the percentages of asphalt used, Figure 13

shows cohesion to increase with increasing asphalt contents for a given

film thickness. For the gradations compared at an estimated film

thickness of 5*^ microns, Figure 13 shows a slight increase in cohesion

from 18 to 27 psi. This represents an increase of 50 percent for asphalt

contents ranging from 7-6 to 12.0. These same gradations compared at an

estimated film thickness of 2.7 microns showed cohesion to increase 100

percent for asphalt centents ranging from 3-8 to 6.0 percent. Thus, for

a given film thickness, cohesion was found to increase with increasing

asphalt contents; however, for the two film thicknesses used, a decrease

in cohesion was observed for the same gradation (See Figure 25). It is

believed, however that a greater range of film thickness would have

established the well known curve giving a maximum cohesion at an optimum

asphalt content (film thickness). The increase in cohesion with

increasing asphalt content at the same estimated film thickness is

believed to be caused primarily by increasing ratios of contant area to

surface area. This is substantiated in work done by Endersby and Vallerga

(1962). Because this ratio increases with increasing fines, and because

this ratio is an inherent property of the sand itself, it follows that the

82

increase in cohesion is caused primarily by increasing fines.

Although Figure 25 is incomplete, it appears that it is in agreement

vith work done by Rice and Goetz (1949) which showed that the addition

of filler caused the optimum asphalt content to increase. Although

incomplete the trends on Figure 25 were curved because of previous

work done by Haas (1963). This work showed the plot of cohesion

versus total dry unit weight or percent asphalt to be curved and to

have a maximum cohesion at an optimum asphalt content or an optimum

total dry unit weight.

Variation with filler to asphalt ratio. Work done by Kallas,

Puzinauskas and Krieger (1962) showed cohesive strengths to increase

with increasing viscosity of the filler-asphalt binder. This work

showed the viscosity of the filler-asphalt binder increased with increasirg

filler to asphalt ratios.

This investigation substantiated work done by the above-mentioned

authors as cohesive strengths were found to increase with increasing

filler to asphalt ratios. As shown in Figure 13, cohesive strengths

increased from 17-0 to 43.0 for respective filler to asphalt ratios of

O.36 and 2.42. This represents a relatively large increase in cohesive

strengths of 153 percent.

Variation with percent passing No. 200 sieve. Work done by Stewart

(1962) and Rice and Goetz (1949) showed compressive strengths to increase

with the percentage of fines passing the No. 200 sieve.

This investigation substantiated work done by the above-mentioned

authors as cohesive strengths were found to increase with increasing

percentages of fines passing the No. 200 sieve. For the gradations of

• • . .( ■ : ' ; • -r ' v ' ■: w:

..■ -'l ... j: ' ’ ■ • ;w- :

mob ‘ : - .

: ; v ' • ■' -;•1: lo /, ;v '■

83

the McGinn pit sand compared at a film thickness of 2.7 microns, Figure 3

shows cohesion to increase from 21.5 to ^3-0 psi, an increase of 100

percent. These same gradations compared at a film thickness of 5-^ microns

showed an increase of only ^8 percent, from 17*2 to 25-5 psi. Thus for

the film thicknesses used, percentages of fines passing the Wo. 200 sieve

are more effective in increasing cohesive strengths at lesser film

thicknesses, that is, increases in cohesion due to increasing fines

decreased with increasing film thicknesses. However, if a greater range

of film thicknesses had been used, it is believed that the effectiveness cf

fines to increase cohesive strengths, would have reached a maximum at an

optimum film thickness. Thus, any film thicknesses both above and below

this optimum would show the fines to be less effective in contributing to

cohesive strengths.

Variation with total dry unit weight. Referring to Figure J,, Series 1 ,

it can be seen that for the original McGinn pit sand stabilized with U.2

percent foamed asphalt and with varying molding water contents, cured

cohesive strengths increased l60 percent. This increase is believed to have

been caused primarily by increased densities. Cured cohesive strengths

increased from 12.0 to 31-2 psi for respective total dry unit weights of

98.3 to 103-2 pcf and molding water contents of 3 to 15 percent. For

the McGinn pit sand with altered gradations, cohesive strengths increased

67 percent for an estimated film thickness of 5-^ microns. Cohesive

strengths increased from 17*2 to 27.0 psi for respective total dry

unit weights of 97*6 and 108.0 pcf. For the same gradations compared

at a film thickness of 2.7 microns, cohesive strengths increased 100 percait.

: •. , ■' - : ' :.

■

.

1

Cohesive strengths increased from 21.5 to 43.0 psi for respective

total dry unit weights of 96-3 to 102.1 pcf.

Thus, for the gradations of the McGinn pit sand investigated,

Figure 3 shows cohesion to increase with increasing total dry unit weight

for a given film thickness (Series Nos. 1, 2A, 2B and 3)l however,

cohesive strengths, compared at the two film thicknesses used, decreased

with increasing total dry unit weight (See Figure 25)- The former

finding is substantiated by work done by Haas (1963). For one gradation

of the McGinn pit sand stabilized with k .2 percent asphalt or with the

same film thickness, he found cohesion to increase with total dry unit

weight. However, as reported by Haas no sharp increase in cohesion due

to changes in total dry unit weight was observed.

The latter finding is substantiated by work done by Haas in which he

found cohesive strengths to reach a maximum with increasing total dry

unit weights and increasing asphalt contents or film thicknesses. In

this particular investigation, the decrease in cohesive strengths for

the two film thicknesses used is believed to represent the downward

trend after the maximum point on the cohesion versus total dry unit

weight curve.

Variation of Angle of Internal Friction with Percent Asphalt. Filler

to Asphalt Ratio, Percent Passing No. 200 Sieve,and with Total Dry

Unit Weight.

For the gradations of the McGinn pit sand used and for the

percentages of asphalt used, Table VI and Figur® 1^ and 15 show that

friction angles do not vary significantly with any of the above-mentioned

variables. Also from the plotted points, no definite trends were observed.

j[ ■ K.\.. v' v: >''I,.!.'

■

85

Stress-Strain Relationships

Figure 26 shows the stress-strain curves obtained by Haas for the

sand-asphalt series at optimum asphalt content, with a 20 psi confining

pressure (full lines). The stress-strain curves for the foamed and

emulsified asphalt at approximately the same molding water contents are

also shown (dotted lines). The initial tangent moduli of deformation

are shown on each curve and illustrate again, on a comparative basis, the

superior shear strengths of the cured, foamed asphalt mixtures in terms

of a higher failure stress at reduced strain. As reported by Haas (1963)

the modulus of defomation for the cured, foamed asphalt mixture was 27,300

psi, as compared to 11,000 psi for the cured MC3 series, 7j700 psi for

the cured 'wet' hot-mix, 7^000 psi for the hot-mix, and 5>000 psi for the

soaked, foamed asphalt and the soaked 'wet' hot-mix. Tangent moduli of

21,lOO psi and 8,300 psi were obtained by the author for foamed and emulsified

cured sand-asphalt specimens. Thus the cured strengths obtained in

this investigation for the emulsion asphalt sand specimens are comparable to

those obtained by Haas for a similar sand stabilized with either a MC3

cutback asphalt or hot-mix asphalt.

It must be emphasized that the reported values of tangent moduli are

approximate only and are based on personal judgment.

' r

■

■

-

■ ■: • W

i

■ >si •' « 6T‘-.3 -JCT: a - ?,-«.• ,'Jao ‘igatt

CHAPTER VI

CONCLUSIONS AND RECOMMENDATIONS

C onclusions

Conclusions presented are limited because of the scope of this

investigation. These limitations include the use of only one sand

with varying gradations, the use of only two types of asphalt and the

use of arbitrary curing and soaking conditions that may not be represent¬

ative of in situ field curing and soaking conditions.

From this laboratory investigation and the literature review,

the following conclusions were made.

(1) For the original McGinn pit sand stabilized with 1.2 percent

foamed asphalt, increased molding water contents up to 15 percent result

in increased cohesive strengths. These increases are considered due to

increased densities through better asphalt distribution. However, it is

realized that excessive molding water contents which result in decreased

densities would probably cause the cohesion to decrease.

(2) There is a minimum required mixing water content necessary to

assist in distribution of foamed asphalt in sands. For the McGinn pit sard

used, this minimum required mixing water content is approximately 7 percent

(3) For the original McGinn pit sand stabilized with 1.2 percent

f oamed asphalt, alternate cycles of wetting and during do not significantly

affect the final cured shear strength.

(1) The original McGinn pit sand stabilized with 1.2 percent foamed

asphalt shows very significant losses in cohesion upon immersion. This

. to l

■' : • ■- ' "j y - -> i ■. ris 1 c

* " /' ■' ■••>.! 0'.'- -.9 ■ 0i t : ■ -/q ,; q I".:;-' JO ,

immersion however dees not affect the angle of internal friction.

(5) The addition of limited amounts of fines is beneficial to the cured

strength of poorly graded sand foamed asphalt mixtures. It is realized

that a maximum upper limit of fines exist and that the addition of fines

above this limit is not beneficial to cured strengths. For the gradations

of the McGinn pit sand stabilized at the same estimated film thickness,

significant increases in the cured cohesive strengths were observed. The®

increases are believed caused by the following:

(i) by increasing percentages of fines passing the No. 200

sieve. However, for the two film thicknesses used,

these increases decreased with increasing film thicknesses.

(ii) by increasing ratios of contact area to available surface

area. This ratio is believed to increase with increasing fines.

(iii) by increasing filler to asphalt ratios, which increases

the viscosity of the filler-asphalt binder.

(iv) by increasing total dry unit weights.

(6) For a particular gradation of the McGinn pit sand stabilized

with varying film thicknesses, there is a maximum cohesion at an optimum

total dry unit weight. Also there is a maximum cohesion for an optimum

asphalt content; this optimum asphalt content, however, is believed to

increase with the addition of fines.

(7) For the gradations of the McGinn pit sand used and for the

percentages and film thicknesses of asphalt used, friction angles do not

vary significantly.

88

(8) Cured sand-asphalt specimens stabilized with foamed asphalt cement

have superior shear strengths when compared to similar mixtures using a

water emulsion asphalt. The very similar cohesive strengths of immersed

specimens of foamed asphalt and water emulsion asphalt mixtures, are

thought to results from the asphalt binder providing similar cohesive

bonds in each of these cases.

(9) For the original McGinn pit sand, there is believed to be no

significant difference in the cured strengths of specimens stabilized

with either an SSI water emulsion asphalt or an MC3 cutback asphalt at

the same molding water content.

(10) A very small change in the gradation of a poorly graded sand

can cause very significant increases or decreases in the cohesive strengths

of foamed asphalt mixtures compared at the same asphalt content. Thus,

there is a very definite need for field control of gradation requirements.

(11) The altered University of Alberta laboratory installation is

adequate to foam asphalt. The addition of asphalt by weight, as done

in the University of Alberta pugmill is believed to be more accurate than

the addition of asphalt by volume as done in the conventional foamed

asphalt laboratory installation.

Recommendations

From the laboratory investigation and literature review, the following

recommendations are made:

(l) It is recommended that accurate conditions of field curing and

soaking be determined. From this more accurate laboratory curing and soaking

procedures could be used. This would result in cured and soaked strengths

that are more representative of field strengths.

. .

• - ■ .

. ■ / . 0 7J' [II '

' : • ' f*ou i 7" . bo -.o- : r

89

For example, curing of specimens in the laboratory should be allowed only

from the top instead of from all sides.

(2) Various amounts of the same filler used in this thesis should be

added and very carefully combined with asphalt. The viscosities of the

various mixtures of filler-asphalt binder should be determined. From this

more definite conclusions could be made regarding shear strengths versus

Ihe viscosity of the filler-asphalt binder.

(3) A comparison of cured shear strengths of the same sands used in

this investigation, at optimum asphalt contents should be made. From this,

the effects of gradation could be compared at optimum asphalt contents. Also

the results could then be compared to those obtained in this investigation.

Thus more constructive comments could be made about the comparison of

shear strengths at the same estimated film thickness. That is, the procedure

of increasing the asphalt content with increasing fines according to Hveem's

surface area constants could be discussed more constructively. The question

"Is the addition of asphalt according to Hveem's theoretical surface area

method a practical one from the point of view of staying close to the

optimum asphalt content for the various gradations used." could be

answered.

(4) Any future investigation involving film thicknesses should be based

on more direct methods of determining surface areas. Therefore, work should

be done on surface area measurements.

• ■ ' % - \n \ •■ , y'.l ;..t:vj 13 s

■ ' • i: ; ■ :i f f ■ ■ : -' '?, .. . ‘ o FiJ.flu.fO:* < o.i i«V ' ' (S) '

■

■

.

90

BIBLIOGRAPHY

Balmer, G., "A General Analytic Solution for Mohr's Envelope", Proc. of ASTM, 1952.

Bishop and Henkel, "The Measurement of Soil Properties in The Triaxial Test", 1962.

Borgfeldt, M.J., and Ferm. R.L., "Cationic Mixing Grade Asphalt Emulsion", Proc. HRB, 1962.

Campen, W.H., Smith, J.R., Erickson, L.G., and Mertz, L.R., "The Relation¬ ship "between Voids, Surface Area, Film Thickness and Stability in Bituminous Paving Mixtures", Proc. AAPT, 1959*

Csanyi, L.H., "The effects of Asphalt Film Thickness on Paving Mixtures", Proc. AAPT, 1948.

Csanyi, L.H., "Foamed Asphalt in Bituminous Paving Mixtures',' HRB Bulletin

160, 1957.

Douglas, A., and Tons, E., "Strength Analysis of Bituminous Mastic Concrete" Proc. AAPT, 1961.

Endersby,V.A., "Fundamental Research in Bituminous Soil Stabilization," Proc. HRB, 1942.

Endersby, V.A., and Vallerga, B.A., "Laboratory Compaction Methods and their Effects on Mechanical Stability Tests for Asphaltic Pavements" Proc. AAPT, 1952.

Goetz, W.H., "Comparison of Triaxial and Marshall Test Results" Proc. AAPT, 1951.

Goetz, W.H. and Schaub, J.H. "Triaxial Testing of Bituminous Mixtures" ASTM STP No. 252, 1959.

Griffith, J.M., and Kallas, B.F., "Influence of Fine Aggregates on Asphaltic Concrete Paving Mixtures", Proc. HRB, 1958*

Haas, R.C.G., "Triaxial Shear Strengths Characteristics of some Sand-Asphalt Mixtures Unpublished Master of Science Thesis, University of Alberta, 1963.

Herrin, M.H. and Goetz, W.H., "Effect of Aggregate Shape on Stability of Bituminous Mixes", Proc. HBR, 1954.

Holmes, A. "The Effect of the Nature and Particle Size of Granular Aggregates on the Properties of Asphalt-Closely Graded Aggregate Mixtures", Proc. AAPT* 19^7*

' '

• -

. ' '( r.&iA K's&l'iu3 , ^L..,. r‘ i-: ;'.'3

'

.'<$1 , - t rl a i

■ A ." t;

Hveem, F.N., "Use of the Centrifuge Kerosene Equivalent as Applied to Determine the Required Oil Content for Dense Graded Bituminous Mixes", Proc., AAPT, 19li2.

Kolias, B.F., Puzinauskas, V.P. and Krieger, H.C., "Mineral Fillers in Asphalt Paving Mixtures", HRB Bulletin 329> 1962.

Knowles, J., "Soil-Asphalt Stabilization with a Lime Additive", Unpublished Master of Science Thesis, University of Alberta, 1962.

Jones, V.C., "Durability Tests on an Asphalt Stabilized Sand ", Unpublished Master of Science Thesis, University of Alberta, 1962.

Lambe, T.William, "Soil Testing for Engineers", John Wiley & Son Inc., 1951

Lefebvre, J., "Recent Investigations of Design of Asphalt Paving Mixtures" Proc. AAPT, 1957.

Mack, Dr. C.M. "Physical Properties of Asphalt inThin Films", Industrial and Engineering Chemistry, March, 1957*

McLeod, N.W., "An Ultimate Strength Approach to Flexible Pavement Design," Proc., AAPT, 195l».

McLeod, N.W., "Void Requirement for Dqnse-Graded Bituminous Paving Mixtures ASTM, STP 252, 1959.

Pauls, J.T. and Goode, J.F. , "Application and Present Status of the Immersion-Compression Test" Proc. AAPT, 19^7*

Pennell, D., "Structural Comparison of a Sand-Asphalt with a Soil Cement" Unpublished Master of Science Thesis, University of Alberta, 1962.

Rice, J.M., and Goetz, W.H., "Suitability of Indiana Dune, Lake and Waste Sands for Bituminous Pavements," Proc. AAPT, 19^9-

Sommer, A., "The Flow of Thin Bituminous Films" Proc. of the Second Annual Conference of Canadian Technical Asphalt Association, 1957*

Stewart, J.A., "An Investigation of the Strength Properties of Sand- Emulsified Asphalt Mixtures and the Grading Characteristics of the Untreated Sands", 1962.

White, R., "Aeration and Curing in Asphalt Sand Stabilization" Unpublished Master of Science Thesis, University of Alberta, 1962.

APPENDIX A

SAND CLASSIFICATION DATA

Standard AASHO Compaction

Moisture Content of Air-Dried Sand

Sieve Analysis

Hydrometer Test

Specific Gravity Test

Bagnold Peak Diameter

UNIVERSITY of ALBERTA

DEP’T- of CIVIL ENGINEERING

SOIL MECHANICS LABORATORY

COMPACTION TFQT

PROJECT .

SITE m : in ; it . ...

SAMPLE , „

LOCATION HOLE DEPTH

Trl

TECHNICIAN DATE

3l Number

c 0

Mold No-

Wt-Sample Wet + Mold a 1702.0 1/7) < .7 17 7:' .0 177 ;. 1 177'. .6 17 7 . -C ■>- CO c

— c 0>- Wt- Mold I.J»21.0 . lh; 1.0 1 h2.1.0 lh21.0 .Lho 1 .0 Wt- Sample Wet hi. 6 vhp 0

T31.Q 72.1 ’lh .6 -;oo.l $ C W Volume Mold .0-0726 .00726 .00 226 . 0( 7Pn . 00'72 6 .07726 . ()(1 (9 6

Un

it

Pet;

Wet Unit Weiaht lb-/ft3 11 - - 8 103-9 1 06 . 1, I07.O 1 ( >r7 /7 1: )Q ,

Dry Unit Weight Ib/ft-^ 9r>.h <-17. h Q';.i4 Qt:.h • 96. h 06.1 o6.q ♦- C " c c a

Container No /ip; r,r>o V66 021 011 Wt- Samole Wet + Tare 172.IQ l00.1 it 1 80. r>8

r

1 67 . LP 16o. 18. 177.ho 1 q •; 10 0 Or Wt-Sample Dry-fTare l6<4. h7 183. 6-> 117.6.1 179.7-. 167.68 1h2. 18 0 c w **z Wt-Water .:Tn 8 . 1 2 in. Lq . io.ni Q 6l G h *2 i i 70 in qo 3 t

w. W <V

Tare Container 71 Q'-i £0 ah 66 .hi 6q . Lq 71.19 68.00 60. hO

Wt- Dry Soil RPi n4 1 18.71 1 1 3 . - 2 88 2 8L 06 07.68 C) -1 '■vQ ill- / . SO Moisture Content 8.7 6 8 81 ft- 87 10 -Q 12.0 1 T L

Method of Compaction_

7- ■ ■• < ' 1 (‘1 - • ..

Diam. Mold 2.00

Height Mold —:—

Volume Mold (,0706

No- of Lavers h

in.

Blows per LayerJ_

Ht- of Free Fall 2s

Wt- of Tamper_2: aii

Shape of Tamping Facex

Description of Sample_

JLmL

; 1-La. 2.66

Non Plastic

.1., =

A~2

Rem arks pp > .1. uri—av£xai;..

l.-triAis.,

10 15 20

Moisture Content %

DEP’T- of CIVIL ENGINEERING SOIL MECHANICS LABORATORY

COMPACTION TEST

PROJECT ” ■ SITE Me Inn SAMPLE , ....

LOCATION 1 ],:n ft;; ..

HOLE DEPTH TECHNICIAN [■ . V ,r DATE]'

Trial Number 1C . .. 12 1)4

c ^ O

Mold No-

Wt-Sample Wet + Mold n178].C; ; i’<'7,6 1780.8- 1791.0 1792.8 1797.0 1790.1 -C *- O’ c

— c ._ Wt- Mold 1 Ji21.0 1.0 1.1*81.0 1.1*23 .0 37(23 .0 3923 .0

Wt- Sample Wet 5< .. 66. ■ - (19 R ,lX’ .0 170.0 .71.0 T/ll.O 17)4.1 W Volume Mold c. . ft. .00726 .OC .00786 .00726 .00726 .00726 .00726

Un

it

Det

f

Wet Unit Weiaht lb/ft3 1.11 .2 n 2.9 112.8 3 3 1.8 317.8 .

Dry Unit Weight Ib/ft-^ 96 Ji . C.7-.U 07.7 . -QJ- /... C

c c o

Container No A3 sn Vi 8 s6 VSO v66 0 s6

Wt-Sample Wet+Tare i 6q . i 6 iflP.L-1 17L .08 loo .r,8 '107.1 8 i9q -2 168 of) o - O n Wt-Sample Dry-FTare io6 jh 167.67 1 A Q 21 •18n. 78 179 - 77 171 86 1 0 2 f)i *> C Wt- Water cm 12.12 11.87 3 8.27 .17.78 17.'16 1*4.86 =3 E

40 (V

o

Tare Container . 0i.6 L 6 -J49 68.78 66.8' 1 . 6l.)iQ

Wt- Dry Soil rn> . . 97 09 ,U.7,29 ini np 90 i 0 SQ Moisture Content ir’ 0 15.7 i 16.6/ i 6. C

j.9(

-- ■■ ■■ -" -— —-- ^ '■ -

m„w. Ilnli \A/i - "/fit IVI U A Vim »

% Opt- Moist- =.

N— N -O

I

.c o a> 5

Method of Compaction

Diam. Mold _

Height Mold -

Volume Mold_

No- of Layers_

Blows per Layer

Ht- of Free Fall Wt- of Tamper_

Shape of Tamping Face.

Description of Sample..

Remarks

M oisture

DEP’T* of CIVIL ENGINEERING


COMPACTION TEST

PROJECT .

SITE SAMPLE LOCATION HOLE DEPTH TECHNICIAN . W1 - DATE

Trial Number , .1 0 ?( '

c o

Mold No-

Wt* Sample Wet + Mold 1 707 0 . 1 . 1 ) 1805.2 l8:i ] . 1 1809.5 -C CT> C

— c Wt- Mold 1U2.1 .0 ll21.0 I)i21.( ill1.0 : .C 11-121.0 i Up. 1.0

Wt- Sample Wet V. i 1 1 • 0 . ) _ 1Q(). 1 ■88.7 ? b

V.

*— c oq

Volume Mold cu. i . .00720 .00726 . 072 ( .00726 .00726 .00726

Wet Unit Weiaht lb-/ft3 117.6 ll6 - 6 118.0 116.8 118.5 118.1:

Dry Unit Weight lb/ft-3 2 *7. 6 Q7.ii c iQ c .. -| 00 5 1 ( ]0- P. .

c " c c o

Container No F3 1 .> Li Lc^C £ PPCl GPP V51 r. t

Wt-Sample Wet + Tare.-m 1 IV] i<)h.50 iaa.ofl in . a ■ T-> 1 II 18 l o,L.

S= Wt-Sample Dry-fTare T 1 L8.h8 1 £0 07 1 IQ 16c . 67 177-7b 177.19 1 n°- U6 . €) C w. * Wt- Water IS .12 15.19 20.17 19.11 19.($'-< 15.75 18.16 3 b ♦- w- Tare Container 60. LO . 7 .1 .91 69. I| 9 , ^ r v c 69.85 59.92 co a?

Wt- Dry Soil - S8.C8 O — ,-yQ - - , - 102.16 100.18 105.7] 87.5-11 107.5U

5Q Moisture Content lv.2 17.8 19.7 1°. 1 18.6 18.0 17.8

1962

CT V

5

Max* Unit Wt- =_"/ft’

Opt- Moist- s_%

Method of Compaction

Diam. Mold

Height Mnlri

Volume Mold

No- of Layers

Blows per Layer

Ht- of Free Fall

Wt- of Tamper

Shape of Tamping Face

Description of Sample

Remarks

Moisture Content %

UNIVERSITY of ALBERTA PROJECT THIS i . SITE McGinn Pit; North

DEPT- of CIVIL ENGINEERING SAMPLE Sand

SOIL MECHANICS LABORATORY LOCATION Stonv Plai] t. Alberta

MOISTURE CONTENT HOLE DEPTH TECHNICIAN A.D.L. DATE Nov.

KQ|g-Ng:

Depth

Somple No

Contoiner No X K7 228 302 E36

Wt Sample Wet t Tore 271*6.1 235.23 157.69 223.17 197.10

Wt- Somple Dry -t- Tore ,27,g.t?. '3^.35 157.12 pop -27

Wt- Water 7.9 0.88 0.57 .iU3q.. 0.65

Tore Container 876.3 32.60 00 £ cL » O 1 30.28 lo.oU

Wt- of Dry Soil 1861.9 201.75 I3I4.I45 192.09 -156. .Ill

Moisture Content usc/o 0.42^ 0.444. ,Q.42l 0.42^ ),4 2j

JtMg-Mfl.

Depth

Somple No-

Container No 1. Moisture Cent. ;nt of Air-Dried Sand

Wt Somple Wet -r Tore

Wt- Somple Dry + Tore

Wt-Woter

Tare Container

Wt of Drv Soil

Moisture Content ar%

Hole No-

Depth

Somple No-

Contoiner No-

Wt Sample Wet-r Tore

Wt-Somple Drv+Tore

Wt Water

Tare Container

Wt- of Dry Soil

Moisture Content u/%

Remarks-.

UNIVERSITY of ALBERTA DEP'T. of CIVIL ENGINEERING SOIL MECHANICS LABORATORY

SIEVE ANALYSIS

PROJECT. T1TECHS SITE inn I'It North SAMPLE Liatifl LP.QA.T tON Lll.QQj' J, .1 a 1 n, i\ I he r l.a, HOLE- DEPTH.. TECHNICIAN A.D.L, DATE t>

Total Dry Weight

of Sample

Sieve

No.

Size of Opening Weight Retained

gms.

Tptal Wt. Finer Than

gms.

Percent Finer Thon

Vo Finer Than wa^is Orig.

Sample Inches Mm.

Initial Dry Weight Retained No. 4

Tare No.

.

V

Wt. Dry + Tare Tare 3/4 1910

Wt. Drv 5/8 9-52 4 •185 4-76

Passing 4

Initial Dry Weight Passing No-4 Tare No

10 •079 2-000 20 •0331 •840 0.0 2??Q 0 C.C.U. J 4 . 100.0$

Wt- Dry + Tare 40 •0165 •420 8.0 2221.0 qq.6$

Tare 60 •0097 •250 861.0 1360.0 61.0$ Wt. Dry 100 •0059 •149 1025.0 335-0 dp. Tf°

2 0 0 •0029 •074 208.0 127.0 5.7$ Passing 200

Description of Sample Tine, uniform rounded quartz POOriy nrc-dsd -and.

Method of Preparotion Air-Dried Semple

Time of Sieving 5 hrvnr^

-l-T^y T..fl:horl -HTprv.lfih t.Vlf» OPVP t.bpn

..pyeiL.dr-led and dr.y~ajjg.YPd. tftr.PHEfo .ftfeg... Remarks complete nest of sieves,_

Represents the average of three sieve analysis ♦

]1 Gravel Sizes Sand Sizes

100 >" T 4 r Sieve

*4 ije *10

41 * *

90

80

70

60

a»50 c

U_ 40 c 4> o30 o ^zo

10

0 TO “W

Note: M-IT- Grain Size Scale

00 10 Grain Size-Mm..

UNIVERSITY of ALBERTA DEPT of CIVIL ENGINEERING SOIL MECHANICS LABORATORY

GRAIN SIZE CURVE

PROJECT M'P KR1 <V ...

SITE McGinn Pit North SAMPLE S&nd . LOCATION R+ftny Plnln A '1 1 .

HOLE DEPTH TECHNICIAN A.i PATE

to

«

CO

>* o o

CO © (0 k~ o o o

to ® im

CO

© > o w.

CD CVI

fO

/ £

o o o

o o 6

o o o o o o o O <J> CO S CO If) — U0l|l J8UUJ

o rO

o 04

4U30J9d

E £ E E

VO VO cu rH

ii II 2 2 ©

d a o

o 6

m ® w «

_ E 6 E

2

I Q> IM

CO

o w o

o2

G L

•H a: O ,Q o

© E a> ac

JL a

4- a J c

L t: t5 s

©

o u

CO

© N

CO

a t-

o

w o

PROJECT SITE SAMPLE F.TNK SAND LOCATION S.tQ^Y.l'leiT’. f Al.Di-rrt.ii . HOLE DEPTH TECHNICIAN ^ White. DATE 1A June...

UNIVERSITY of ALBERTA DEP'T- of CIVIL ENGINEERING SOIL MECHANICS LABORATORY

SPECIFIC GRAVITY

Somple No.

Flask No.

Method of Air Removal VA. bus

Wb+w+s 787. ;Q Temperature T pa

.'.Tm

Evaporating Dish No. nw 1 Wt. Sample Dry + Di$h Rm

Tare Dish pm 228.As

Ws i£L .U7~>A.

Gs 2.66

Wb*w+s 8 Weight of flask

Wb-i-w s Weight of flask s Weight of dry W

+ water ♦ sample at T°.

+ water at T° (flask calibration curve),

soil w

Ss ^Specific grovlty of soil particles «■ -tfh'+w+I

Determination of Ws from wet soil 6ample:

Sample No. Sample No.

Container No. Container No.

Wt. Sample Wet+Tare Wt.Test Sample Wet + Tare

Wt. Sample Dry + Tore Tare Container

Wt. Water Wt.Test Sample Wet

Tare Container W,

Wt- of Dry Soil

Moisture Content W %

Description of Sample-. Fine Sand

AASHO Classification - A--

Sample oven-dried init.TP.1Ty

Remarks;

A-10

Bagnold Peak Diameter

The Bagnold relationship (peak diameter included) is a method of

expressing the grading of sand; its principal advantage is that it

reasonably defines the complete grading of a sand instead of defining

it only at several isolated points.

The Bagnold peak diameter is dependent on the Bagnold size distribution

lav which states that the frequency of occurence of a given grain size

decreases at a constant log rate for grain sizes coarser and finer than

the peak or model grain size. A theoretical plot of the Bagnold size

distribution law is shown in Figure 1A. Figure 2Ashews estimated Bagnold

curves based on three points for each gradation of the sands used.

However, it is realized that more points are required to give accurate

Bagnold peak diameters. Thus because of the inaccuracies in the

determination of the Bagnold peak diameter, discussion of shear strengths

versus Bagnold peak diameters for the various gradations of the sands

used did not seem justified.

One point worthy of mentioning here is that for the sands used many

divisions of grain sizes are required to adequately describe the Bagnold

peak diameter. Consequently the use of the Bagnold lav/ as a means of

expressing the grading of sands, seems to be impractical and therefore

not economical.

APPENDIX B

PROCEDURE USED TO OBTAIN DIFFERENT

GRAIN SIZE DISTRIBUTIONS

APPENDIX B

R-

PROOEDURE USED TO OBTAIN DIFFERENT GRADATIONS

The procedure used is summarized ns follows:

1. The air dried sand was placed in the top sieve of the 20

inch by 20 inch mechanical motor operated soil test siever.

2. The motor was then started and the sand sizes were sorted

using the number 30, 50, 100 and 200 sieves.

3. After each application of approximately 5 pounds of sand

it was necessary to clean the sieves. This was done using air and a

special brush.

b. Once the sieves were cleaned, they were placed in their

proper slots and the procedures described above were repeated.

5. After most of the remaining sand was sieved, it was necessary

to determine the exact grain size distribution of the portions of sand

retained on the number 50, 100 and 200 sieves. This was necessary

because many smaller particles, mostly silt sizes, clung to the bigger

particles of sand; thus the sand, retained on the number 50, 100 and 200

sieves, retained particles which were smaller than openings of these

respective sieves. The actual distributions were determined by performing

a wet sieve analysis on representative samples from each sieve. These

are shown in Figures IB, 2B and 3B.

6. After the grain size distribution of the sands retained on the

number 50, 100 and 200 sieves was determined, it was necessary to

recombine these sands with the original sand so as to obtain the

required grain size distribution. Grain size distribution curves that

would give 20, 15, 12, 9> 7 and 2 percent passing the number 200 sieve

were used.

'

.

.

• . -1

■ n' H'

B-2

A fev sample calculations are shown in Table IB. Table IIB shows the mix

proportions of oven-dried sands used to obtain ^500 gram samples of

the sands with various gradations.

f

UNIVERSITY of ALBERTA DEP'T. of CIVIL ENGINEERING SOIL MECHANICS LABORATORY

SIEVE ANALYSIS

PRoJ EC T Titrflr, SITE L .11! i l. ■ No rlli

SAMPLE fund LOCATION lon-.rMs.ln A 1 h, ■ r

HOLE DEPTH TECHNICIAN DATE .rmh. k/£

Total Dry Weight Sieve Size of Opening Weight Reta'ned

gms.

Total Wt. Finer Than

gms.

Percent Finer Than

%T:mer'TtRjn

of Sample N o. Inches Mm.

Initial Dry Weight Retained No. 4

Tore No Wt. Dry + Tare Tare 3/4 1910

Wt- Dry 3/8 9-52 4 •185 4-76

Passing 4

Initial Dry Weight Passing No-4 Tare No-

10 -079 2-000 20 •0331 •840

Wt- Dry + Tare 4XT30 •0165 •420 '■> Retained

Tare 63T5C •0097 •250 100.0

Wt- Dry l6[.a.2 100 •0059 •149 1CI9.5 ,14.7 8.2 91.8

200 •0029 •074 Zte“ hi .1 2.5 5*7 . Passing 200

Description of Sample i fo. i i -Ua, plus No. ICO sieve material

Method of Preparation T- - shed through I-io. 200 sieve, dried and then sieved

obtained from original sieving oi through next of sieves. dirty sand Remarks

Time of Sieving

Gravel Sizes -jtr 1

Sand Sizes Si

IOC

3.'

r 4 Sieve *4

Size it dt & dt dfc

*10 20 40 60 100 200

90

80

70

60

u10 D6Q5 Cy :

.ram

.ram.

4>50 C

u. 40

c a> o30 o

°"20

10

0 •o in Note: M IT- Grain Size Scale

100

Grain Size-Mm..

FIGURE 2 B

UNIVERSITY of ALBERTA DEP’T of CIVIL ENGINEERING


SIEVE ANALYSIS

PROJECT SITE McGinn Pit North . SAMPLE LOCATION Eton. 1 inl.nf Alberta.

HOLE DEPTH TECHNICIAN A.D.L. DATE Fob. h/ 0

Total Dry Weight of Sample

Sieve

No-

Size of Opening Weight Retained

gms.

Tptal Wt. Finer Than

gms.

Percent Finer Thon

%fWThpfv 8<J£ts Qf\fr'

Inches Mm.

Initial Dry Weight Retained No. 4 Tflro N| o

wt. Dry + Tore Tare 3/4 19 1 0

Wt. Dry 3/8 9-52 4 •185 4-76

Passing 4

Initial Dry Weight Passing No-4 Tare No

10 •079 2-000 20 •0331 •840

Wt- Dry + Tare 40 •0165 •420 Retainer

Tar e 6 0 •0097 •250

Wt- Dry 1CV7J* 100 •0059 •149 1037.h 100.0

200 •0029 •074 076.7 6r .7 .5-8 qU.2 Passing 200

Description of Sompie " •

-T.Vur. _ri.-YC ffir.; ■, ’i;,j_

_obtained fror critTv al rlevins of dirty sand

Time of Sieving

Method of Preporotion Washed, through

No. 20C sieve and then dried.

Remarks_

Gravel Sizes ,it

Sand Sizes Si

100

s: 4 ¥ eve

4 Size

*10 20%,Q *6Q .\qQ.*.2QQ

90

80

70

UI0 D60=-

Cu =-

.nwrv

.mm-

60

<u50 c

U. 40

c <u 0 30 o

^20

10

0 o-(5l

Note: M IT- Grain Size Scale

100 10 Grain Size-Mnv.

FIGURE -'B /

UNIVERSITY of ALBERTA DEPT of CIVIL ENGINEERING SOIL MECHANICS LABORATORY

GRAIN SIZE CURVE

SITE i nn I it. Nnrth SAMPLE ‘nr.rl LOCATION it nny lcin( A IYhm’Il. ...... . HOLE J1F.EJH TECHNICIAN A.P. JDAI.E

o o o

o o 6

o 6

m 0 w ♦- «

_ £ 6 E

 1 w 0)

H <L) -P a

©

o u <n

<n c o k. o

H-

« ♦- o Z

FIGURE mB

B-7

TABLE I B

SAMPLE CALCULATIONS INVOLVED IN THE COMBINATION OF TWO OR THREE SANDS OF VARIOUS GRADATIONS TO OBTAIN THE REQUIRED PERCENTAGES PASSING THE NO. 200 SIEVE

COMBINATION OF SANDS USED TO OBTAIN APPROXIMATELY l4. 5$ PASSING NO. 200

MATERIAL Material retained or passing various screens in grams

+No. 30 +No. 50 +No. 100 +No. 200 -No. 200

100 grams of original sand 27 grams of -100 +200 sand 13 grams of - 200 sand

0 20 64.5 9-8 25.4

5-7 1.6

13.0

TOTALS 0 20 64.5 35.2 20.3

SUMS OF TOTALS - 140.0 grams .

Screen No. Weight Retained

Weight Passing:

Percentage Passing

30 0 140.0 100$ 50 20 120.0 85.7$

100 64.5 55-5 39-7$ 200 35-2 20.3 i4.5 $

COMBINATION OF SANDS USED TO OBTAIN APPROXIMATELY 7.2% PASSING NO. 200

MATERIAL Material retained or passing various screens in grams

+ No. 30 fNo. 50 +No. 100 +No. 200 -No. 2 00

100 grams or original sand 0 20 64.5 9.8 5-7 6 grams of -100 +200 sand 5-6 .4

1.6 grams of -200 sand 1.6

TOTALS 0 20 64.5 15-4 7-7

SUMS OF TOTALS - 107-6 grams

Screen No. Weight Retained

Weight Passing

Percentage Passing

30 0 107-6 100.0$

50 20 87.6 8l. 5$ 100 14.5 23-1 21.5$ 200 15-4 7-7 7-2$

- ir ' ■■ ' I ' ■■■■■: ■ : : Vv

" X u ■ .

.. .*■ 1.L .

:j ■ :

■

•' l

B-8

TABLE II B

RELATIVE PROPORTIONS OF SANDS USED TO OBTAIN 4500

GRAM SAMPLES OF THE VARIOUS CALCULATED GRADATIONS

Mix No. 8 Mix No. 9

2915 grains of original sand 3215 grams of original sand

877 grains of -100 +200 sand 868 grams of -100 +200 sand

703 grains of -200 sand 418 grams of -200 sand

Mix No. 10 Mix No. 11 3700 grains of original sand 3930 grams of original sand

482 grains of -100 4200 sand 393 grams of -100 +200 sand

315 grains of -200 sand 176 grams of -200 sand

Mix No. 12 Mix No. 13

4185 grams of original sand 900 grams of -30 +50 sand 251 grains of -100 t200 sand 3600 grams of -50 +100 sand

67 grams of -200 sand

N.B. - Refer to Figure 4 in this Appendix for grain-size distribution curves.

3 :..i ......

APPENDIX C

MEASUREMENT OF SURFACE AREAS OF SOIL PARTICLES

Hveera's Theoretical Surface Area Constants

Application of Surface Area Constants to the Sands Used

Amount of Asphalt Used in the Various Sands

■

. ...

.m.; "

'

C-J

APPENDIX C

Hveem's Theoretical Surface Area Constants

As previously mentioned in the main body of the text, Hveem's

theoretical surface area constants were used as an indirect measure of

surface area. This method assumes perfectly spherical particles and is

therefore somewhat unrealistic for the sand used. However, as previously

mentioned because this was a preliminary investigation considering surface

areas, it was decided that the procedure of estimating surface areas using

Hveem's indirect method was satisfactory.

As given by the Highway Division of the California Department of Public

Works in their "Materials Manual" Volume I, the following are the surface

area constants used in this investigation.

Passing Particular Sieve Surface Area Constants

Passing Max. sieve size 2 Passing No. b sieve 2 Passing No. 8 sieve ^ Passing No. l6 sieve 8 Passing No. 30 sieve lb Passing No. 50 sieve 30 Passing No. 100 sieve 60 Passing No. 200 sieve l60

Application of Surface Area Constants to the Sands Investigated

For the particular application of the above surface area constants

to particular sands, it was necessary to multiply the surface area factors

by the percentages passing the various sieves. The sum of these multiplications

gave an answer in square feet per pound of aggregate.

The following is a sample calculation involved in the determination

of the surface areas of the various sands used in this investigation.

'

■

■

- /■' j .<■[.: V:. \ ■ ■ r livV '■:/ +■ >u

C-2

Mix No. 8

Passing Particular Surface Area $ Passing Product of °jo Passing in dec Sieve Constant Sieve of surface area Constants

Passing Max. Sieve 2 100.0 2

Passing No. H Sieve 2 100.0 2

Passing No. 8 Sieve k 100.0 k Passing No. l6 Sieve 8 100.0 8 Passing No. 30 Sieve Ik 100.0 lk.O Passing No. 50 Sieve 30 87.O 26.1 Passing No. 100 sieve 60 1+5-0 27.0 Passing No. 200 sieve 160 20.2 32.3

Sum of Product 115.!+

As listed the following are the sums of the above mentioned product

obtained for the various gradations of the sands used.

Mix No. 1o Passing No. 200 Sieve Sum of Product

8 20.2 115.1+ 9 11.5 102.7

10 12.3 93-1 11 9-^ 81*. 5 12 7.2 78.9 13 2.7 65.0

3 5-7 72.1+

Amount of Asuhalt used in the various sands

Once the surface areas of the sands were obtained, they were used to

determine the amount of asphalt necessary to compare results at the

estimated film thicknesses of 5.U and 2.7 microns. A film thickness of

2.7 microns was used because it was necessary to compare results to the

original McGinn pit sand used in this investigation. As shown on the

above tabulation, the surface area per pound of original sand (Mix No. 3)

was 72.H square feet. Based on this estimated surface area, the calculated

film thickness for an asphalt content of h.2 percent was 2.7 microns.

V d 3' . n:~ 1

■

ov:. 3 Of .a"/1' viJraff

znsv t u:c :rv^ anJ" io ax. tv.-'/'U/o -idt spaO

. , , ■■; i

. -j! ■ -svx;-■ ■ ■ ./»1 [■■■. -x. ■■■■••■ . 3: f r ■ . 'r ’ ■ ■ ,' ■:>

■

■. ' J . 3ix - ;T

C-3

The amount of asphalt required to give an estimated film thickness of

2.7 microns at varying gradations was in direct proportions to the estimated

surface area of the sand. The calculated asphalt contents for the various

sands used are shown as follows:

^Passing No. 200 Total Surface °]o Asphalt i» Asphalt Sieve Area Used Used

ti.Z- 2.7 Microns ><^ = 5.4 Microns

20.2 115.14 13-5 111. 5 102.7 5.98 12.0 12.3 93-1 10.8

9.1+ 84.5 4.95 9.9 7-2 78.9 9.2

5-7 72.4 4.20 8.4

2.7 65.0 3.80 7.6

APPENDIX D

DETAILS OF PREPARATION OF BATCHES. MOLDING OF SPECIMENS AND OF

APPARATUS AND TESTING PROCEDURES FOR TRIAXIAL TESTS.

.

: ■ iCidOOy i W.VX23S Q1A

D-l

APPENDIX D

Preparation of Batches

In Series 1A, the sand was mixed with 3,7, 11 and 15 percent water and

all mixes were left sitting for 15 minutes so as to obtain a uniform

dispersion of water. In the meantime, the asphalt was heated to 350° F

in a special Sta-Warm automatic electric heater (See Plate 5) and the

asphalt container (See Plate 5) was heated to 300°F. Once the water

was dispersed uniformly throughout the sand, the sand was poured in

the pugmill (See Plate l) and the asphalt was poured in the 300°F

pre-heated cylindrical container at 350°F. The asphalt container was

then moved into position (See Plate 5) • The quick coupler between

the asphalt container and the foamed asphalt nozzle was connected aril

the top of the asphalt container was screwed into position. The air

pressure hose was then connected to the quick coupler on the top of

the container. In the meantime, an assistant set the air pressure at

30 psi and set the steam pressure at 50 psi. Because the steam condensed

in the cold pipes it was necessary to prevent this water from dripping

into the sand. Once a dry steam of 50 psi was obtained, the valve (See

Plate 5) between the asphalt container and the patented nozzle (See Plate

l) was opened; also simultaneously the pugmill, revolving at 120

revolutions per minute, was started. The resulting mixture was allowed

to mix for various periods of time, depending on the mixing water present

in the mixture.

As arbitrarily done by Haas (1963)> it was decided to mix the batches

with less mixing water for longer periods so as to compensate for greater

amounts of mixing water and thus more uniform dispersion of asphalt with

D-2

the same amount of mixing. It was decided to mix the 3; 7> 11 and 15

percent mixes for respective periods of 2-jj, 2 and 1 3/1' minutes.

Once the mixing operation was completed the resultant hatches were

bagged and allowed to cool to room temperature before compaction started.

Molding the Specimens

All specimens were molded in a 2 inch by 1 inch mold (See Plate l)

and a reverse extrusion process employing a 3>000 pound Ashcroft

hydraulic press was used to advantage (See Plate 2). After the specimens

were extruded the required amount, the top of the specimen was trimmed

flush with the top of the mold. The final extrusion resulted in the 2.0

inch diameter by 1.0 inch long specimen. Immediately after this specimens

were weighted and some were measured.

The specimens were then placed in an oven at 100°F for 7 days. The

cured series were tested immediately after the curing period. The

immersed series, following curing, was placed in a large tank at room

temperature for ll days with one Inch of water covering the sample.

Apparatus and Testing Procedures for Triaxial Tests

The list of apparatus used in this testing program follows:

(1) A 3000 pound Ashcroft hydraulic compression testing machine was

used for extruding the samples from the compacting mold. (See Plate 2).

(2) Two 2.0 inch diameter triaxial compression cells (See Plate 3)

were used to test the specimens. Employing two cells allowed setting up

and preparing a sample while compression of another specimen was in progress .

(3) A Farnell loading machine was utilized for applying a uniform

strain rate of 0.01 inches per minute (See Plate U).

D-3

(M For application of the air over water lateral pressure, a

constant supply of air was available to introduce into the special Farnell

air chamber and then into the cell. It was necessary to adjust the

pressure a few times during the test in order to maintain the cell pressure

constant.

(5) All samples and ingredients were weighed on a ^000 gram capacity

scale (See Plate 5) with one gram as the smallest division.

(6) The load application to the specimen was measured by means of

proving rings. For most of the testing, a 1500 pound capacity ring. Wo. 783

was used and for the unconfined specimens a ^00 pound capacity ring Wo. 782

was used. The calibration curves supplied by the manufacturer were checked

and found to be correct.

Before testing the sand asphalt specimens, diameters, lengths, and

weights were recorded.

The specimen was placed on the base of the triaxial cell and a

rubber membrane was attached to the base and loading head of the cells by

means of 0-rings. (See Plate 2). The cell was then assembled and filled

with water. The oily piston was then inserted into the hole and allowed to

drop into the loading head. A final hand adjustment was required to assure

proper seating into the loading head.

The cell was then placed on the loading table of the Farnell automatic

tester and the loading yoke was brought into contact with the piston.

After connecting the pressure line, the cell pressure was applied using

the air pressure device of the Farnell machine. The recording dial of

the proving ring was then zeroed so that the measured load would be the

deviator load. A strain rate of 0.01 inches per minute was used because

of previous work done by Pennell (1962), Haas (1963) and Knowles (1962).

Leaks in the rubber membranes were detected immediately by a

significant drop in the cell pressure. Of course, when leaks developed,

it was necessary to dissemble the cell and reset with a new membrane.

Finally, loading was commenced with proving ring readings being

taken at percent strain readings of 0.1, 0.2, 0.3, 0.1, 0.5; 0.6, 0.8,

1.0, 1.4, 1.8, 2.2 and every 0.4 percent from then on.

At failure, that is, when the deviator stress dropped, loading was

discontinued, and samples taken from the cell were placed in an oven

at 105'°C.

The lateral pressures used were 80, 70, 60, 50, 40, 30, 20, 10 and 0

pounds per square inch with one specimen at each lateral pressure.

In some cases fewer lateral pressures were used because of an insufficient

number of specimens. In these cases the lateral pressures used were

distributed evenly between 0 and 80 pounds per square inch.

APPENDIX E

SAMPLE SHEETS

Sample Calculation Sheet

Sample of Deviator Stress Circles Obtained for Both

Foamed and Emulsified Asphalt Mixtures

Sample Data Sheets

E-l

APPENDIX E

Basic Equations Used for Calculations

Where A.C. = Asphalt Cement

Where G = Specific Gravity

VTotal

WTotal

Vs — Ws Gs

= Vs+ VA.0tVV

r Ws + WA „ t W volatiles A.C .

V W. A.C. - A.C.

A.C.

W^ ^ Weight of dry soil x percent of asphalt cement added

Use G A.C

Degree of Saturation

Sample Calculations

■=•1.03 and Gs = 2.66

Vwater

Vvoids

Vv

Vv

Volume of Specimen =• Cross-Sectional Area x Height

--rrd%. x height

=-3.Mx(l£5t*4>0\

— \2-'50 <zms>

Volume of specimen in Cubic cms.^ 12.50 x 16.39 = 205-0

Weight of soil residue asphalt = 3-22.0 gms ( h .2$=A.C ,<t>)

Total Dry Unit Weight = 322.0 x g2 h = 98 0 lbs/ft3 205-0

Weight of Soil - 322.0 _ ~ 1^2 “ 309 gms

Volume of Soil — 309 -- 2.16 cm3 2.66

Volume of Asphalt — 13 _ ^ q “ i7o3 " 12-6 cm

. r,:'.', • ; .v ■' . A

■ • 'V

.

Volume of Air Voids « 205 - (ll6+ 12.6) =. 76.k cm'

Volume of H^P -Weight of specimen after molding minus weight of specimen after drying

= 333-9 “ 322.0 = 11.9 crn^

Degree of Saturation = 15 M

e-3

T

Q is cv)

Q 2

< oo u>

N

2 2

O

j Q _ Lit

(fi

h 0 o-

10 U. fO

£2NiaV3.USi

w- ■1 UNIVERSITY of ALBERTA DEPT of CIVIL ENGINEERING

TPTAXIAL COMPRESSION

- RTTUMTNOIK MIXTURES

.SiIE-rVi STonV SAMPLE..

■ 1 1 .. Mach1ne Data

-Leonard Darnell Trowf/af 'fcfnrh,

-2rot//svp A/o, 7&r$_

. SPECIMEN

(XL..

mojEc.i-7

IQ£ATJQN_ ^T5 Ny PLAIN .JM-BfcRTA

HOLE _

-lIQHJjlSiAN *VPk. DATE JaKilfi/feA -D.EP.II1

Description of Sompte_.___ -..Said._floMiied ISi-tk noz ^ -foamed .teafaftpad /s.% mdd/w

Specimen Number

Lfltieral Pressure

Length Diameter

Aica

-Vo luma WC .—Sni.1—4- Vnlat i Ips fiftpr Comp

fairing Period.

JalL. Soil 4- Residue Asphalt

-laLfll—Volatile Content

60/751 aAQQ

JL&LL -2J21

ttlQ

22&,te

DATA

8Q&2/ j^ss: ?.o/'

JUJl

31ZBL

3247.

.4Qp£l JL22L

3/4'

xzs-.z. .225,6

JLnlfltile Content af Testing

Void Rati n

Decree of Saturation

Dry I'nit Weight

Loss t?£ Yol during, mring

fc- Pr ov Ring

Load lbs.

Str. Dial .001"

Str . %

Corr. Area

^>pc C/J <7fn Vo. 1

-J2 O CD 5.17 /oo 4 n./

-2J£> . <6 0.7

-2Q5 . IZ 0,5 /(Z> 0.4.

5QQ 24 0.4

S2Q . 32 OA .650 . 4C /.O 520 -2/5 S& /.A 3> 2.1 -Z24 -IQg 7Z /<8 3 Z3>

-Z?J 706> 27 5 74

. 155 . 703 104 ZA 5.2 <7 . 75Q -£9? IZO 5.0 5.27

7?/rs) O/o. 2

O o ?./7 -Bo . 4 O.i

250 ..A 0.7. 2550 vz. CD-7 2>2<? 16 04

A-Zci 74 02 -520 3Z OA 223.5. 40 / o 5ZO -300 36 \lA -5£L

Prov Ring Div .

. Load lbs.

Str , Dial .001'

Str. %

Corr. Area 0T CT-CC

70.2 * 7£ /■& 3.22 2o3 -2feS ZZ 3.24 3/7 237 \8&. 7<?o 104 j Zb 5Z5" 3Z2 74i

!<?? /20 2.o 327 fz3 242 .32h 72£l. . 36 3A .3.2*? 323 2.4Z,

[ Jagg 26//?? in / /o. 4$

0 O £> 3.(4 L£Z3 -£/jL -.20. -4 Q.l Z13- .g/g 2222 . e OZ 225. 2£c. 72 03 277 .M3 .3fP /£ 041

..£?4 -UB. . -24 <20 -gP 21 03

,542. L ^ 1.0 3.17 20 /7/ 5<e2 Situ !.A 3/6 2-/7 /77

.^ar 2360- 77 L6 3 >9 2/Or, na

CURING, SOAKING AND DIMENSIONAL DATA

FOR SAND-ASPHALT SPECIMENS

PROJK CT T 1 1 G | To date; mixed PV P> 8 tQ^T.

MATER I At (Vfc , HK p , f <n l/U^

- ■■-■■. — ■ ' ii-Mia.W y 1 i 1 _i -

MIXING TIME y Nl//7(J^S

£LHI 2Aoa{ O nl 16_C?° Y OAl_E COMPACTED Pt Ij 8. 1 <?

SOAKING compa sffort ,'/faA:Wd Pi-ncrVor-

AVE. DRY UNIT W f „ sampu.es _ TO after COMPACTION

AFTER Cl if? | no

joe-h IbsJft?

Ave. (f) SWELL, SAMPLES - TO

AFTER SOAKING

AFTER SOAKING

ADOITIONAL REMARKS f?rof>a'''?/dyx 5

32* ^//7, o)

4- A/^/aap) af ^ o) _

- 706 fori, d{-/fro ) fa {///,_(>')__

- s/c£> ^

—■Altered Mf Lsnn #,4_

-~4i7O- Z nfrr-egf

tf<r A/o.loO S/>W

Triaxial shear strength characteristics of some sands ...

Documents

Transcript of Triaxial shear strength characteristics of some sands ...