RAMMED EARTH: FIBER-REINFORCED, CEMENT-STABILIZED

by

ERIC WALTER SIMENSON

B.S., University of Colorado Denver, 2011

A thesis submitted to the

Faculty of the Graduate School of the

University of Colorado in partial fulfillment

of the requirements for the degree of

Master of Science

Civil Engineering

2013

ii

This thesis for the Master of Science degree by

Eric Walter Simenson

has been approved for the

Civil Engineering Program

by

Dr. Frederick Rutz, Chair

Dr. Kevin Rens

Dr. Nien-Yin Chang

Date: November 12th

, 2013

iii

Simenson, Eric Walter (M.S., Civil Engineering)

Rammed Earth: Fiber-Reinforced, Cement-Stabilized

Thesis directed by Assistant Professor Dr. Frederick Rutz

ABSTRACT

This thesis examines the use of cement and synthetic plastic fiber additives to

improve the strength of rammed earth walls. These additives can be an economical

solution to increasing the strength of rammed earth, which by itself is typically a low-

strength material when additives are not used. Cement can be easily incorporated into the

soil mixture and it adds strength and durability to the wall. Modern fibers, such as

polypropylene plastic, have high tensile strength, they are durable, and they can be

incorporated into rammed earth material. The fibers can add shear strength and flexural

strength to rammed earth material. Three mix designs were created: soil only, soil-

cement, and soil-cement with fiber. Based upon the results of this thesis, the use of

cement additive in rammed earth significantly increased the compressive, shear, and

flexural strength. The addition of fiber did not increase the ultimate strengths, but it did

provide a secondary benefit of keeping material bound together after failure and

increased residual strength after failure.

The form and content of this abstract are approved. I recommend its publication.

Approved: Dr. Frederick Rutz

iv

DEDICATION

I dedicate this work to the family and friends who supported my efforts.

v

ACKNOWLEDGMENTS

I would like to thank my advisor Dr. Frederick Rutz. His interest in analyzing

historic structures encouraged me to do the same and follow my passion.

vi

TABLE OF CONTENTS

CHAPTER

I. EARTH CONSTRUCTION ............................................................................................ 1

Advantages of Earth Homes ................................................................................... 1

Disadvantages of Earth Homes ............................................................................... 2

II. HISTORICAL BACKGROUND ................................................................................... 4

III. EARTH CONSTRUCTION RESEARCH ................................................................... 9

Strength of Compacted Earth ................................................................................ 10

Soil Selection Criteria ........................................................................................... 10

Direct Tensile Testing ........................................................................................... 11

Structural Behavior of Rammed-Earth Walls ....................................................... 12

Assessing the Anisotropy of Rammed Earth ........................................................ 12

Soils Reinforced with Plastic Fiber....................................................................... 13

IV. MATERIALS ............................................................................................................. 15

Soil ........................................................................................................................ 15

Sand and Aggregate .............................................................................................. 15

Cement .................................................................................................................. 16

Reinforcement ....................................................................................................... 17

V. RAMMED EARTH PROCESS ................................................................................... 18

VI. OBJECTIVE OF STUDY........................................................................................... 24

VII. MIX DESIGN............................................................................................................ 25

Soil Selection Requirements ................................................................................. 25

Optimum Moisture Content and Maximum Dry Density ..................................... 32

Cement and Fiber Additives ................................................................................. 36

vii

VIII. EXPERIMENT DESCRIPTION ............................................................................. 38

Unconfined Compression Test Preparation .......................................................... 38

Direct Shear Preparation ....................................................................................... 44

Modulus of Rupture Preparation ........................................................................... 50

IX. TEST RESULTS ........................................................................................................ 53

Unconfined Compression...................................................................................... 53

Modulus of Elasticity ............................................................................................ 61

Direct Shear .......................................................................................................... 66

Modulus of Rupture .............................................................................................. 79

X. CONCLUSION ............................................................................................................ 86

Recommendations for Future Research ................................................................ 87

REFERENCES ................................................................................................................. 89

APPENDIX ....................................................................................................................... 92

New Mexico Building Code Excerpts .................................................................. 92

Fiber Specifications from Manufacturer ............................................................. 104

Unified Soil Classification System ..................................................................... 105

Gypsum Cement Specifications .......................................................................... 106

Portland Cement Specifications .......................................................................... 108

viii

LIST OF TABLES

TABLE

VII.1 Gradation Results from the Soil. ............................................................................ 27

IX.1 Unconfined Compressive Strength for Soil Samples. ............................................. 56

IX.2 Unconfined Compressive Strength for Soil-Cement Samples. ................................ 56

IX.3 Unconfined Compressive Strength for Soil-Cement-Fiber Samples. ...................... 57

IX.4 Modulus of Elasticity for Soil-Cement Samples. .................................................... 64

IX.5 Modulus of Elasticity for Soil-Cement-Fiber Samples. .......................................... 65

IX.6 Adjusted Modulus of Elasticity for Soil-Cement-Fiber Samples. ........................... 66

IX.7 Soil Direct Shear Test Results. ................................................................................ 69

IX.8 Soil-Cement Direct Shear Test Results. .................................................................. 70

IX.9 Soil-Cement-Fiber Direct Shear Test Results. ......................................................... 71

IX.10 Soil Modulus of Rupture Test Results. .................................................................. 81

IX.11 Soil-Cement Modulus of Rupture Test Results. .................................................... 81

IX.12 Soil-Cement-Fiber Modulus of Rupture Test Results. .......................................... 82

ix

LIST OF FIGURES

FIGURE

II.1 Pueblo de Taos. ........................................................................................................... 6

II.2 Hakka Earth Houses. ................................................................................................... 7

II.3 Miller House. ............................................................................................................... 8

V.1 Rammed Earth Forms. .............................................................................................. 19

V.2 Concrete Foundation. ................................................................................................ 21

V.3 Modern adobe wall.................................................................................................... 22

V.4 Rammed Earth Wall. ................................................................................................. 23

VII.1 Sieve Stack. ............................................................................................................ 26

VII.2 Liquid Limit. ......................................................................................................... 28

VII.3 Casagrande Device. ................................................................................................ 29

VII.4 Plastic Limit. .......................................................................................................... 31

VII.5 Standard Proctor Tools. .......................................................................................... 33

VII.6 Air-drying Soil. ...................................................................................................... 34

VII.7 Compaction Curve. ................................................................................................. 35

VII.8 Polypropylene Fibers. ............................................................................................. 37

VIII.1 Color of Soil-Cement and Soil. ............................................................................. 39

VIII.2 Fibers in Soil. ........................................................................................................ 40

VIII.3 Soil-Cement in Proctor Mold. ............................................................................... 41

VIII.4 Extracted Soil-Cement Sample. ............................................................................ 41

VIII.5 Capping Cylinders................................................................................................. 43

VIII.6 Unconfined Compression Sample. ........................................................................ 44

VIII.7 Liner on Proctor Mold........................................................................................... 45

x

VIII.8 Liner inside Mold. ................................................................................................. 45

VIII.9 Liner filled with Soil. ............................................................................................ 46

VIII.10 Pneumatic Jack. ................................................................................................... 47

VIII.11 Soil and Liner. ..................................................................................................... 48

VIII.12 Cut Shear Samples. ............................................................................................. 49

VIII.13 Beam Mold. ......................................................................................................... 51

VIII.14 Modulus of Rupture. ........................................................................................... 52

IX.1 MTS Machine. ......................................................................................................... 54

IX.2 Forney Testing Machine. ......................................................................................... 55

IX.3 Crushed Soil Sample................................................................................................ 59

IX.4 Crushed Soil-Cement Sample. ................................................................................. 59

IX.5 Crushed Soil-Cement-Fiber Sample. ....................................................................... 60

IX.6 Soil-Cement Stress-Strain. ....................................................................................... 62

IX.7 Soil-Cement-Fiber Stress-Strain. ............................................................................. 63

IX.8 Soil Shear Stress. ..................................................................................................... 69

IX.9 Soil-Cement Shear Stress. ....................................................................................... 70

IX.10 Soil-Cement-Fiber Shear Stress. ............................................................................ 71

IX.11 Shear Stress Failure of Soil.................................................................................... 73

IX.12 Shear Stress Failure of Soil-Cement. ..................................................................... 74

IX.13 Shear Stress Failure of Soil-Cement-Fiber. ........................................................... 74

IX.14 Soil Direct Shear Plot. ........................................................................................... 75

IX.15 Soil-Cement Direct Shear Plot............................................................................... 76

IX.16 Soil-Cement-Fiber Direct Shear Plot. .................................................................... 77

IX.17 Soil-Cement Modulus of Rupture. ......................................................................... 83

IX.18 Soil-Cement-Fiber Modulus of Rupture. ............................................................... 84

xi

IX.19 Soil-Cement-Fiber Modulus of Rupture Close-Up................................................ 84

xii

LIST OF EQUATIONS

EQUATION

IX.1 Unconfined Compressive Strength. ......................................................................... 53

IX.2 Modulus of Elasticity. .............................................................................................. 61

IX.2 Shear Strength. ......................................................................................................... 68

IX.4 Modulus of Rupture for 4-point loading. ................................................................. 80

xiii

LIST OF ABBREVIATIONS

in inch

lb pound

MOR modulus of rupture

OMC optimum moisture content

psi pounds per square inch

SC soil-cement

SCF soil-cement with fiber

tsf tons per square foot

UCS unconfined compressive strength

1

CHAPTER I

EARTH CONSTRUCTION

Advantages of Earth Homes

Rammed earth houses are universally available. Earth is a free building material

and no person or company can monopolize the building material. Anyone with a plot of

land and a few helping hands can construct an earthen home. These homes require only

simple building techniques, and materials can be locally sourced. Earth homes offer a

sustainable housing solution for everyone in the world.

These houses are very durable and require little maintenance over the life of the

structure. The cost of maintaining a rammed earth can be very low. Typically, the only

regular maintenance is refinishing the exterior with a fresh coat of plaster. This coating

helps protect the wall from moisture and erosion. The soil in the wall contains no organic

matter and it will not biodegrade overtime. Unlike rammed earth homes, wood homes

will slowly deteriorate as the wood rots from moisture and microorganisms. The earthen

walls do not attract rodents or insects because there is nothing in the wall for these pests

to eat. Termites can cause serious problems for wood homes, which makes earth homes

an excellent choice in termite prone regions.

The massive walls, usually at least 18 inches thick, perform well in adverse

weather conditions. The walls can handle high winds, which can be beneficial in

hurricane and tornado prone areas. The weight of walls is self-anchoring against uplift

forces and the thick walls are much more rigid than typical wood framed homes. Earth

homes offer a natural barrier from cold temperatures and extreme fluctuations in

2

temperature. The thick walls act as an energy collector and store heat from the sun. This

energy is slowly radiated inside and it helps keep the indoor air temperature constant.

Rammed earth homes are very quiet inside. The mass of the walls provides

exceptional insulation from outside noise. As David Miller notes (Miller 1980), “Earth

structures are also very quiet due to mass, added insulation, and double glazing. This

lack of noisy and irksome furnaces, compressors, fans, and ducts helps contribute to

peace of mind.” Many modern earth homes typically utilize radiant heat in the floors;

thus, they do not have noisy forced-air furnaces for heating. Air-conditioning is typically

not needed, even in desert climates, because the thermal mass of the walls regulates the

indoor air temperature.

Disadvantages of Earth Homes

Earth homes are very labor intensive to build. In areas with a high labor cost,

there is a large upfront cost to construct an earthen home. The walls must be built onsite

and there is no mass production of prefabricated building materials when it comes to

building a rammed earth home. In developed countries, rammed earth homes have

difficulty competing with cheap building products, like wood. Wood homes can be

rapidly built, which reduces costs associated with labor.

Modifications are to existing rammed earth homes are challenging. Rammed

earth walls are much more permanent than typical wood framed walls. It is not easy to

cut new openings in walls for windows or doors. This can increase future construction

costs if modifications to the structure are needed.

Rammed earth is an unusual and uncommon building material. It carries a stigma

because people tend to think only the poor live in “mud huts”. The stigma is primarily a

3

result of society being unfamiliar with modern rammed earth homes. Many examples of

high-end rammed earth homes can be found in the American West. Still for many

people, building a home from uncommon materials creates some apprehension in their

mind. People might be uncertain about how the building will perform or who will be

qualified to inspect and service this unique structure. These are valid concerns and

thankfully all these concerns can be addressed through education about rammed earth.

4

CHAPTER II

HISTORICAL BACKGROUND

Building a shelter is a primitive instinct in humans. In the last several thousand

years, ancient cultures across the globe built dwellings of earth and stone. More recently,

the industrial revolution brought assembly-line mechanization that produces cheap

building products of wood, steel, and concrete. As a result of the industrial revolution,

the prevalence of earth buildings has declined in the last century. A new era of earth

building is on the horizon with people looking for sustainable, economical, and durable

housing. This renewed interest brings challenges as building experts must bring this

ancient technology up to modern building codes and standards.

Examples of rammed earth can be found all around the world. Many of the

ancient examples have disappeared as they succumbed to the forces of nature. Some of

the first cities in the world were built from earth, since earth was a readily available

resource. Building materials in the Middle East were scarce and they mainly consisted of

earth and stone, as good timber was not abundantly available. The Tower of Babel in the

seventh century B.C. was constructed of sun baked bricks and parts of the Great Wall of

China were built with rammed earth (Easton 2007). In the first century A.D., the

historian Pliny the Elder documented rammed earth construction in his book, “Natural

History” (Bostock 1885).

And then, besides, have we not in Africa and in Spain walls of earth,

known as "formaceoan" walls from the fact that they are moulded, rather

than built, by enclosing earth within a frame of boards, constructed on

either side. These walls will last for centuries, are proof against rain,

5

wind, and fire, and are superior in solidity to any cement. Even at this day,

Spain still beholds watch-towers that were erected by Hannibal, and

turrets of earth placed on the very summits of her mountains. It is from the

same source, too, that we derive the substantial materials so well adapted

for forming the earth-works of our camps and embankments against the

impetuous violence of rivers. What person, too, is unacquainted with the

fact, that partitions are made of hurdles coated with clay, and that walls

are constructed of unbaked bricks?

From the Middle East, earth construction spread to parts of Africa and Asia. The

Moors brought rammed earth and mud brick technology called adobe to Spain. Earth

building technology then traveled to South America and eventually north to Mexico

during the Spanish conquests. Prior to the arrival of the Spanish conquistadors, Native

Americans, mainly the Pueblo people, used basic earth buildings techniques throughout

the American southwest around A.D. 700. Taos Pueblo community in northern New

Mexico, Figure II.1, was established in the late 13th

and early 14th

centuries and located

there is the oldest continuously occupied structure in North America (Easton 2007). It is

one of the best examples of prehispanic architecture in the Americas. It is made from

adobe mud bricks and some of the walls are several feet thick. The main multistory

structure consists of residential apartments that usually have two rooms, one for sleeping/

living and the other for cooking. The primary access to each apartment was by ladder

through an opening on the roof. The roof entrance was designed for defensive purposes

and the ladders could be removed if the village came under attack. Traditional doors

6

were added later when the Spanish settled in the area and spread their architectural

influence.

Figure II.1 Pueblo de Taos.

The north side of the main structure at Taos Pueblo. Photo by Captain-tucker. Used with

permission.

For the last several centuries different methods of earth construction have been

used throughout the world. A variation of adobe called cob was common in the United

Kingdom. Cob uses softball sized lumps of wet clay stacked together to form thick walls.

Many examples of historic earthen buildings exist in Rhone River valley of France. The

French call rammed earth pisé de terre and they used wooden forms filled with

compacted layers of earth. The Chinese have used rammed earth for thousands of years

and the Hakka people of southern China currently occupy multistory buildings built from

rammed earth, Figure II.2.

7

Figure II.2 Hakka Earth Houses.

A cluster of circular and square rammed earth homes in southern China. Photo by yazi-

swallow. Used with permission.

More recently a significant amount of research into rammed earth was performed

by CRATerre (Center for the Research and Application of Earth) in France during the

1970’s to 1990’s under the leadership of Hugo Houben, Patrice Doat, and Hubert

Guillard (Houben and Guillard 2008). CRATerre is still conducting research and

students can earn a graduate degree in Architecture. Australia is another pioneer of

modern rammed earth construction. The lack of good timber in Australia spawned a

8

building industry focused around masonry, concrete, and earth. Both Australia and New

Zealand have well established and detailed building codes for earthen structures.

Even Colorado had two modern day rammed earth pioneers, Lydia and David

Miller. The Millers built five houses around Greely, Colorado during the 1940’s and

early 1950’s (Miller 1980). The private residence of the Millers is shown in Figure III.3.

California, Arizona, and New Mexico have many examples of earth buildings and these

states have a small, but very capable community of earth building experts. New Mexico

even has a section in their current building code that specifies requirements for earth

structures. Excerpts of the New Mexico building code are in the Appendix.

Figure II.3 Miller House.

Ranch style house built with rammed earth by the Millers. Photo by Mother Earth News.

Used with permission.

9

CHAPTER III

EARTH CONSTRUCTION RESEARCH

The initial work on this thesis involved the review of numerous research papers,

books, and other sources related to rammed earth. No published research was found on

the use of plastic fiber reinforcement in rammed earth. Geotechnical papers discussed the

use of fiber reinforcement, but these papers were not addressing the use of fiber

reinforcement in rammed earth construction. Building a house from vertical walls of soil

is a unique blend of structural and geotechnical engineering, and relatively few research

papers focus on this topic.

Based upon the literature review, several research papers were selected for

discussion in this thesis. These selected research papers guided the work in this thesis to

a subject that had not been investigated. Specifically, the subject of using fiber and

cement together to improve the strength of rammed had not been thoroughly investigated.

Also, there were no research papers on rammed earth that tested compressive strength,

shear strength, and flexural strength in the same study. The New Mexico and Australian

building codes both have strength requirements, so research should focus on ways to

improve the strength of rammed earth in order to meet building codes. Furthermore,

research papers did not examine how fiber reinforcement changes the material properties

of rammed earth. Fiber reinforcement could be an economical means of increasing the

strength of rammed earth.

The content of the research papers that were selected for further discussion is

summarized below. The main purpose of discussing these research papers is to help the

reader understand the current body of knowledge. The papers on rammed earth include

10

studies on compressive strength, developing soil selection criteria, tensile strength

testing, structural behavior, anisotropic properties, and fiber reinforcement.

Strength of Compacted Earth

One study (Burroughs 2006) examined the strength of rammed earth with varying

amounts of stabilizers, like Portland cement. This study is significant due to the large

number of soils tested, which numbered 104 different soils. The author performed 219

strength tests using varying amounts of lime, cement, and asphalt emulsion for

stabilization. Stabilizers were added at an amount of 0-6% per dry soil weight. The

strength test was unconfined compressive strength (UCS). Each soil was tested for liquid

limit, plastic limit, linear shrinkage, and gradation.

Burroughs was able to classify soils as favorable or unfavorable for stabilization.

To be favorable for rammed earth construction, the UCS must be over 2 MPa (~290psi).

Soils that consistently performed well had sand contents >65%, plasticity index (PI) of

<15, and a shrinkage of <6%. It was observed that these favorable soils gain relatively

little compressive strength when stabilized with lime and/or cement. Unfavorable soils

(<2 MPa) exhibited more strength gain when stabilized, but their compressive strength

still did not meet the 2 MPa criteria.

Soil Selection Criteria

Another study (Burroughs 2008) continued the research started in the 2006 study

mentioned previously. In this study by Burroughs, soil samples from other research

studies were compared to the 104 samples collected by the author. Parameters for

determining suitable soils for rammed earth include: clay/silt content, sand content,

11

gravel content, liquid limit, plastic limit, plasticity index, and linear shrinkage. Optimal

parameters are as followed: clay/silt content ranged from 21% to 35%; sand content

ranged from 30% to 70%; gravel ranged from 13% to 62%; liquid limit ≤35; plastic limit

16-19; plasticity index <15; and linear shrinkage <6%. Linear shrinkage is the best

discriminator in determining the suitability of a soil for rammed earth and it was found

that 93 of 100 samples with a LS<6% meet the 290 psi criteria for unconfined

compressive strength. In other words, soils that have low shrinkage generally have good

strength for rammed earth walls. Gravel content and plastic limit are the worst

discriminators in determining the suitability of a soil.

Direct Tensile Testing

This study by Mesbah et al. (Mesbah et al. 2006) focused on how to test a

compressed earth block containing randomly distributed natural fibers for reinforcement.

The natural fibers were sisal fibers from the Agave plant and they ranged in length from

0.8 inches to 2.0 inches long. Natural fibers have been used throughout history as a

means to reinforce earth and they are commonly found in adobe blocks. In this case,

fibers were added at a rate of 0.5% per dry weight of soil.

The authors devised a means to directly pull apart the block and test the tensile

strength. It was noted that the fibers kept cracks from propagating and thus the tensile

strength was increased. The longer fibers, 2 inches, gave the blocks better residual

strength after cracks formed. The fibers did not fail, but they did pull out of the block.

Overall, the fibers improved the tensile strength and more research is needed to determine

the optimal length and quantity of fibers.

12

Structural Behavior of Rammed-Earth Walls

This study by Reddy and Kumar (Reddy and Kumar 2011) was conducted to

examine the ultimate compressive strength of full scale rammed earth walls. This is an

important study because actual full scale walls were tested. Previous studies reviewed for

this thesis tended to focus on the testing of small cylindrical samples. This study by

Reddy and Kumar draws correlations between the strengths of small samples and full

scale walls. Three different sized specimens were tested: a prism measuring 6” long x 6”

thick x 12” tall, a small wall measuring 24” x 6” x 28”, and a wall 30” x 6” x 118”. The

prism had the highest average compressive strength, followed next by the small wall, and

finally the wall had the lowest strength. Much of the variation between the strength of

the three different specimens can be attributed to the slenderness ratio of each specimen.

The walls were the most slender and this leads to buckling as the walls deflect laterally

when loaded vertically.

Some important recommendations were collected from this study. The samples

should have enough time to cure and all the samples should be roughly the same moisture

content before testing. Compaction of the soil on the wet-side of the optimum moisture

content (OMC) curve results in higher compressive strength. Testing samples that are

fully saturated is recommended since this approach is done in testing of masonry.

Testing prisms rather than cylinders will better simulate field conditions and this

approach is done in masonry testing.

Assessing the Anisotropy of Rammed Earth

This study by Bui and Morel (Bui and Morel 2009) evaluates the possible

anisotropic nature of rammed earth. Since monolithic walls are constructed with layers

13

of compacted soil, there is concern about bonding between layers and the variations

within each layer. Layers that are too thick will not be compacted evenly. The upper

portion of the layer will be more compacted than the bottom, which means the density

will be higher in the upper portion.

To evaluate the rammed earth material, samples were loaded perpendicular to the

layers. This simulated how a wall is typically loaded from the dead loads of the building.

The samples were also loaded parallel to the layers. This simulated how the horizontal

forces of wind or earthquakes might load a wall. The results for compressive strength

and failure modulus (the ratio between maximum stress and maximum deformation) were

very similar. The study concluded at low stress, before separation of layers, the rammed

earth acts as a continuous material.

The possible density variation between the top and the bottom of a lift was also

examined. A homogenization process determined an equivalent elastic modulus for a

layer of rammed earth. The homogenized sample was compared to a typical non-

homogenized rammed earth sample. Again, the compressive strengths were similar.

Ultimately this study concluded rammed earth is isotropic if the layers remain

bonded together. The results of this research have a significant impact on the study of

rammed earth. By assuming the material is isotropic, simplified models can be used to

predict the behavior of rammed earth. Anisotropic material requires more detailed and

complicated analysis.

Soils Reinforced with Plastic Fiber

This study by Jiang et al. (Jiang et al. 2010) compared the engineering properties

of soil reinforced with different lengths and percentages of polypropylene fiber. These

14

samples were not rammed earth samples, but the study did provide general information

on how fibers change the strength properties of soil. Fiber lengths were 0.4 inches, 0.6

inches, 0.8 inches, and 1.0 inch. Fiber contents as a percentage of dry soil weight were

0.1%, 0.2%, 0.3%, and 0.4%. The researchers measured the unconfined compressive

strength and the shear strength using the direct shear test. It was concluded that fibers 0.6

inches or less increased both compressive strength and shear strength. Longer fibers

decreased the compressive and shear strength. A fiber content of 0.3% was the optimal

content to increase both compressive and shear strength.

This study also examined the effect of aggregate sizes in the fiber-reinforced soil

mixture. Aggregate sizes were <1mm, 1-2mm, 2-5mm, and 5-10mm. In general,

increasing the aggregate size lowered the unconfined compressive strength. The angle of

internal friction peaked at 3.5mm aggregate. Cohesion decreased as the aggregate size

increased. Overall, the conclusion of this study is polypropylene fiber can increase the

strength and stability of soil.

15

CHAPTER IV

MATERIALS

Soil

One soil was obtained and it was the only soil used in this experiment. The soil

was a Clayey Sand (SC) as defined by the Unified Soil Classification criteria. The liquid

limit was 26 and the plasticity index was 6. The optimum moisture content (OMC) was

11.5% and the maximum dry density was 121 pounds per cubic foot. The optimum

moisture content is the percentage of moisture in the soil that allows for maximum

density after compaction. All these tests were performed by the author of this thesis.

Linear shrinkage was not tested because Atterberg tests indicate this was a low plasticity

soil and these soils have very little shrinkage; thus, the linear shrinkage was likely below

the <6% criteria defined in other research (Burroughs 2006). This soil conformed to the

ideal properties developed in other studies mentioned in this thesis. It did have a very

low clay content, which means stabilization with cement was needed to achieve the

desired strength parameters.

Sand and Aggregate

Sometimes it is necessary to add sand and aggregate to soil mixture. For this

thesis, it was not necessary to add sand and aggregate. The native soil used in this project

already had a high sand content; hence, its classification as a Clayey Sand.

16

There are instances when sand and aggregate can improve the rammed earth soil

mixture. If the native soil was pure clay, it would be beneficial to add both sand and

gravel. Pure clays have the tendency to shrink and swell, and this can lead to cracking in

the wall. Another disadvantage of pure clay soil is cement does not bind well with it. A

general rule of thumb is the clay content of the rammed earth soil should be around 20 to

30 percent.

It is worth noting that some studies (Burroughs 2008, Jiang et al. 2010) have

found large aggregate reduces the strength of rammed earth. The larger material has a

smaller surface area, which will reduce the friction surface and this can lead to lower

strength in compression and shear.

Gravel and other aggregate are not always used in rammed earth projects. One

study (Burroughs 2008) found gravel content >13% can increase the strength of rammed

earth. It should be noted that gravel, like pea gravel, is typically graded to 3/8” diameter

and smaller gravel can be about 1/8” diameter or larger. Material passing through the

No. 4 sieve (~1/8” diameter) is considered sand. Gravel used in rammed earth projects is

typically pea gravel or smaller, and the use of bigger aggregate is not common.

Cement

Type I/II Portland cement was used to stabilize the soil. For this project, cement

was used at a rate of 6% per dry soil weight. The New Mexico building code specifies a

minimum of 6% cement for stabilized rammed earth. Other studies mentioned previously

in this thesis have used ranges of 5% to 10% cement. One author found low cement

contents, ~2%, can actually lower the strength because the cement interferes with the

17

inter-particle bonding of the clay (Minke 2006). That author and other researchers

consistently find cement contents above 2% increase the strength of the soil mixture.

Hydrated lime is another means of stabilization. Lime is well suited to soil with

high clay content. The lime does not interfere as much with inter-particle bonding of the

clay. For this study, lime was not preferred as a stabilizer because the overall clay

content of the final soil mixture was only ~20%. Also, cement seems to be the preferred

method of stabilization amongst the research studies.

Reinforcement

Polypropylene fiber was used in this study. The fibers were short (0.5 inch) in

length and mixed into the dry soil at a quantity of 0.3% per dry weight of soil. These

parameters are based upon the study of fiber-reinforced soil (Jiang et al. 2010). Other

studies (Consoli 1998, 2011) have used longer fibers and incorporated greater quantities

of fibers into the soil. Longer fibers have dispersion problems and tend to not mix well

with the dry soil as observed by the author of this thesis. Greater quantities of fibers can

lower compressive strength and shear strength.

Based upon trail batches of soil with various amounts of polypropylene fiber, it

was determined that 0.3% fiber content was acceptable for this thesis. Higher amounts

tended to ball up and disperse poorly in the soil. Lower amounts resulted pockets of soil

that had very little fiber reinforcement.

18

CHAPTER V

RAMMED EARTH PROCESS

Earth buildings are still being built all over the world. The buildings range from

simple one room homes to high-end luxury homes. Thousands of years of refining

building techniques and more recently the use of applied science have resulted in a

construction method that is a mix of art and science. Rammed earth research is ongoing at

many universities throughout the world and the body of knowledge surrounding rammed

earth is continuously expanding (Maniatidis and Walker 2003).

Rammed earth is an old building technique that has benefited from modern

science. The basic principle is compaction of a soil mixture into rigid forms to create

monolithic walls. The soil mixture is typically ~25% clay, ~60% sand, and ~15% gravel

(Burroughs, 2008). Modern practice is to stabilize the soil mixture with about 5-10%

hydrated lime or Portland cement as a percentage of the dry weight of soil. Concrete

forms or other heavy duty forms are needed to withstand the compaction force and

contain the earth wall until it hardens. Typical monolithic walls are 18 to 24 inches thick

and usually limited to two stories in height.

19

Figure V.1 Rammed Earth Forms.

These plywood and steel framed forms are held together in the middle with steel straps.

Photo by Eric Simenson.

The loose soil mixture is wetted to its optimal moisture content (OMC) and

placed in a loose layer, or “lift”, about 8 inches thick. Pneumatic “pogo-stick” tampers,

similar to those used in small back-fill compaction jobs, are used to compact the soil

(Adobe Builder 2001). Layers that are too thick will potentially have areas that are not

compacted because the tamper is not powerful enough to compact deep lifts of soil. In

practice, the final compacted layer typically is about half the thickness of the loose lift.

The forms are usually left in place for a few days until the cement partially cures.

Depending upon the wall thickness, the wall may take several months or more to fully

cure.

20

Modern rammed earth buildings can incorporate many features to improve the

performance and safety of the structure. The most common features include concrete

foundations and concrete bond beams at the bottom and top of the walls, respectively.

Other design features include the use of steel reinforcement with vertical and horizontal

rebar embedded in the wall. The use of rebar is relatively uncommon in rammed earth

construction and it is use in other types of earth construction techniques has not been

documented. The structural performance of the building can also be improved by

limiting the size and location of doors, windows, and other openings. These basic

principles are very similar to masonry design. Architectural features usually include

large roof overhangs to keep rain from eroding the walls.

To illustrate some of these modern features, Figure V.2 shows a typical rammed

earth foundation. A keyway is constructed in the top of the foundation to transfer shear

forces. Figure V.3 shows both a concrete foundation and a concrete bond beam on an

adobe house. Modern rammed earth and adobe walls have some similar features, like

bond beam and concrete foundations, because building codes require these features.

21

Figure V.2 Concrete Foundation.

Concrete foundation with a keyway serves as the foundation for a rammed earth wall.

Photo by Eric Simenson.

22

Figure V.3 Modern adobe wall.

Concrete foundation and concrete bond beam on an adobe wall. Photo by Eric Simenson.

Special admixtures can be incorporated into the soil and special coatings can be

applied to the finished wall to improve performance. Many modern rammed earth

buildings are coated with plaster and newer products like acrylic sprays are utilized as

wall coatings that provide weatherproofing. Coatings, like acrylic, act as waterproofing,

which helps prevent the wall from eroding due to moisture. Plaster and stucco coatings

23

provide a durable exterior surface that takes the wear and tear from mother nature. It is

routine maintenance to restore coatings at regular intervals.

Overall, the design of rammed earth buildings largely depends upon the soil

properties and the location of the building with regards to climate and environmental

hazards. Some earth buildings are currently hundreds of years old and they were made

without the use of cement and special coatings. Other buildings in moist climates need

additional detailing to prevent water from destroying the walls.

Figure V.4 Rammed Earth Wall.

The plywood forms were removed after one day on this cement stabilized rammed earth

wall in New Mexico. Photo by Eric Simenson.

24

CHAPTER VI

OBJECTIVE OF STUDY

The purpose of this study was to test the engineering properties of rammed earth

with polypropylene plastic fibers and Portland cement additives. Currently, there is no

published research on the use of plastic fiber reinforcement in rammed earth. The

materials included a suitable soil that meets criteria established by other researchers in

peer-reviewed journals (Bryan 1988, Burroughs 2006, 2008). Additives to the soil

included Portland cement Type I/II, polypropylene fiber, and water. The three mix

designs for the experiment were soil, soil-cement, and soil-cement-fiber. For the soil-

cement and soil-cement-fiber mixes, the minimum strength requirements were 300

pounds per square inch compressive strength and 50 pounds per square inch modulus of

rupture (New Mexico Building Code 2009). According to the Australian building code,

the rammed earth material was assumed to have no shear strength (HB 195 Standards

Australia 2001).

Unconfined compression, direct shear, and beam flexure tests were performed on

the three mix designs. Analysis of the test results included comparison of ultimate

strengths and also the behavior of the material during testing.

25

CHAPTER VII

MIX DESIGN

Finding a suitable soil is a crucial step in rammed earth construction because the

mix design consists of over 90% moist soil and the remainder being soil additives, like

cement. Poor soils can be used if larger quantities of cement are added to increase

strength. Cement is a costly additive and one of the advantages of rammed earth

construction is minimizing material cost; thus, minimizing cement is a goal for most

builders.

Soil Selection Requirements

Several research papers (Bryan 1988; Burroughs 2006, 2008) offer guidance on

selecting soils. The soil mixture is typically about 25% clay, 60% sand, and 15% gravel.

The gravel content needs to be fairly small with a limited amount of material being

retained on the No.4 sieve. To determine clay/sand/gravel content, dry soil was passed

through a stack of sieves with progressively smaller openings. Upon inspection, the soil

for this experiment contained some percentage of clay and this clay portion needed to be

removed from the soil before putting the soil through the sieve stack. The reason for this

procedure was the clay sticks on the finest sieve (No. 200) instead of passing through it.

To measure the clay content, the soil was washed through a No. 200 sieve. The clay

content was determined by comparing the original dry weight of the soil to the dry weight

of the soil remaining on the No. 200 sieve. The soil washed through the No. 200 sieve

was considered the clay portion of the soil. Below is Figure VII.1 showing the washed

and dried soil sample ready for gradation through the sieve stack.

26

Figure VII.1 Sieve Stack.

The soil sample is placed on the screen and it is shaken through progressively finer

screens.

For this project, approximately 3% was retained on the No.4 screen and this was

considered to be the gravel fraction of the soil. Approximately 20% passed through the

No. 200 sieve and this was the clay fraction. The remaining 77% was sand of various

grain sizes. For this experiment, the fraction passing the No. 200 sieve was considered

27

clay and not silt due to the properties exhibited during the Atterberg tests. Clay is

cohesive and it displays predictable behaviors during the Atterberg test. On the hand, silt

can be unpredictable and can be very sensitive to moisture. For example, adding one

percent moisture to silt can cause it to go from a stiff soil to almost liquid. It is very

difficult to perform an Atterberg test on silt. For these reasons, it can be justifiably

assumed that clay was in the soil and not silt.

Below are the results from the gradation test and Atterberg test. Based upon these

results, the soil could be classified with the Uniform Soil Classification System.

Table VII.1 Gradation Results from the Soil.

Sieve

Size

[mm]

Sieve Size

Cumulative

Soil Retained

[g]

%

Retained

%

Passing

4.75 4 18.4 3.2 96.8

2 10 71 12.5 87.5

1.19 16 115.8 20.4 79.6

0.425 40 231.4 40.8 59.2

0.25 50 281.5 49.7 50.3

0.15 100 368.7 65.1 34.9

0.075 200 435 76.7 23.3

Pan 451.6 79.7 20.3

28

Figure VII.2 Liquid Limit.

The results of the Atterberg Liquid Limit test.

As shown in Table VII.1 above, the soil for this experiment meets the gradation

criteria explained in previous research (Bryan 1988, Burroughs 2006, 2008) for clay,

sand, and gravel. The approximate ideal quantities are 20% to 30% clay, 70% to 80%

sand, and a small percent of less than 10% as gravel. The soil was obtained from Pioneer

Sand Company at 6379 Valmont Road Boulder, Colorado 80301. Pioneer supplies a

variety of landscaping materials and they called this soil “fill dirt”. According the

supplier, it is essentially soil used for backfilling and it is not meant for growing plants.

Figure VII.2 displays the liquid limit results from the Atterberg test. The test

method used was ASTM D4318 “Standard Test Methods for Liquid Limit, Plastic Limit,

and Plasticity Index of Soils”. The test was performed using a standard device developed

by Professor Casagrande.

29

Figure VII.3 Casagrande Device.

The standard device used to test the liquid limit of soils.

A layer of moist soil was placed smoothly into the brass cup shown in Figure

VII.3. A groove was cut through the middle of the sample using a special grooving tool

as specified in ASTM D4318. The test was performed by turning a crank on the

Casagrande device, which drops the brass cup a standard distance. Each drop is called a

“blow” and the total blow count was recorded when the groove in the soil closes at which

30

time the test was completed. As seen in Figure VII.3, a portion of the groove has closed

and the test is completed. The test was performed three times using soil with different

moisture contents. The liquid limit is defined as the moisture content at which it takes 25

blows to close the groove. The purpose of testing three samples is to have one point

below 25 blow counts, one point near 25, and the last point above 25. The three points

should form a straight line and graphically the 25 blow count point on the line is used to

determine the liquid limit. ASTM does allow a liquid limit to be determined from testing

just one sample. This procedure is necessary when a limited amount of soil is available

for testing. For this thesis, 100 grams of soil was reserved for the Atterberg tests, which

allowed three liquid limit tests and the plastic limit test to be performed.

In general, pure clays tend to have a higher liquid limit than sandy soils and

testing the liquid limit of a soil can give someone a rough idea of the suitability soil for

rammed earth. Pure clays should be avoided due to their tendency to shrink and swell,

which causes cracking. On the other hand, pure sand is not wanted because it has no

cohesion between the sand grains and it cannot be formed into a wall. Suitable soil has a

small fraction of clay and the remainder is sand. The sticky, cohesive property of clay

helps bond together all the sand grains.

Three liquid limit samples were tested at varying moisture contents to establish a

linear trend line. Since the trend line closely follows the points, the sample at 25 blows is

used to determine the liquid limit of 26% for this soil.

The plastic limit was tested in accordance with ASTM D4318. The test involved

rolling thin ribbons of soil on a glass plate. The plastic limit is defined as the moisture

content at which a ribbon of soil can be rolled 1/8” thick. Higher moisture contents allow

31

the ribbons to easily be rolled. The procedure is to slowly reduce the moisture content by

actively rolling the soil on the glass plate. At the plastic limit, the soil will be at the

lowest moisture content needed to form the 1/8” thick ribbon. The plastic limit was

tested and determined to be 15.

Figure VII.4 Plastic Limit.

The 1/8” thick ribbon of soil has reached the plastic limit.

The plasticity index is determined by subtracting the plastic limit from the liquid

limit. The plasticity index was 11. According to the Unified Soil Classification System,

a soil with these gradation and Atterberg properties is classified as a Silty Sand/Clayey

Sand with the symbol SM-SC. As previously described, the soil exhibited some cohesion

during the Atterberg limit tests; thus, the soil is more likely a Clayey Sand (SC). Silt is

non-cohesive and it can be very moisture sensitive when performing these tests. For

32

example, some silt can have a liquid limit that is very close to the plastic limit. The

unusual behavior is common with silts and this behavior was not observed during testing

for this thesis. Overall, this soil passed the general requirements for rammed earth

material.

Optimum Moisture Content and Maximum Dry Density

Upon determination of a suitable soil, the next step was to calculate the optimum

moisture content and maximum dry density of the soil. These values were calculated

using a Standard Proctor test in accordance with ASTM D698 “Standard Test Method for

Laboratory Compaction Characteristics of Soil Using Standard Effort (12400 ft-lbf/ft3)”

(ASTM D698 2012). A split mold was the best choice because the compacted samples

could be easily extracted from the mold. Without a split mold, the compacted sample

was very difficult to remove from the mold. Typically, the sample would be pried out of

the mold in small pieces and a composite sample of small pieces would be used to

determine the moisture content.

The split mold, shown in Figure VII.5, was also needed to prepare rammed earth

samples for unconfined compression and direct shear testing. The unconfined

compression samples were prepared by following the standard proctor procedure and

creating a compacted cylinder of soil. Those tests are described in more detail in Chapter

VIII.

33

Figure VII.5 Standard Proctor Tools.

The split mold seen on the left and the standard 5.5 lb hammer on the right.

The soil was prepared for the standard proctor test by air-drying it for 24 hours

and the moisture content of the air-dried was measured. This was the preferred procedure

because it is relatively easy to spread the soil on a tarp in the laboratory or outside on a

sunny day. Smaller samples can be dried more rapidly by placing them in an oven set at

a low temperature. Given the volume of soil needed for this thesis, air drying was the

most practical method that satisfied the ASTM D698 standard.

34

Figure VII.6 Air-drying Soil.

The soil was dried outside on a sunny day.

Five samples were prepared for the standard Proctor test and the moisture content

was increased by approximately 3% for each sample. Any material larger than the No. 4

sieve was removed in accordance with the test procedure.

35

Figure VII.7 Compaction Curve.

The compaction curve was developed using five samples at varying moisture contents.

Based upon the test results, the optimum moisture content (OMC) was 11.5%

with a maximum dry density of approximately 121 pounds per cubic foot. The OMC is

the point where the soil obtains the highest density when compacted. For rammed earth,

it is very important to prepare soil at the OMC in order to achieve the highest density and

strength.

For this thesis, Portland cement was added to some of the samples being tested.

When water is added to cement, a chemical process occurs and the water is consumed

during the reaction. In effect, some portion of the water in the soil will be lost when

cement is added. For this thesis, it was decided that water will be added only to bring the

soil to its optimum moisture content. One study (Reddy and Kumar 2011) found the

optimum moisture content was slightly higher, 0.4%, for soil-cement when compared to

the original soil. It was decided that increasing the moisture content for the samples

36

containing cement would introduce a new variable and that would make it more difficult

to compare the three soil mixtures.

Cement and Fiber Additives

Three mix designs were used in this experiment. The first mix was simply soil at

the OMC of 11.5%. The second mix was soil at the OMC plus 6% Portland cement. The

6% Portland cement was calculated as a percentage of the dry weight of soil. The third

mix was soil at the OMC plus 6% Portland cement and 0.3% ½” polypropylene fiber.

The 0.3% fiber was calculated as a percentage of the dry weight of soil.

The percentage of Portland cement was specifically chosen because the NM

Building Code and the research cited in this thesis used 6% as the minimum amount of

cement required to stabilize rammed earth. Other researchers (Minke 2006) have shown

lower quantities of cement can reduce strength of the rammed earth due to chemical

interaction with the cement and clay. Higher quantities of cement will most likely

increase the strength of rammed earth, but higher quantities are not always necessary and

are not economical. Typically, a suitable soil for rammed earth will easily meet strength

requirements outlined in the NM Building Code when low quantities, such as 6%, of

Portland cement are used.

The percentage of fiber was based upon the work of other researchers (Jiang et al.

2010). Typically low percentages, such as below 1%, of fiber are used in mix designs.

High amounts of fiber can reduce workability as the fibers tend to impede the mixing of

materials. This phenomenon was observed during trail batching for this thesis. Several

trial batches were created to test the workability of the addition of fiber to the soil. Based

upon previous research and trial batches, it was concluded that 0.3% fiber reinforcement

37

would be sufficient for this thesis. Smaller quantities left pockets in the soil where there

were little or no fibers. Higher quantities tended to bunch up during mixing and this

could have created weak spots in the samples because the fibers do not bond to each

other. The 0.3% fiber generated the highest strength samples for other researchers and

the objective of this thesis was to create higher strength rammed earth by incorporating

synthetic fibers.

Figure VII.8 Polypropylene Fibers.

The ½” long polypropylene fibers used in the fiber-reinforced samples.

38

CHAPTER VIII

EXPERIMENT DESCRIPTION

Three tests were performed on the three mix designs. The first test performed

was unconfined compression on cylindrical samples. The second test performed was

direct shear on cylindrical samples. The third test performed was 3-point modulus of

rupture on rectangular beams. Both the unconfined compression and modulus of rupture

required 6 samples of each mix per test. The direct shear required 9 samples of each mix.

Unconfined Compression Test Preparation

Samples for unconfined compression were prepared in accordance with the

ASTM standard “Compressive Strength of Molded Soil-Cement Cylinder” (ASTM

D1633 2007). The samples were made by compacting soil into a standard Proctor mold

following the same procedure previously mentioned for determining the OMC and

maximum dry density. The benefit of following this procedure was the samples were

uniform size and each sample received the same amount of compactive force.

The procedure used for this experiment was to mix enough soil to make two

samples at a time. Materials were mixed in a 5-gallon bucket and stirred together with a

hand trowel. Initially, larger batches were attempted, but the problem was the soil-

cement would become noticeably stiffer about 20 minutes after adding water to the dry

soil-cement mixture. The stiffer soil-cement made it difficult to compact into the mold.

It was determined that only two samples could be prepared at a time before the soil-

cement created any issues with compaction.

39

Figure VIII.1 Color of Soil-Cement and Soil.

After adding the Portland cement, the mixture turns a gray color.

The method was to mix all the materials dry then add water. It was much less

effort to mix dry materials and materials mixed together better when dry. The fibers have

a tendency to clump together and extra mixing was needed to evenly disperse the fibers.

If clumps of fibers formed, the clumps were pulled apart by hand. Mixing was completed

once the gray color of the cement covered all the soil and the original brown color of the

soil could no longer be seen. Mixing of fibers was completed once random sampling of

the mixture showed even dispersion of fibers throughout the mixture.

40

Figure VIII.2 Fibers in Soil.

Fibers are randomly distributed in the soil and the cement has not been added.

The soil mixes were placed into the mold in three lifts following the same

procedure for the standard Proctor test. Each lift was scarified to promote bonding

between the lifts. Excess material above the mold was scraped off using a beveled edge

and the tops of the cylinders were made smooth as possible. To facilitate sample

extraction from the mold, the sides were lightly lubricated with WD-40 prior to placing

any soil mix in the mold. The lubrication was necessary because the soil-cement mixes

had a tendency to stick to the sides of the mold.

41

Figure VIII.3 Soil-Cement in Proctor Mold.

The top surface has been trimmed and sample is ready for extraction from the mold.

Figure VIII.4 Extracted Soil-Cement Sample.

Sample removed from mold and ready for curing procedure.

42

The molded samples were individually placed in sealed plastic bags and allowed

to cure for 28 days in a humid concrete curing room. This was somewhat of a redundant

procedure. The most important step in the curing procedure was seal the sample in a

plastic bag. This ensured no moisture was gained or lost from the sample. Controlling

the moisture was one variable that was held constant throughout this experiment. Placing

the samples in a humid curing room followed the ASTM standard. The reality was the

samples did not benefit from being in the humid room; however, it was convenient to

store the samples in the curing room, which was located in the laboratory where the

testing was performed. In practice, it is beneficial to moist-cure soil-cement similar to

moist-curing concrete. However, moisture content was controlled throughout the

experiment and kept at 11.5%, so moist-curing was not done on any samples.

One deviation from the ASTM procedure was made in this experiment. The

standard calls for submerging the cylinder in water for 4 hours prior to testing. It was

decided not to do this because the soil and soil-cement cylinder would absorb water at

different rates. There was also some concern that the soil cylinders would fall apart if

they were submerged. The ASTM procedure was specific to soil-cement, which can

handle being submerged for prolonged periods. It was determined that since moisture

was being held constant in the samples, it would have been contradictory to suddenly

submerged the samples immediately before testing.

Prior to testing, gypsum cement (plaster) caps were placed on both ends of the

cylinder. Several initial specimens were tested without caps and the results varied

significantly. The primary reason for this is the surface of the cylinders was inherently

uneven and compressive load was not distributed evenly, which leads to stress

43

concentrations and premature failure of the specimen. Hydrocal® gypsum cement was

used to cap all the cylinders. The end of the cap was made smooth with a piece of glass

and a small level.

Figure VIII.5 Capping Cylinders.

Glass plates were used to create smooth caps made of gypsum cement.

44

Figure VIII.6 Unconfined Compression Sample.

A capped sample of rammed earth ready for unconfined compression testing.

Direct Shear Preparation

The first steps in preparing the shear samples started with the procedure for

making an unconfined compressive sample. The soil mixture was compacted into the

standard Proctor mold following the same procedure previously mentioned. Once the

mold was filled and leveled with the beveled edge, it remained in the mold. A small soil

sample liner of stainless steel measuring 2.5 inches in diameter and 3 inches long was

placed on top of the soil in the mold. The metal liner was then forced down into the soil,

thus filling the liner with the prepared soil sample. This procedure is illustrated in

Figures VIII.7, VIII.8, and VIII.9.

45

Figure VIII.7 Liner on Proctor Mold.

Soil sample liner ready to be driven down into the Proctor mold.

Figure VIII.8 Liner inside Mold.

Soil sample liner successfully driven into the Proctor mold.

46

Figure VIII.9 Liner filled with Soil.

The liner filled with soil after the excess soil is trimmed away.

47

Figure VIII.10 Pneumatic Jack.

The jack used to extract soil sample from soil sample liner.

48

Figure VIII.11 Soil and Liner.

A stainless steel soil sample liner and the extracted sample.

The soil filled liner was placed in a sealed plastic bag and stored in a humid

concrete curing room for 7 days. After 7 days, the soil sample was removed from the

liner using a pneumatic soil extraction jack, shown in Figure VIII.10. The soil sample

measuring approximately 3 inches long was then saw-cut into smaller samples measuring

about 0.9 inches long. Typically, one 3 inch long liner would yield two shear samples.

The soil at the top and bottom of the liner was usually damaged and not intact, so those

areas were cut off. Figure VIII.12 shows examples of the cut shear samples that are

ready for testing.

49

Figure VIII.12 Cut Shear Samples.

Three shear samples have been cut from a 3 inch long soil sample liner.

The reason for using the 2.5 inch diameter stainless steel soil liner was the direct

shear machine has a 2.5 inch diameter shear box. Once the samples were saw-cut, they

neatly fit into the shear box. One difficulty was this experiment required all samples

receive the same compactive force and it would be very challenging to compact soil into

the shear box. The liner solved this problem because it was driven into a standard

Proctor mold. This method ensured all shear samples were compacted the same amount.

Another issue was the soil-cement samples needed time to cure. It would be

inefficient to compact soil-cement directly into the shear box and wait at least a week for

it to cure. There was only one shear box available and it was not practical to wait one

week to test one sample when 27 samples needed to be tested.

50

One advantage of using the liner was multiple samples were cut from one liner.

Another advantage is soil samples in the field are collected by placing these liners into

hollow stem drill rigs. Soil laboratories prepare shear samples by extracting soil out of

the liner and cutting it to the desired length. When possible, this experiment used

industry practices that could easily be replicated by other laboratories.

Shear samples were tested in accordance with ASTM standard “Direct Shear Test

of Soils Under Consolidated Drained Conditions” (ASTM D3080 2011). The shear box

on the direct shear machine forced a failure plane through the midsection of sample.

Cutting the samples to the correct thickness ensured the failure plane occurred through

middle of the sample. A force normal to the failure plane was applied by adding weights

to a moment arm on the machine. This force was varied and three samples were tested at

three different normal forces.

Modulus of Rupture Preparation

Modulus of rupture beam samples were created in custom built rectangular forms,

shown in Figure VIII.13. The testing procedure followed ASTM standard “Flexural

Strength of Soil-Cement Using Simple Beam with Third-Point Loading” (ASTM D1635

2012). The beams measured 3 inches by 3 inches by 11.25 inches. In order to achieve

uniform density, the volume of the mold was calculated and an amount of soil mixture

was weighed to fill the mold. In theory, all of the weighed soil would fit into mold if the

proper amount of compaction was applied. The soil was placed in three lifts and each lift

was scarified to promote bonding between layers. This procedure was similar to the

standard Proctor procedure used to create unconfined compression samples and direct

shear samples.

51

Figure VIII.13 Beam Mold.

Laminated melamine board was used to create a mold for the modulus of rupture test.

Before adding the soil, the sides were lightly greased with WD-40 to prevent the

soil mixture from sticking to the laminated wood. Once the mold was filled, a beveled

edge was used to scrap off excess soil and create a smooth surface on the top of the beam.

The mold filled with soil was wrapped with plastic and the sample was allowed to cure in

the mold for 7 days in a humid concrete curing room. Beam samples were removed from

the mold by unscrewing one side of the mold, which allowed the samples to easily be

removed.

52

Figure VIII.14 Modulus of Rupture.

The loading apparatus required to perform the modulus of rupture test.

The Figure VIII.14 shows a beam placed in the testing apparatus. Steel pipes with

a diameter of 1.25 inches were used instead of steel rods as specified in the ASTM

standard. The pipes were a suitable alternative because the soil mixes were expected to

fail at very low strengths and any deflection in the pipe would be insignificant. Stronger

steel rods would be needed for testing high strength material.

53

CHAPTER IX

TEST RESULTS

Unconfined Compression

All unconfined compression tests were performed in accordance with ASTM

D1633 standard “Compressive Strength of Molded Soil-Cement Cylinder” (ASTM

D1633 2007) on samples that were cured 28 days. The unconfined compressive strength

(σc) is calculated by dividing the ultimate load (Pu) by the area (A) of the cylinder.

σc = Pu/A

Equation IX.1 Unconfined Compressive Strength.

The unconfined compression tests were performed using the MTS Model

#204.63LUBT Compression Testing Machine that has a 20,000 pound capacity. In

accordance with ASTM D1633, the machine was programmed to apply a constant load

rate of 20 ± 10 psi per second. The load and vertical displacement were automatically

recorded with the machine’s computer software. The raw data was entered into Excel,

where plots of the stress versus strain were created for each mix design.

Before testing samples in the MTS machine, several trial samples were created of

each mix. The purpose of the trial samples was to understand how the samples would

perform during the compression tests and the main concern was how the plaster caps

would affect the results. It was observed that the plaster caps sometimes cracked during

the early stages of the load application from approximately zero to 100 psi compressive

stress. The cracking of the caps caused the vertical displacement to rapidly increase

54

during this initial phase of loading. After the initial settlement or cracking of the cap, the

stress-strain curve began to follow a linear path that correlated with the predicted

properties of the mix designs. The results of the trial tests are not included here because

they were simply used to calibrate and hone the procedural steps for actual testing of the

mix designs.

Figure IX.1 MTS Machine.

MTS Machine used for unconfined compression tests.

The samples of soil only were tested using the Forney Model F-40EX-F-TPILOT

Compression Testing Machine following ASTM D1633 procedures. These samples were

the first mix design tested and the MTS machine was unavailable at the time of testing.

Unfortunately this machine does not automatically record data and only ultimate stress

55

was recorded for each sample. The ultimate strengths of the three mix designs were still

compared, but stress-strain plots for the soil only mix design were not created.

Figure IX.2 Forney Testing Machine.

Forney machine used for compression tests and modulus of rupture tests.

Below are summary tables of the unconfined compression tests for each of the

three mix designs. Six samples of each mix design were tested. The ultimate

compressive strength is listed for each sample in pounds per square inch. The average

compressive strength of the samples, the standard deviation, and the coefficient of

variation are also listed.

56

Table IX.1 Unconfined Compressive Strength for Soil Samples.

Soil

Sample

Pressure

[psi]

1 53

2 74

3 64

4 68

5 66

6 67

Average

Strength [psi] 65

Std. Deviation

[psi] 7

COV 0.11

Table IX.2 Unconfined Compressive Strength for Soil-Cement Samples.

Soil-Cement

Sample Pressure [psi]

1 732

2 806

3 766

4 632

5 634

6 651

Average

Strength

[psi]

704

Std.

Deviation

[psi]

75

COV 0.11

57

Table IX.3 Unconfined Compressive Strength for Soil-Cement-Fiber Samples.

Soil-Cement-Fiber

Sample Pressure [psi]

1 766

2 735

3 720

4 716

5 658

6 655

Average

Strength [psi] 708

Std.

Deviation

[psi]

44

COV 0.06

The soil-cement and soil-cement-fiber samples had nearly identical average

compressive strengths. This suggests that the fibers had little or no influence upon the

compressive strength. The fibers were found to hold cracks together after failure. The

soil-cement matrix was the sole provider of strength in compression. The strength

contribution due to stabilization with cement is clearly evident in these test results.

Adding 6% Portland cement increased the compressive strength of the original soil by

approximately 11 times.

Numerous researchers referenced in this thesis have documented the increase in

compressive strength when cement is added to soil. However, studies have not compared

the compressive strengths between soil-cement and soil-cement with fibers. This thesis

found the average strength of soil-cement and soil-cement with fibers were essentially the

same, but the standard deviation and COV was lower for the soil-cement-fiber mixture

indicating the fibers gave the material more predictable average strength. A larger

58

sample size would need to be studied before conclusively stating that fiber additives

produce samples with less strength variability.

The results of this test confirm the work that others have already done. The

primary reasons for performing this test were two-fold: 1) validate cement-stabilization is

possible with the selected soil, and 2) validate the results with other studies. Clearly,

cement-stabilization occurred because the compressive strength greatly increased when

cement was added to the soil. The results were also similar to studies referenced in this

thesis. One additional goal was achieved through performing compression tests and that

was the cement-stabilized samples all exceeded the minimum compressive strength of

600 psi specified in the New Mexico building code. This soil and mix design would be

suitable for construction in the state of New Mexico.

The failure mechanisms for the three mixes were markedly different. The soil

tended to slowly deform under light loading in a ductile, plastic type behavior. Both the

soil-cement and soil-cement-fiber displayed sudden, brittle failure at much higher

compressive stress levels than the soil only mixture.

59

Figure IX.3 Crushed Soil Sample.

Failed soil sample after completion of the unconfined compression test.

Figure IX.4 Crushed Soil-Cement Sample.

Failed soil-cement sample after completion of the unconfined compression test.

60

Figure IX.5 Crushed Soil-Cement-Fiber Sample.

Failed soil-cement-fiber sample after completion of the unconfined compression test.

The soil-cement cylinders would suddenly fail with vertical cracking along the

outer edges, which indicates a splitting type failure. Once failure occurred approximately

50% of the cylinder would crumble and fall off the platen. The soil-cement-fiber

cylinders also failed suddenly with vertical cracking; however, nearly all of the material

would remain connected through the fibers and the cylinder would remain on the platen.

The fibers were not able to increase the load capacity of the cylinder; however, they did

significantly change the behavior during failure. The plaster caps influenced the failure

type by restraining the ends of the cylinder. These restraints forced the stress path

straight-down through the cylinder. This is a known phenomenon, so the caps were

poured as thin as possible to minimize the restraining effect.

61

Modulus of Elasticity

From data collected during the compression testing, it was possible to create

stress-strain curves for the soil-cement and soil-cement-fiber mix designs. Using ASTM

standard “Static Modulus of Elasticity and Poisson’s Ratio of Concrete in Compression”

(ASTM C469 2010) the modulus of elasticity was calculated. The Poisson ratio was not

tested because it requires a special apparatus to measure lateral deformation of the

cylinder. The modulus of elasticity (E) was calculated using the chord modulus of

elasticity formula where S2 is the stress at 40% of the ultimate load, S1 is the stress at ɛ1,

ɛ2 is the strain at S2, and ɛ1 corresponds with a longitudinal strain of 0.000050.

E = (S2 - S1)/(ɛ2 - ɛ1)

Equation IX.2 Modulus of Elasticity.

The following stress-strain plots were generated from the compression testing of

the soil-cement and soil-cement-fiber mix designs.

62

Figure IX.6 Soil-Cement Stress-Strain.

The stress versus strain plots for the six soil-cement samples.

Sample #2 and Sample #6 exhibit the characteristic initial curve shape that

indicates cracking and settlement of the plaster cap. The first portion of the curve rises at

a small angle until approximately 50 psi then the curve rises steeply. The linear elastic

portion of the curve occurs after this initial portion. To avoid recording data from the

cracking of the cap, Sample #1, #3, #4, and #5 were preloaded. The preload was applied

by the machine and the load was adjusted using the machine’s computer controls. The

preload varied from less than 40 psi to 120 psi depending upon condition of the plaster

cap and how the platens on the machine contacted the plaster caps.

63

Figure IX.7 Soil-Cement-Fiber Stress-Strain.

The stress versus strain plots for the six soil-cement-fiber samples.

The preload for the soil-cement-fiber samples was much smaller. This was due to

better pouring of the caps. The plaster had a tendency to harden very quickly and within

as little as 10 minutes the material was sometimes unworkable. Practice with mixing and

pouring the plaster led to the creation of thinner, smoother, and more level plaster caps.

These higher quality caps required less preloading because they had less of a tendency to

crack and settle. Despite all this attention to detail, Sample #4 showed larger initial strain

with little loading and this indicates a potentially uneven cap. Sample #1 also showed

some initial issues with the cap and the overall curve did not take the same shape as the

64

other sample. Sample #1 reached a high strength, but the stress-strain curve indicates

some type of anomaly with this sample.

The modulus of elasticity was calculated using Equation IX.2. According to the

ASTM standard, the initial stress is taken at a strain of 0.00005. For this thesis, this

procedure was not practical because the plaster caps created some anomalies during the

first portion of the stress-strain curves. After reviewing the data and plots, it was decided

that the initial stress/strain point would be taken where the slope of curve began to rise

steeply. At this point, it was assumed that soil-cement was behaving linearly elastically.

If the initial stress/strain point was taken prior to any cracking of the plaster cap, the

result was a modulus of elasticity that was unrealistically low.

Table IX.4 Modulus of Elasticity for Soil-Cement Samples.

Soil-Cement

Sample

Modulus of

Elasticity [psi]

1 118,409

2 111,571

3 143,740

4 142,231

5 162,460

6 171,074

Average

Strength

[psi]

141,581

Std.

Deviation

[psi]

23,445

COV 0.17

65

Table IX.5 Modulus of Elasticity for Soil-Cement-Fiber Samples.

Soil-Cement-Fiber

Sample

Modulus of

Elasticity [psi]

1 29,965

2 111,394

3 176,875

4 188,992

5 112,752

6 126,053

Average

Strength

[psi]

124,339

Std.

Deviation

[psi]

56,835

COV 0.46

It is should be noted that Sample #1 of the soil-cement-fiber mix exhibited an

unusual stress-strain curve. The cylinder was able to deform much more than the other

samples and it was the strongest sample of the soil-cement-fiber mix. One possible

explanation for this anomaly was a small portion of the cylinder may have not been fully

compacted. This could happen if one of the three lifts in the cylinder was placed too

thick, which would have caused the bottom portion of the lift to be less compacted. As

the sample was tested, the testing machine compressed the weak layer until it was fully

consolidated and this caused extra deformation in the cylinder. Regardless of the

explanation, this sample was suspect and another table without Sample #1 was created.

Removal of outlying data was done in accordance with ASTM “Standard Practice for

Dealing with Outlying Observations” (ASTM E178 2008). The ASTM standard uses

statistical analysis to determine if a data point can be classified as an outlier. In this case

66

Sample #1 was determined to be an outlier and removed from the Modulus of Elasticity

calculations.

Table IX.6 Adjusted Modulus of Elasticity for Soil-Cement-Fiber Samples.

Soil-Cement-Fiber

Sample

Modulus of

Elasticity [psi]

1 -

2 111,394

3 176,875

4 188,992

5 112,752

6 126,053

Average

Strength

[psi]

143,213

Std.

Deviation

[psi]

36,958

COV 0.26

With Sample #1 removed, the adjusted modulus of elasticity for soil-cement-fiber

mix design in Table IX.6 was very similar to the soil-cement mix design. The average

modulus of elasticity for soil-cement-fiber was E = 143,213 psi and the soil-cement was

E = 141,581 psi. Given the results of the compression tests and the calculated modulus

of elasticity, it appeared that adding fibers to the soil-cement had little or no impact on

these material properties. Again, the biggest changes in material properties came from

the addition of cement.

Direct Shear

Direct shear tests were performed in accordance with ASTM D3080 standard

“Direct Shear Test of Soils Under Consolidated Drained Conditions” (ASTM D3080

67

2011) on samples that were cured for 7 days. The decision was made to test samples a 7

days rather than 28 days due to the large number of samples needed for testing.

Furthermore, it was very difficult to prepare shear samples and it was not feasible to wait

28 days in between batches given time constraints on the project. Many of the samples

were damaged during saw cutting or during extraction from the soil sample liner.

Typically the middle of the soil sample liner would yield one good, undamaged sample

that could be used for testing. Originally, it was thought that one soil sample liner would

yield three samples, but this proved impossible due to damage during extraction from the

liner and fracturing during saw cutting.

Trial samples were tested to gauge how the machine would handle soil-cement,

which is much stronger than typical soil tested in this machine. It was observed that high

normal force was needed to keep soil-cement samples stationary during testing. The

samples had a tendency to shift in the shear box because the shear force was large, so a

large normal force was needed to hold the sample in place. Typically for soils, a larger

normal force will cause failure at higher shear force. In order to predict the shear

strength of a material, samples were tested at different normal forces and plotted in Excel.

Typically for soils, a linear relationship occurs between shear force and normal force.

Shear stress (τ) in pounds per square inch is calculated by dividing the shear force

(Fs) in pounds by the cross-sectional area (A) of the sample measured in square inches.

Three samples were tested at the same normal stress to obtain an average shear stress for

that specific normal stress. The normal stress was applied at three different stress

increments. The purpose for varying the normal stress was to create a plot of average

68

shear stress versus normal stress. In this plot, a linear trend line was drawn and from the

trend line, the cohesion and friction angle were determined.

τ = Fs /A

Equation IX.3 Shear Strength.

All shear samples were tested in accordance with ASTM standard D3080 using a

Forney Direct Shear Machine Model #2050 SN 814. The machine has an adjustable

displacement rate, which was set at 0.02 inches per minute. The ASTM Standard has a

range of acceptable displacement rates from 0.0001 to 0.04 inches per minute. For soils,

the displacement rate is chosen based upon the time it takes the soil to consolidate. For

this thesis, the soil-cement was already consolidated and it did not consolidate even under

the large normal stress applied to the samples during shear testing. The displacement rate

calculation outlined in the ASTM standard was not applicable in this case; however, the

displacement rate within the standard’s range was selected. At this rate of displacement,

it took approximately 15 minutes to complete testing on one sample. It total, 27 samples

were tested.

Below are the summary tables and charts for direct shear testing of the three

mixes.

69

Table IX.7 Soil Direct Shear Test Results.

Soil

Normal

Stress

[psi] 3.5

Average

Shear Stress

[psi] 7.65

Normal

Stress

[psi] 6.9

Average

Shear Stress

[psi] 13.25

Normal

Stress

[psi] 13.9

Average

Shear Stress

[psi] 19

Figure IX.8 Soil Shear Stress.

Plot of shear stress and normal stress.

70

Table IX.8 Soil-Cement Direct Shear Test Results.

Soil-Cement

Normal

Stress [psi] 111.1

Average

Shear

Stress

[psi] 156

Normal

Stress [psi] 138.9

Average

Shear

Stress [psi] 194

Normal

Stress [psi] 166.7

Average

Shear

Stress [psi] 212

Figure IX.9 Soil-Cement Shear Stress.

Plot of shear stress and normal stress.

71

Table IX.9 Soil-Cement-Fiber Direct Shear Test Results.

Soil-Cement-Fiber

Normal

Stress

[psi] 111.1

Average

Shear

Stress

[psi] 164

Normal

Stress

[psi] 138.9

Average

Shear

Stress

[psi] 190

Normal

Stress

[psi] 166.7

Average

Shear

Stress

[psi] 245

Figure IX.10 Soil-Cement-Fiber Shear Stress.

Plot of shear stress and normal stress.

The soil had a friction angle of 46.4° and cohesion of 4.8 psi. This friction angle

corresponded with the typical upper limit for dense sand (Das 2008). The friction angle

was calculated by taking the inverse tangent of the slope of the line and the cohesion was

72

calculated as the y-intercept of the line. Based upon the soil classification and the

compaction applied to each sample, these results corresponded to what would be

expected for this type of soil.

The soil-cement had a friction angle of 45.2° and cohesion of 47.3 psi. The

friction angle remained close to the soil, but the cohesion increased nearly 10 times. The

increase in cohesion came from the cement. The cement created much greater bonding

between soil particles and this caused the cohesion to substantially increase.

The soil-cement-fiber and soil-cement had similar average shear strengths at

normal stresses of 111.1 psi and 138.9 psi. At a normal stress of 166.7 psi, the soil-

cement-fiber had a significantly higher average shear stress when compared to the soil-

cement, 245 psi and 212 psi shear strength respectively. This higher shear stress caused

the trend line on the soil-cement-fiber plot to rise at a much steeper angle of 55.5°. The

cohesion was calculated at the point the trend line intercepts the y-axis, and in this case,

the value was -2.8 psi. It is theoretically impossible to have a negative value for

cohesion, so it is reasonable to assume this value should be zero based upon the plot.

The anomaly for the soil-cement-fiber samples was likely the higher shear

strength at the maximum normal stress tested. This value of 245 psi seemed unusually

high and lower value that was closer to the soil-cement value of 212 psi would be

expected given the similarity in all the test results thus far. At lower normal stress, the

soil-cement and soil-cement-fiber samples produced similar shear strengths.

Furthermore, the unconfined compression tests and modulus of elasticity calculations

showed soil-cement and soil-cement-fiber samples have very similar properties. If the

average shear stress of 245 psi for the soil-cement-fiber was closer to 212 psi, the trend

73

line for the soil-cement-fiber would be very similar to the trend line for the soil-cement.

The new trend line would create a lower friction angle and higher cohesion, similar to the

soil-cement.

If shear tests were performed again, some changes could be made that might

improve the results. Letting the samples cure for 28 days would give the samples

additional time to gain strength. The extraction of the samples from the liner and also

saw-cutting the samples may have introduced micro-cracks that influenced the results.

Stronger samples might have been able to resist potential cracking caused by these

procedures. Additionally, cutting only one sample from the middle of the liner might

yield higher quality samples. The ends of the extracted samples always showed signs of

damage and these portions were cut away, but it is possible that all the damaged material

was not removed.

Figure IX.11 Shear Stress Failure of Soil.

Failed soil sample from direct shear testing.

74

Figure IX.12 Shear Stress Failure of Soil-Cement.

Failed soil-cement sample from direct shear testing.

Figure IX.13 Shear Stress Failure of Soil-Cement-Fiber.

The two halves are still connected together with fiber after failure.

75

The most significant observation between the soils tested in shear was the soil-

cement-fiber samples remained connected together after failure. The two halves of the

soil-cement-fiber samples had the typical failure plane through the middle of the sample

and the fibers kept the halves connected together.

Figure IX.14 Soil Direct Shear Plot.

Horizontal load and horizontal displacement plot from direct shear testing.

76

Figure IX.15 Soil-Cement Direct Shear Plot.

Horizontal load and horizontal displacement plot from direct shear testing.

77

Figure IX.16 Soil-Cement-Fiber Direct Shear Plot.

Horizontal load and horizontal displacement plot from direct shear testing.

The direct shear machine had a maximum displacement of 0.3 inches. Two of the

soil-cement tests ended before 0.3 inches because the shear box was not positioned to

allow for maximum displacement. Luckily, the samples still failed and some residual

strength was recorded, so the data from these samples was considered valid. Almost all

of the curves have an initial portion where the curve is horizontal. This is caused by the

shear box moving until it was firmly against the sample. Typically a slight gap existed

between the sample and the shear box, so the data showed some horizontal displacement

until this gap was closed. This does not alter the maximum shear stress results.

Overall, the soil-cement and soil-cement-fiber exhibited residual strength after

failure. The residual strength comes from the roughness between the two halves of the

sample sliding over each other. Three of the soil-cement samples had a rapid loss of

78

strength after failure, Sample 8tsf-01, Sample 10tsf-01, and Sample 12tsf-01. The soil-

cement-fiber did not have these sharp declines. It was expected that the fibers would add

more residual strength because the fibers would keep the two halves connected. It is

possible that the fibers helped prevent a sudden loss in strength.

The observation that every sample with fiber remained connected together after

shear failure indicated the fibers were well bonded in the soil-cement matrix. The fibers

have a high tensile strength of 83-96 kips per square inch as specified by the

manufacturer (see Appendix for material data sheet) and they have a large capacity to

elongate before failure. As the two halves sheared apart, the fibers were put into tension

as they tried to hold the halves together. The direct shear machine was unable to create

enough displacement to fail the fibers.

After reviewing the results, it was assumed that the residual strength of the

samples would be even more dramatic if the direct shear machine allowed for greater

displacement. Alternatively, direct tensile testing might be required to fully test the

strength and limits of the fibers in the soil-cement. In future studies, it would be

interesting to continue the test until the fibers failed. At that point of failure, it would be

possible to analyze how much strength the fibers actually contributed. Overall, the

ultimate shear stress of the samples came from the strength of the soil-cement matrix.

The fibers appeared to influence the strength only after the ultimate shear stress of the

soil/cement was reached.

79

Modulus of Rupture

The modulus of rupture (MOR) tests were performed in accordance with ASTM

D1635 standard “Flexural Strength of Soil-Cement Using Simple Beam with Third-Point

Loading” (ASTM D1635 2012). This test measures the flexural strength of beams using

a four point loading apparatus. The samples were tested using the Forney Model F-

40EX-F-TPILOT Compression Testing Machine. This machine had a large lower platen

that could accommodate the size of the beams. This test is specified in the New Mexico

building code and cement-stabilized rammed earth must have a minimum MOR of 50 psi.

It was one objective to determine if the cement-stabilized samples in this thesis met the

minimum strength requirement. Performing research on a material that met building

code requirements would be more relevant to the rammed earth industry than testing a

material that was unsuitable for building.

It was hypothesized that the fibers would increase the MOR by delaying the onset

of cracking and then holding the cracks together. Using natural fibers to delay the onset

of cracking was documented during direct tensile testing research performed by other

researchers (Mesbah et al. 2006). In the MOR test, the bottom of the beam was put into

tension and the top of the beam was in compression. Soil-cement was assumed to have

relatively low tensile strength, but the fibers have relatively high tensile strength. It was

assumed that the fibers would carry the tensile stresses; thus, the soil-cement-fiber beams

would have a higher MOR than the soil-cement beams.

The modulus of rupture (R) is calculated by multiplying the load (P) and span

length of the beam (L) and dividing those numbers by beam width (b) multiplied by the

beam depth squared (d). It should be noted that the span length, width, and depth are

80

measured in inches. The load is measured in pounds force. The span length is not the

overall length of the beam rather it is the distance between the lower supports on the

beam. In accordance with the ASTM standard, the span length was 9.25 inches for all the

specimens.

R = (PL)/(bd2)

Equation IX.4 Modulus of Rupture for 4-point loading.

It should be noted that Equation IX.4 was used to calculate the MOR for the

loading setup shown in Figure VIII.14, which was the loading setup specified by the

ASTM standard. The ASTM standard calls the test a three point loading test, but the

setup they specify is actually a four point loading test. The supports on the top of the

beam were spaced a distance of L/3 per the ASTM standard. Placing the top supports at

this distance modifies the 3-point MOR equation to the 4-point MOR equation.

Trial beams were not created for this test because experience had been gained

using the Forney machine during the compression tests. Also, the ASTM standard

provided an average flexural strength of 94 psi for specimens with 6% cement. This

information was helpful because it provided some estimation of the expected beam

strength. The machine was adjusted to provide a constant load rate of approximately 5

pounds per second. The ASTM standard specifies a constant load rate of 100 ± 5 psi/min

in the extreme fibers of the beam, which equates to approximately 5 pounds per second

applied by the machine.

All of the beams cracked within the middle third of the beam. Adjustments to the

MOR calculation must be made if the crack is outside of the middle third. Below is a

summary of the modulus of rupture tests for the three mixes. The maximum load was

81

recorded by the machine and Equation IX.4 was used to calculate the MOR for each

sample.

Table IX.10 Soil Modulus of Rupture Test Results.

Soil

Sample

Pressure

[psi]

1 6

2 8

3 4

4 4

5 6

6 4

Average

Strength

[psi]

5

Std.

Deviation 2

COV 31

Table IX.11 Soil-Cement Modulus of Rupture Test Results.

Soil Cement

Sample Pressure [psi]

1 163

2 179

3 100

4 171

5 158

6 150

Average

Strength

[psi]

154

Std.

Deviation 28

COV 18

82

Table IX.12 Soil-Cement-Fiber Modulus of Rupture Test Results.

Soil Cement Fiber

Sample Pressure [psi]

1 142

2 138

3 154

4 142

5 158

6 171

Average

Strength

[psi]

151

Std.

Deviation 13

COV 8

Both the soil-cement and soil-cement-fiber samples had very similar results with

an average modulus of rupture of 154 psi and 151 psi, respectively. The soil-cement-

fiber had a lower standard deviation and coefficient of variation. The soil-cement-fiber

beams remained bonded together even after failure, but the beams could not handle any

additional loading. The fibers would stretch if additional load was applied.

The soil had a modulus of rupture of only 5 psi and it could barely handle any

load. The soil almost no flexural strength as was anticipated before testing. One reason

for using reinforced concrete lintel over doors and windows is unstabilized rammed earth

homes have virtually no flexural strength and these lintels are needed to withstand

bending that occurs over the openings. The results of these tests demonstrated the benefit

of adding Portland cement to increase the flexural strength.

83

Figure IX.17 Soil-Cement Modulus of Rupture.

Failed soil-cement beam from modulus of rupture test.

84

Figure IX.18 Soil-Cement-Fiber Modulus of Rupture.

Failed soil-cement-fiber beam from modulus of rupture test.

Figure IX.19 Soil-Cement-Fiber Modulus of Rupture Close-Up.

Fibers hold the failed beam together.

85

Again the most significant difference between the soil-cement and soil-cement-

fiber was the failure mechanism. The soil-cement would suddenly crack and fall off the

testing apparatus. The soil-cement-fiber would suddenly crack, but the fibers would hold

the beam together and the beam would remain on the testing apparatus. The fibers held

the beam together, but the beam could not take any additional load.

The fibers did not delay the onset of cracking, which was unexpected based upon

the research mentioned in this thesis. As soon as a visible crack appeared, it would

propagate to the top of the beam and the beam would fail. If the fibers had delayed the

onset of cracking, the soil-cement-fiber beams would have handle more load, but this did

not happen.

After reviewing the results it seemed the fibers were not stiff enough to allow the

beam to take more load. If the fibers were stiffer, it is likely the crack would have not

propagated as quickly and the crack would have remained tight. Once the beam cracked,

the fibers would begin to stretch, which allowed the crack to widen and propagate.

Fiberglass or steel fibers might have increased the MOR because they are significantly

stiffer than the polypropylene fibers used in this thesis.

86

CHAPTER X

CONCLUSION

The initial hypothesis was the addition of fibers to the soil-cement would greatly

increase the shear strength and flexural strength of the material. For the soil-cement and

soil-cement-fiber samples, the results for unconfined compression, shear, and flexural

testing were similar. The soil-cement-fiber samples had statistically less variation for the

unconfined compression and modulus of rupture tests. The sample size of six is

relatively small to make statistical conclusions, but the evidence in this thesis suggests

soil-cement-fiber material has less variation in strength compared to soil-cement.

The main difference between the soil-cement and soil-cement-fiber was the

failure mechanisms. Brittle failure happened on both the soil-cement and soil-cement-

fiber. However, the material in the soil-cement-fiber samples remained bonded together

after failure. While the fiber held the material together, it did not significantly increase

the load capacity. The cement additive provided the additional strength and it also made

the material brittle.

One possible benefit of the fibers could be in the form of safety. During failure,

the material in a rammed earth wall with fiber reinforcement would remain connected

together rather than falling off the wall. A soil-cement wall without fiber could

potentially be hazardous if large chunks of the wall were falling down on the occupants

during an earthquake or collapse. Adding fibers seems to be a relatively easy way from a

construction standpoint to gain some additional safety.

The results were somewhat surprising in that the fiber did not increase the

strengths more than the results showed. One possible explanation is the plastic fibers are

87

much more elastic than the soil-cement. The difference in engineering properties causes

the soil-cement matrix to resist the entire load until failure. After failure the fibers are

stretched and begin to resist the applied load, but most material strength is gone once the

soil-cement matrix cracks.

This thesis explored a new topic in rammed earth research and tested the strength

of rammed earth with fiber and cement additives. No previous study had been published

on the use of synthetic fiber reinforcement in rammed earth. From the test results, the

soil-cement and soil-cement-fiber mix design had nearly identical compressive, shear,

and flexural strength. The soil-cement-fiber mixture did have less statistical variation in

the test results. The soil-cement-fiber mixture also exhibited different behavior during

failure compared to the soil-cement, and it appears that the soil-cement-fiber mixture

could improve the safety of rammed earth buildings during failure or collapse.

Recommendations for Future Research

The use of fibers in rammed earth is a topic that must be researched more

thoroughly. Currently, there is very little published research on this topic and relevant

research about fiber-reinforced soils is specific to geotechnical applications and not

rammed earth walls. The published geotechnical papers focus on a wide variety of soils

and these soils are not selected specifically for their use in rammed earth. Future research

should follow the recommended soil selection criteria mentioned in this thesis.

There are endless possibilities for conducting more tests on rammed earth.

Testing more samples in compression, shear, and flexure would improve the statistical

analysis of the material by providing a larger data set to analyze. Other test methods, like

triaxial testing, are strongly recommended. This test would likely produce the most

88

accurate results for compressive and shear strength. Researchers have not used this test

on rammed earth and one reason for this might be due to the difficulty and time involved

in performing this test. Triaxial testing might overcome some of the challenges

mentioned in this thesis that occurred with preparing the shear samples. Triaxial tests

could be performed on samples using the same preparation procedures for compression

test samples outlined in this thesis.

Experimenting with different mix designs is highly recommended. The use of

different fiber material is a promising area of research. Fiberglass and steel fiber could

be used instead of polypropylene fiber. These materials have been successfully used to

increase soil and concrete strengths, but their use in rammed earth has not been explored.

Studies could also examine the possibility of reducing the cement content when

increasing amounts of fiber reinforcement are used.

89

REFERENCES

Adobe Builder (2001). “Rammed Earth, Book #9.” New Mexico.

ASTM. (2007). “Standard Test Methods for Compressive Strength of Molded Soil-

Cement Cylinders.” ASTM D1633-07, West Conshohocken, PA.

ASTM. (2007). “Standard Practice for Making and Curing Soil-Cement Compression and

Flexure Test Specimens in the Laboratory.” ASTM D1632-07, West

Conshohocken, PA.

ASTM. (2008). “Standard Practice for Dealing with Outlying Observations.” ASTM

E178-08, West Conshohocken, PA.

ASTM. (2010). “Standard Test Method for Liquid Limit, Plastic Limit, and Plasticity

Index of Soils.” ASTM D4318-10, West Conshohocken, PA.

ASTM. (2010). “Standard Test Method for Static Modulus of Elasticity and Poisson’s

Ratio of Concrete in Compression.” ASTM C469-10, West Conshohocken, PA.

ASTM. (2011). “Standard Test Method for Direct Shear Test of Soils Under

Consolidated Drained Conditions.” ASTM D3080-11, West Conshohocken, PA.

ASTM. (2012). “Standard Test Method for Flexural Strength of Soil-Cement Using

Simple Beam with Third-Point Loading.” ASTM D1635-11, West Conshohocken,

PA.

ASTM. (2012). “Standard Test Method for Laboratory Compaction Characteristics of

Soil Using Standard Effort (12400 ft-lbf/ft3 (600 kN-m/m

3)).” ASTM D698-12,

West Conshohocken, PA.

Bryan, A. J. (1988). “Criteria for the suitability of soil for cement stabilization.” Building

and Environment, 23(4), 309-319.

Bryan, A. J. (1988). “Soil/cement as a walling material—I. Stress/strain properties.”

Building and Environment, 23(4), 321-330.

Bryan, A. J. (1988). “Soil/cement as a walling material—II. Some measures of

durability.” Building and Environment, 23(4), 331-336.

90

Bui, Q., and Morel, J. (2009). “Assesing the anisotropy of rammed earth.” Construction

and Building Materials, 23, 3005-3011.

Burroughs, S. (2006). “Strength of compacted earth: linking soil properties to

stabilizers.” Building Research and Information, 34(1), 55-65.

Burroughs, S. (2008). “Soil property criteria for rammed earth stabilization.” Journal of

Materials in Civil Engineering, 20(3), 264-273.

Consoli, N. C., Prietto, D. M., and Ulbrich, L. A. (1998). “Influence of fiber and cement

addition on behavior of sandy soil.” Journal of Geotechnical and

Geoenvironmental Engineering, 124(12), 1211-1214.

Consoli, N. C., Zortea, F., Souza, M. D., and Festugato, L. (2011). “Studies on the dosage

of fiber-reinforced cemented soils.” Journal of Materials in Civil Engineering, 23,

1624-1632.

Das, B. (2008). Advanced Soil Mechanics, Taylor and Francis, New York, NY.

Easton, D. (2007). The Rammed Earth House, Chelsea Green, White River Junction, VT.

Houben, H., and Guillard, H. (2008). Earth Construction—A comprehensive guide,

Intermediate Technology Publications, London.

Jaquin, P. A., Augarde, C. E., Gallipoli, D., and Toll, D. G. (2009). “The strength of

unstabilised rammed earth materials.” Géotechnique, 59(5), 487-490.

Jiang, H., Cai, Y., and Liu, J. (2010). “Engineering properties of soils reinforces by short

discrete polypropylene fiber.” Journal of Materials in Civil Engineering, 22(12),

1315-1322.

Kaniraj, S. R., and Havanagi, V. G. (2001). “Behavior of cement-stabilized fiber-

reinforced fly ash-soil mixtures.” Journal of Geotechnical and Geoenvironmental

Engineering, 127(7), 574-584.

King, B. (1996). Buildings of earth and straw: Structural design for rammed earth and

straw bale architecture, Ecological Design Press, Sausalito, CA.

91

Maniatidis, V., and Walker, P. (2008). “Structural capacity of rammed earth in

compression.” Journal of Materials in Civil Engineering, 20(3), 230-238.

Mesbah, A., Morel, J. C., Walker, P., and Ghavami, K. (2004). “Development of a direct

tensile test for compacted earth blocks reinforced with natural fibers.” Journal of

Materials in Civil Engineering, 16(1), 95-98.

Miller, B., and Miller, L. (1980). Manual For Building A Rammed Earth Wall, Self-

Published by Authors, Greeley, CO.

Minke, G. (2000). Earth Construction Handbook, WIT Press, Southampton, UK.

Minke, G. (2006). Building with earth: design and technology of a sustainable

architecture. Birkhäuser, Boston, MA.

NM Building Code (2009). “2009 New Mexico Earthen Building Materials Code.”

Housing and Construction Building Codes General, 14.7.4.

Reddy, B. V. V., and Kumar, P. P. (2011). “Structural behavior of story-high cement-

stabilized rammed-earth walls under compression.” Journal of Materials in Civil

Engineering. 23(3), 240-247.

Standards Australia. (2002). The Australian earth building handbook, Sydney, Australia.

Walker, P. J., and Dobson, S. (2001). “Pullout tests on deformed and plain rebars in

cement stabilized rammed earth.” Journal of Materials in Civil Engineering, 13,

291-297

92

APPENDIX

New Mexico Building Code Excerpts

93

TITLE 14 HOUSING AND CONSTRUCTION

CHAPTER 7 BUILDING CODES GENERAL

PART 4 2009 NEW MEXICO EARTHEN BUILDING MATERIALS

CODE

14.7.4.1 ISSUING AGENCY: Construction Industries Division of the

Regulation and Licensing Department.

[14.7.4.1 NMAC - Rp, 14.7.4.1 NMAC, 1-28-11]

14.7.4.2 SCOPE: This rule applies to all earthen building materials

contracting work performed in New Mexico on or after January 28, 2011, that is subject

to the jurisdiction of CID, unless performed pursuant to a permit for which an application

was received by CID before that date.

[14.7.4.2 NMAC - Rp, 14.7.4.2 NMAC, 1-28-11]

14.7.4.3 STATUTORY AUTHORITY: NMSA 1978 Section 60-13-9 and

60-13-44.

[14.7.4.3 NMAC - Rp, 14.7.4.3 NMAC, 1-28-11]

14.7.4.4 DURATION: Permanent.

[14.7.4.4 NMAC - Rp, 14.7.4.4 NMAC, 1-28-11]

14.7.4.5 EFFECTIVE DATE: January 28, 2011, unless a later date is

cited at the end of a section.

[14.7.4.5 NMAC - Rp, 14.7.4.1 NMAC, 1-28-11]

94

14.7.4.6 OBJECTIVE: The purpose of this rule is to establish minimum

standards for earthen building materials construction in New Mexico.

[14.7.4.6 NMAC - Rp, 14.7.4.1 NMAC, 1-28-11]

14.7.4.7 DEFINITIONS:

A. Amended soil means improving an unqualified soil to a qualified state

with the addition of other soils or amendments.

B. Amendments means additive elements to soil, such as lime, portland

cement, fly ash, etc. which are “dry-mixed” into the main soil body as a percentage of

total weight to achieve stabilization.

C. Buttress means a projecting structure providing lateral support to a wall.

The buttress shall be incorporated into the foundation and wall system. (Refer to figure 1

of the earthen building figures supplement).

D. CEB means compressed earth block.

E. Keyway means a groove on the vertical rammed earth wall surface for

interlocking purposes. Refer to figure 3 of the earthen building figures supplement).

F. Lift means a course of rammed earth, placed within the forms, and then

compacted.

G. Nailer means any material rammed into the wall that serves as an

attachment device. Refer to figure 4 of the earthen building figures supplement).

H. Optimum moisture means sufficient water (generally no more than ten

(10) percent) mixed into the soil to attain sufficient compaction.

I. psi means pounds per square inch.

J. Qualified soil means any soil, or mixture of soils, that attains 300 psi

compression strength and attains 50 psi. modulus of rupture.

K. Rammed earth means qualified soil that is mechanically or manually

consolidated to full compaction.

L. Round-cap nails means fasteners that include nails or screws in

combination with caps of at least three-fourths (3/4) inches diameter or three-fourths ( ¾)

inch square.

95

M. Stabilization, stabilized means qualified soils that pass the wet strength

test under ASTM D1633-00 or contain a minimum of six (6) percent portland cement by

weight. Stabilization is achieved through the use of amendments.

N. Wet strength compression test means an approved testing laboratory

process in which a fully cured rammed earth cylinder is completely submerged in water a

minimum of four hours according to ASTM D1633-00, then subjected to a compression

test.

[14.7.4.7 NMAC - Rp, 14.7.4.7 NMAC, 1-28-11]

14.7.4.8 EARTHEN BUILDING MATERIALS:

A. General. The provisions of this rule, 14.7.4 NMAC, shall control the

design and construction of one- and two-family dwellings in which earthen building

materials form the bearing wall system.

B. Allowable wall heights for earthen structures. All earthen structures

whether adobe, burned adobe, compressed earth block, rammed earth or terrón, shall

conform to table 1. For purposes of using table 1, height is defined as the distance from

the top of the slab or top of stem wall to the underside of the bond beam.

Table 1 ALLOWABLE WALL HEIGHTS FOR EARTHEN STRUCTURES

Maximum

Sds

Wall

Thickness

Maximum

Height

Maximum

Sds

Wall

Thickness

Maximum

Height

.25

10 120”

.3

10 120”

12 128 12 128

14 144 14 144

16 144 16 144

18 144 18 144

24 144 24 144

.35

10 120”

.4

10 120”

12 128 12 128

14 144 14 144

16 144 16 144

18 144 18 144

24 144 24 144

10 104” 10 96”

12 128 12 112

14 144 14 136

96

.45

16 144

.5

16 144

18 144 18 144

24 144 24 144

This table is based on two story maximum, one and two family residential with seismic

soil site class D1.

[14.7.4.8 NMAC - Rp, 14.7.4.8 NMAC, 1-28-11]

14.7.4.12 RAMMED EARTH CONSTRUCTION:

A. General. The following provisions shall apply.

(1) Rammed earth shall not be used in any building more than (2) stories

in height. The height of every wall of rammed earth without lateral support is specified

in 14.7.4.8 NMAC table 1. The height of the wall is defined as the distance from the top

of the slab or top of stem wall to the underside of the bond beam.

(2) Exterior rammed earth walls shall be a minimum of eighteen (18)

inches in thickness. Exception: Exterior walls that are also designed as solar mass walls

(trombe) as defined by the passive solar heating worksheet, dated June 2004 and prepared

by the state of New Mexico energy, minerals and natural resources department, are

allowed and shall be minimum thickness of ten (10) inches, not to exceed twelve (12)

inches. They shall be fully attached to or integrated with any adjacent structural wall and

topped with a bond beam that fully attaches them to the bond beam of any adjacent

structural wall as described in 14.7.4.17 NMAC.

(3) Interior rammed earth walls shall be a minimum of twelve (12) inches

in thickness.

(4) Unstabilized rammed earth walls must be covered to prevent

infiltration of moisture from the top of the wall at the end of each workday and prior to

wet weather conditions, whether the walls are contained within forms or not.

(5) Fully stabilized rammed earth walls may be left unprotected from the

elements.

(6) In no case shall a rammed earth wall be reduced in thickness with back

to back channels or nailers. Channels or nailers rammed on both sides of a running wall

shall not be opposite each other to avoid an hourglass configuration in the wall section.

Channels or nailers on both sides of a running wall shall be separated from each other

vertically at a distance no less than the rammed earth wall thickness. (Refer to figure 4 of

the earthen building figures supplement).

97

(7) An architect or engineer registered in the state of New Mexico shall

design and seal structural portions of two-story residential rammed earth construction

documents.

(8) The general construction of the building shall comply with all

provisions of the 2009 New Mexico Residential Building Code (NMRBC), unless

otherwise provided for in this rule.

(9) Passive solar structures incorporating the use of solar mass walls

(trombe), direct gain arrays or sunspaces (greenhouses) as defined by the passive solar

heating worksheet, dated June 2004 and prepared by the state of New Mexico energy,

minerals and natural resources department, are allowed.

B. Fireplaces. Adobe or masonry fireplaces and chimneys in rammed earth

structures shall comply with 14.7.3.18 NMAC. They shall be integrated into adjacent

rammed earth walls during construction or secured to them by suitable steel ladder

reinforcement or reinforcing rods.

C. Count Rumford fireplaces. Count Rumford fireplaces are allowed as

provided in 14.7.3.18 NMAC.

D. Stop work. The building inspector shall have the authority to issue a

“stop work” order if the provisions of this section are not complied with.

E. Lateral support. Lateral support shall occur at intervals not to exceed

twenty-four (24) feet. Rammed earth walls eighteen (18) inches to less than twenty-four

(24) inches thick shall be laterally supported with any one or combination of the

following: A rammed earth wall of bond beam height that intersects the running wall

with at least sixty (60) degrees of support (refer to a figure 5 of the earthen building

figures supplement); an adobe wall of bond beam height and at least ten (10) inches in

width that intersects with and attaches to the running wall with at least sixty (60) degrees

of support (refer to figure 5 of the earthen building figures supplement); a minimum

twenty 20 gauge steel frame or wood frame wall of full height that intersects with and

attaches to the running wall with ninety (90) degrees of support, that is properly cross-

braced or sheathed (refer to figure 6 of the earthen building figures supplement); a

buttress configuration that intersects the running wall at ninety (90) degrees, of adobe or

rammed earth. The buttress base must project a minimum of three (3) feet (or thirty-three

(33) percent of the wall height) from the running wall and support at least seventy-five

(75) percent of the total wall height (refer to figure 7 of the earthen building figures

supplement). The thickness of a rammed earth buttress shall be at least eighteen (18)

inches. The thickness of an adobe buttress shall be a minimum fourteen (14) inches.

Rammed earth walls greater than twenty-four (24) inches in thickness are self-buttressing

and do not require lateral support provided their design adheres to 14.7.4.8 NMAC table

1 and the other applicable provisions of this rule.

98

F. Openings. Door and window openings shall be designed such that the

opening shall not be any closer to an outside corner of the structure as follows.

(1) In rammed earth walls eighteen (18) inches to less than twenty-four

(24) inches thick, openings shall not be located within three (3) feet of any corner of the

structure. (Refer to figure 8 of the earthen building figures supplement). Exception:

Openings may be located within three (3) feet of any corner provided a buttress extending

at least three (3) feet from the structure supports the corner. A continuous footing below

and a continuous bond beam above, shall be provided across such openings.

(2) Rammed earth walls greater than twenty-four (24) inches thick are

self-buttressing, with no special consideration for placement of openings within the area

of the wall.

G. Piers. Rammed earth piers supporting openings shall measure no less

than three (3) square feet in area and no dimension shall be less than eighteen (18) inches.

(Refer to figures 9-A and 9-B of the earthen building figures supplement).

[14.7.4.12 NMAC - Rp, 14.7.4.12 NMAC, 1-28-11]

14.7.4.13 FOUNDATIONS:

A. General. Foundation construction shall comply with applicable

provisions of the 2009 New Mexico Residential Building Code, and the following: a

minimum of three (3) continuous #4 reinforcing rods are required in minimum 2500 psi.

concrete footings supporting rammed earth walls. Stem walls shall be the full width of

the wall supported above or wider to receive forming systems. Footings shall be a

minimum of ten (10) inches in depth.

B. Perimeter insulation. For the purposes of placement of perimeter

insulation, rammed earth walls may overhang the bearing surface up to the thickness of

the perimeter insulation, but in no case greater than two (2) inches.

C. Keyway. A key way shall be provided where the rammed earth wall

meets the foundation system. The keyway shall be established at the top of the stem a

minimum of two (2) inches deep by six (6) inches wide formed at the time of the pour,

and shall run continuously around the structure to include any intersecting rammed earth

wall sections. The rammed earth wall shall be fully rammed into this keyway (refer to

figure 2 of the earthen building figures supplement). Exception: Placement of vertical

reinforcing rods extending a minimum twelve (12) inches into the rammed earth wall.

The vertical rods shall be minimum #4, imbedded into the concrete and spaced forty-

eighty (48) inches on center, maximum.

D. Concrete grade beam. Rubble filled foundation trench designs with a

reinforced concrete grade beam above are allowed to support rammed earth wall

99

construction. An architect or engineer registered in the state of New Mexico shall certify

the grade beam/rubble-filled trench design portion.

[14.7.4.13 NMAC - Rp, 14.7.4.13 NMAC, 1-28-11]

14.7.4.14 RAMMED EARTH SOIL SPECIFICATIONS:

A. General. The soil shall not contain rock more than one-and-a-half (1 1/2)

inch in diameter. The soil shall not contain clay lumps more than one-half (1/2) inch in

diameter. The soil shall be free of all organic matter. The soil shall not contain more

than two (2) percent soluble salts.

B. Soil compressive strength. Prior to the start of construction, fully-cured

rammed earth soil samples shall be tested at an approved testing laboratory for

compressive strength. The ultimate compressive strength of all rammed earth soil,

stabilized or non-stabilized, shall be a minimum three-hundred (300) psi. The

compressive strength report shall be submitted with the permit application. This report

may be waived if the builder provides certification of compliance. The certification must

be dated within one year of the date on the application for the building permit. Samples

tested shall be representative of soil to be used on the project for which the permit

application is submitted.

C. Stabilized rammed earth soil. The following shall apply to stabilization

of rammed earth soil: Asphalt emulsion may not be used for stabilization of rammed

earth soil. Thorough mixing of additives to the soil may be achieved by any method that

assures a complete blending to a uniform color and texture. Stabilized soil is suitable soil

that contains six (6) percent or more portland cement by weight or that passes ASTM

D1633-00. Samples tested shall be representative of soil to be used on the project for

which the permit application is submitted. The compressive strength report shall be

submitted with the permit application. Laboratory testing shall indicate rammed earth

samples attained a minimum of two-hundred (200) psi. after seven (7) days. If a different

soil is provided at any time during construction, it must meet the minimum requirements

outlined above, prior to use in the structure.

D. Unstabilized rammed earth soil. Unstabilized rammed earth soil is that

containing less than six (6) percent portland cement by weight or that fails to pass ASTM

D1633-00. The exterior of such walls shall be protected with approved stucco systems or

other method approved by the building official. Refer to 14.7.4.19 NMAC for weather-

resistive barrier requirements.

E. Amended soil. The following guidelines shall apply when amending soils

to attain a qualified soil. Soil shall not contain rock greater than one-and-a-half (1 1/2)

inch in diameter. Soil shall not contain clay lumps greater than one-half (1/2) inch

diameter. Soil shall be free of organic matter. Soil shall not contain more than two (2)

100

percent soluble salts. Soils to be mixed shall be sufficiently dry to blend completely to

one uniform color and texture. The amended soil shall be tested prior to use as per

Subsection B of 14.7.4.14 NMAC.

F. Forming systems. The forming system shall be adequate to contain the

material under compaction. It shall be properly plumbed and braced to withstand the soil

pressures as well as construction activity on and around it.

G. Placement of material, compaction and curing.

(1) No amount of portland cement stabilized soil will be mixed that will

not be placed in the wall system within sixty (60) minutes of its preparation.

(2) Lifts of prepared soil shall be placed in the forms in relatively even

layers not to exceed 8 inches in depth. Each lift shall then be rammed to full compaction.

(3) Optimum moisture content as determined to meet minimum

compressive strength shall be maintained for stabilized and unstabilized walls.

(4) Work will progress, lift-by-lift, until the work approaches bond beam

height.

(5) Forms may be stripped immediately after ramming is completed for a

section of wall, providing ramming of adjacent sections does not affect the structural

integrity of completed walls.

(6) Portland cement stabilized walls not in forms shall be lightly spray-

cured with water at least five (5) spaced times during daylight hours. This procedure

shall continue for at least three (3) days starting from the time that the wall is exposed to

the elements. Exception: Rammed earth walls left in forms three (3) or more days shall

not require water-spray curing.

H. Placement of attachment materials.

(1) Nailers: Nailers incorporated into the rammed earth wall shall be

installed as follows (Refer to figure 4 of the earthen building figures supplement); the

rammed earth wall shall not be reduced in thickness with back-to-back nailers. To avoid

an hourglass configuration in the wall section, nailers on either side of a running wall

shall not be opposite each other. Nailers on either side of a running wall shall be

separated from each other vertically a distance not less than the rammed earth wall

thickness. Nailers shall be placed onto the wall such that the narrow dimension of the

nailer is exposed on the race of the wall prior to ramming. Nailers shall be cured and

sealed against moisture penetration prior to installation in forms. The nailers shall not

extend the full depth of the wall. Box wood nailers are not allowed. (Refer to figure 11

of the earthen building figures supplement). The nailer shall be no more than two (2)

inches by four (4) inches by its length.

101

(2) Channels: Channels may be incorporated into the rammed earth wall

as follows (Refer to figure 2 of the earthen building figures supplement); To avoid an

hourglass configuration in the wall section, channels on either side of a running wall shall

not be opposite each other. (Refer to figure 4 of the earthen building figures supplement).

Channels shall be no more than two (2) inches by four (4) inches by their length in

dimension. Vertical channels shall not be placed closer than twelve (12) inches to a

rammed earth wall finished edge or corner.

[14.7.4.14 NMAC - Rp, 14.7.4.14 NMAC, 1-28-11]

14.7.4.17 BOND BEAMS:

A. General. The bond beam shall be secured to the rammed earth wall.

Refer to Subsections H and I of 14.7.4.16 NMAC above. Bond beams may be of wood

or concrete construction. Bond beams shall measure six (6) inches nominal depth and

extend the full width of the wall. Exception: The bond beam width may be reduced as

follows: Two (2) inches maximum in an eighteen (18) to less than twenty-four (24) inch

thick rammed earth wall, or three (3) inches maximum in a rammed earth wall twenty-

four (24) inches or greater in thickness. Bond beams must be continuous, running the full

perimeter of the structure. Interior rammed earth or adobe walls shall be incorporated

into the bond beam. Varying height bond beams shall extend into the adjoining rammed

earth wall one-half (1/2) the thickness of the adjoining rammed earth wall. The concrete

bond beam may secure anchoring and strapping devices.

B. Wood bond beam construction. In addition to the general requirements

of Subsection A of 14.7.4.17 NMAC, wood bond beams may be constructed as approved

by the building official. Light wood bond beam construction may be utilized as shown in

figure 10 of the earthen building figures supplement.

C. Concrete bond beam construction. In addition to the general

requirements of Subsection A of 14.7.4.17 NMAC, concrete bond beams shall be

constructed of minimum twenty-five hundred (2500) psi. concrete and shall contain steel

reinforcement as follows: For eighteen (18) to less than twenty-four (24) inch thick

rammed earth wall construction, a minimum of two (2) continuous number four (4)

reinforcing rods shall be used. For walls equal to or greater than twenty-four (24) inches

in thickness, a minimum of two (2) continuous number five (5) reinforcing rods shall be

used. Provide two (2) inch minimum reinforcement concrete cover over all horizontal

reinforcing rods. Concrete bond beams may be used to secure anchoring and strapping

devices.

D. Concrete bond beam cold joints. Concrete bond beam cold joints are

limited to corners of perpendicular intersections with other structural, full-height walls.

Cold joints shall be tied into the adjoining bond beam with three (3) number four (4)

reinforcing rods. The reinforcement shall extend a minimum of twenty-four (24) inches

into both portions of the concrete bond beam.

102

[14.7.4.17 NMAC - Rp, 14.7.4.17 NMAC, 1-28-11]

14.7.4.18 LINTELS OVER OPENINGS:

A. General. All openings require a lintel or semi-circular arch over the

opening. All lintels, whether of wood or concrete shall bear a minimum of twelve (12)

inches into the length of the wall. Exception: Nichos and other shaped voids as defined

in 14.7.4.15 NMAC.

B. Bearing limitations. Lintels shall bear a minimum of twelve (12) inches

beyond coved, splayed or rounded bearing portions of openings that are less than the full

width of the wall. (Refer to figure 15 of the earthen building figures supplement).

C. Lintels over openings in stabilized rammed earth walls. Openings less

than twenty-four (24) inches in width shall not require a lintel or semi-circular arched

opening. Openings greater than twenty-four (24) inches in width require lintels as

defined in table 4.

Table 4 Concrete Lintels Over Openings in Rammed Earth Walls (1)

Wall

width

Lintel span Lintel depth Reinforcement

(2)

Reinforcement

Concrete Cover

(3)

Uniform

Load

18”

24” 6” 3- #4 @ 4”o.c.

3” minimum

concrete cover

on all sides

1000

PLF

36” 6” 3- #4 @ 4”o.c.

48” 6” 3- #4 @ 4”o.c.

60” 6” 3- #4 @ 4”o.c.

72” 8” 3- #5 @ 4”o.c.

84” 8” 3- #5 @ 4”o.c.

96” 8” 3- #5 @ 4”o.c.

20”

24” 6” 3- #4 @ 4”o.c.

4” minimum

concrete cover

on all sides

1350

PLF

36” 6” 3- #4 @ 4”o.c.

48” 6” 3- #4 @ 4”o.c.

60” 6” 3- #4 @ 4”o.c.

72” 8” 3- #5 @ 4”o.c.

84” 8” 3- #5 @ 4”o.c.

96” 10” 3- #5 @ 4”o.c.

24” 6” 3- #4 @ 5”o.c.

36” 6” 3- #4 @ 5”o.c.

48” 6” 3- #4 @ 5”o.c.

60” 6” 3- #4 @ 5”o.c.

103

22”

72” 8” 3- #5 @ 5”o.c. 3 1/2” minimum

concrete cover

on all sides

1700

PLF

84” 10” 3- #5 @ 5”o.c.

96” 10” 3- #5 @ 5”o.c.

24”

24” 6” 3- #4 @ 6”o.c.

3” minimum

concrete cover

on all sides

2000

PLF

36” 6” 3- #4 @ 6”o.c.

48” 6” 3- #4 @ 6”o.c.

60” 6” 3- #4 @ 6”o.c.

72” 8” 3- #5 @ 6”o.c.

84” 10” 3- #5 @ 6”o.c.

96” 12” 3- #5 @ 6”o.c.

1. 3000 psi minimum concrete at approximately 28 days.

2. Grade 40 steel reinforcement minimum.

3. Steel reinforcement at mid-depth of lintel.

[14.7.4.18 NMAC - Rp, 14.7.4.18 NMAC, 1-28-11]

104

Fiber Specifications from Manufacturer

105

Unified Soil Classification System

106

Gypsum Cement Specifications

107

108

Portland Cement Specifications