UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Plastic shrinkage in dry mix shotcrete Campbell, Kevin Neil 1999

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1999-0337.pdf [ 13.59MB ]
Metadata
JSON: 831-1.0050144.json
JSON-LD: 831-1.0050144-ld.json
RDF/XML (Pretty): 831-1.0050144-rdf.xml
RDF/JSON: 831-1.0050144-rdf.json
Turtle: 831-1.0050144-turtle.txt
N-Triples: 831-1.0050144-rdf-ntriples.txt
Original Record: 831-1.0050144-source.json
Full Text
831-1.0050144-fulltext.txt
Citation
831-1.0050144.ris

Full Text

PLASTIC SHRINKAGE IN DRY MIX S H O T C R E T E  by  KEVIN NEIL C A M P B E L L  B.A.Sc. The University of British Columbia, 1995 A THESIS SUBMITTED IN PARTIAL FULFILLMENT O F T H E REQUIREMENTS F O R T H E D E G R E E O F MASTER O F APPLIED S C I E N C E  in  T H E FACULTY O F G R A D U A T E STUDIES DEPARTMENT O F CIVIL ENGINEERING We accept this thesis as conforming to the required standard  T H E UNIVERSITY O F BRITISH COLUMBIA  May 1999 © Kevin Neil Campbell, 1999  ln  presenting  degree  at the  this thesis  in partial fulfilment  of the  requirements  for an advanced  University of British Columbia, I agree that the Library shall make it  freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes department  or  by  his  or  her  representatives.  may be granted by the head of my It  is  understood  that  copying  or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department of The University of British Columbia Vancouver, Canada Da,e  DE-6 (2/88)  '<ff  Abstract  Plastic shrinkage cracking of dry-mix shotcrete is a common problem encountered within the construction industry. Dry-mix shotcrete is used in a wide variety of applications most commonly, the repair and rehabilitation of deteriorated concrete structures. Dry-mix shotcrete is also used as a tunnel liner, for slope stabilization, and for refractory applications. Plastic shrinkage cracks pose a significant threat to the durability and aesthetic appeal of shotcrete structures. This work was intended to develop a simple test used to study restrained plastic shrinkage in shotcrete overlays.  Plastic shrinkage cracking of bonded dry-mix shotcrete overlays subjected to a severe drying environment is investigated. A novel test method was developed to study the effects of fibers and various admixtures on restrained plastic shrinkage. Shotcrete is cast directly onto fully matured, roughened sub-base that provides dimensional restraint. The assembly is immediately transferred to a drying chamber where plastic shrinkage cracking is monitored under controlled conditions. It is found that this method is effective in estimating the potential and degree of plastic shrinkage cracking in shotcrete, and also in assessing the effectiveness of various fiber types and admixtures. Steel and synthetic fibers were found to delay the formation of plastic shrinkage cracks as well as substantially reduce both the amount and width of cracking.  TABLE OF CONTENTS  Abstract  ii  Table of Contents  iii  List of Tables  vii  List of Figures  viii  Chapter 1 - Introduction 1.1 - Plastic Shrinkage in Shotcrete  1  1.2 - Previous Research of Restrained Plastic Shrinkage  1  1.3 - Objectives  3  Chapter 2 - Literature Survey 2.1 - Introduction to Shotcrete  4  2.2 - Shotcrete Fundamentals: Wet-Mix vs. Dry-Mix  4  2.3 - Dry Mix Shotcrete 2.3.1 - Pumping Equipment  5  2.3.2 - Dry-Mix Shotcrete Material  5  2.3.3 - Dry-Mix Shotcreting Production Methods  8  2.3.3.1 - The Bone Dry Method  8  2.3.3.2 - The Pre-Moisturizing Method  9  2.3.3.3- The Wet Sand Method  10  2.3.3.4- The Brazilian Method  10  2.3.4 - Uses of Dry-Mix Shotcrete  11  2.3.4.1 - Repair and Rehabilitation  12  2.3.4.2 - Substrate  12  2.4 - Shrinkage  13  2.4.1 - Material Related Shrinkage  13  2.4.2 - Placement Related Shrinkage  14  iii  2.4.3 - Consequences of Shrinkage  15  2.4.4 - Control of Shrinkage in Shotcrete  15  2.4.4.1 - Fiber Reinforcement  15  2.4.4.2 - Shrinkage Controlling Admixtures  16  2.4.5 - Shrinkage Cracking  16  2.4.5.1 - Crack Formation Due to Desiccation  17  2.4.5.2 - Plastic Shrinkage  17  2.4.5.3 - Cracking Due to Drying  20  2.4.5.4 - Cracking Due to Self-Desiccation  21  2.5 - Review of Restrained Shrinkage Testing  21  2.5.1 - Linear Specimen Testing  22  2.5.1.1 - Beam Specimens  22  2.5.1.2 - Dog-Boned Specimens  23  2.5.1.3 - Internal Restraint  24  2.5.2 - Plate Specimen Testing  24  2.5.2.1 - Continuous Edge Restraint  24  2.5.2.2 - Discrete Edge Restraint  25  2.5.3 - Ring Specimen Testing  26  2.5.4 - Restrained Plastic Shrinkage Tests  28  2.6 - The Effect of Fiber Reinforcement on Restrained and Free Shrinkage  30  Chapter 3 - Development of Test Procedure 3.1 - Initial Stages  33  3.1.1 - Roughened Base Development  33  3.1.2 - Test Chamber Development  34  3.1.3 - Formwork Development  35  3.1.4 - Sample Development  37  3 . 1 . 5 - Mix Design Development  38  3.1.5.1 - Cement Content  39  3.1.6 - Effect of Finishing  40  3.1.6.1 - As-Shot Finish  41  3.1.6.2 - Steel Trowel Finish  41  3.1.6.3 - Excessive Steel Trowel Finish  41  3.1.6.4 - Comparison of Finishing Techniques  42  iv  3.2 - Development of the Final Test Procedure  43  3.2.1 - Equipment  45  3.3 - Comments on the Test Procedure  45  3.3.1 - Panel Set-up  45  3.3.2 - Panel Finishing  46  3.4 - Standard Test Procedures  46  3.4.1 - Compressive Strength  46  3.4.2 - Boiled Absorption and Permeable Voids  47  3.5 - Non Standard Test Procedures  47  3.5.1 - Determination of In-situ Water Content  47  3.5.2 - Determination of In-situ Cement and Aggregate Content  47  3.5.3 - Measurement of Crack Widths and Lengths  48  3.6 - Calibration of Test Procedure  49  3.6.1 - Variables  49  3.6.1.1 - Water Injection System  50  3.6.1.2 - Material Water Content  50  3.6.1.3 - Placement Technique  50  3.6.2-Results  :  50  3.6.2.1 - Water Injection System  54  3.6.2.2 - Material Water Content  55  3.6.3 - Discussion of Results  56  3.6.3.1 - Placement Technique  57  3.6.3.2 - Paste Content  61  Chapter 4 - Evaluation of Variables Affecting Plastic Shrinkage Cracking 4.1 - Summary of Variables  63  4.1.1 - Presentation of Data  64  4.2 - Fiber Analysis  67  4.2.1 - Hooked End Steel Fiber  68  4.2.2 - Flattened End Steel Fiber  69  4.2.3 - Crimped Steel Fiber with Crescent Section  70  4.2.4 - Straight 25 mm Polyolefm Fiber  72  4.2.5 - Straight 50 mm Polyolefin Fiber  74  4.2.6 - Comparison of Fibers  75  v  4.2.6.1 - Fiber Geometry  78  4.2.6.2 - Fiber Rebound  81  4.2.6.3 - Optimum Fiber Design for Plastic Shrinkage Control  82  4.3 - Shrinkage Controlling Admixture  83  4.4 - Wet Mix Study  85  4.4.1 - Waterproofing Admixture  86  4.4.2 - Results of Waterproofing Admixture Study  86  4.5 - General Discussion of Test Program  87  Chapter 5 - Conclusion 5.1 - Conclusion  89  5.2 - Suggestions and Recommendations for Future Research on the Subject  89  References  91  Appendix A - Summary of Test Data  93  Appendix B - Photo Log  100  Appendix C - Restrained Plastic Shrinkage Test Procedure  114  vi  LIST OF TABLES  Table 2.1 - Gradation Limits for Combined Aggregates (ACI 506R-8)  7  Table 3.1 - Roughened Test Base Concrete  34  Table 3.2 - Initial Test Mix  39  Table 3.3 - Standard Test Mix  40  Table 3.4 - Effect of Finishing on Plastic Shrinkage Cracking  41  Table 3.5-Variables Considered in Calibration Study  51  Table 3.6 - Summary of In-situ Cement and Aggregate Contents in Calibration Study  52  Table 3.7 - Summary of Plastic Shrinkage Cracking Results of Calibration Study  54  Table 3.8 - Comparison of Over Finishing with Cast Samples  61  Table 4.1 - Summary of Variables in the Final Test Program  64  Table 4.2 - In-situ Material Content Analysis of Final Test Program  66  Table 4.3 - Summary of Plastic Shrinkage Crack Analysis for the Hooked End Fiber  68  Table 4.4 - Summary of Plastic Shrinkage Crack Analysis for the Flattened End Fiber  69  Table 4.5 - Summary of Plastic Shrinkage Crack Analysis for the Crimped Fiber  71  Table 4.6 - Summary of Plastic Shrinkage Crack Analysis for the 25 mm Polyolefin Fiber..73 Table 4.7 - Summary of Plastic Shrinkage Crack Analysis for the 50 mm Polyolefin Fiber...74 Table 4.8 - Unit Quantity of Fibers  80  Table 4.9 - Results of Plastic Shrinkage Crack Analysis of the Shrinkage Controlling Admixture  84  Table 4.10 - Results of Plastic Shrinkage Crack Analysis of the Waterproofing Admixture..86  vii  LIST OF FIGURES  Figure 2.1 - Effect of aggregate on plastic shrinkage. (From R.G. L'Hermite, Proceedings, Fourth International Symposium on the Chemistry of Cement, Washington, D.C., 1960, Vol.2 pp.659-694.) 18 Figure 2.2 concrete  -  Nomograph to calculate  the  rate of evaporation  from freshly  placed 19  Figure 2.3 - Plastic Crack Window  20  Figure 2.4 - Restrained shrinkage sample devised by Pihlajavaara and Pihlam  23  Figure 2.5 - Restrained shrinkage sample devised by Paillere et al  23  Figure 2.6 - Restrained shrinkage sample used by Pan et al  24  Figure 2.7 - Border restraint used by Shaeles and Hover  25  Figure 2.8 - Typical Ring Test setup  26  Figure 2.9 - Modified Ring Test setup used by Dahl  28  Figure 3.1 - Test setup for studying restrained plastic shrinkage cracking in shotcrete overlays 34 Figure 3.2 - Early restrained shrinkage sample  36  Figure 3.3 - "Coated edge" test sample  38  Figure 3.4 - Effect of finishing on crack area  42  Figure 3.5 - Comparison of crack areas with different shotcrete methods  56  Figure 3.6 methods  Comparison of maximum crack widths produced with different shotcrete 57  Figure 3.7 - Comparison of crack areas between cast and shot samples produced using the traditional dry method 58 Figure 3.8 - Comparison of crack areas between cast and shot samples produced using the Brazilian damp method 59 Figure 3.9 - Comparison of crack area between cast and shot samples produced with the Brazilian dry method 60 Figure 3.10 - Effect of paste content on crack area  62  Figure 4.1 - Hooked End Fiber; crack area vs. fiber volume  69  Figure 4.2 - Flattened End Fiber; crack area vs. fiber volume  70  viii  Figure 4.3 - Crimped Fiber; crack area vs. fiber volume  71  Figure 4.4 - 25 mm Polyolefin fiber; crack area vs. fiber volume  73  Figure 4.5 - 50 mm Polyolefin fiber; crack area vs. fiber volume  74  Figure 4.6 - Comparison of fibers; crack area vs. fiber volume  76  Figure 4.7 - Comparison of fibers; crack area vs. fiber mass  77  Figure 4.8 - Comparison of fibers; maximum crack width vs. fiber volume  77  Figure 4.9 - Comparison of fibers; crack area vs. number of fibers  79  Figure 4.10 - Fiber efficiency; crack area vs. batched fiber volume  82  Figure 4.11 - Shrinkage Controlling Admixture; crack area vs. admixture dosage  85  Figure 4.12 - Time vs. final crack width  87  ix  Chapter 1 - Introduction  1.1 - Plastic Shrinkage in Shotcrete  Plastic shrinkage cracking of dry-mix shotcrete is a common problem encountered within the construction industry. Dry-mix shotcrete is also used in a variety of applications, most commonly the repair and rehabilitation of concrete structures. Dry-mix shotcrete is also used as a tunnel liner, for slope stabilization, and for refractory applications. Plastic shrinkage cracks pose a significant threat to the durability and aesthetic appeal of concrete structures.  Shotcrete is more prone to plastic shrinkage cracking for several reasons.  Shotcrete  typically has high cement contents, low coarse aggregate contents and often contains accelerators, known to aggravate shrinkage. Almost all shotcreteing is placed over a rough surface that provides excellent bond. Such a bond allows for no movement in the shotcrete overlay so even small amounts of shrinkage causes strains sufficient to cause cracking in the plastic state.  1.2 - Previous Research of Restrained Plastic Shrinkage  Considerable research has been conducted on restrained shrinkage in concrete, although there are at present no standardized procedures for assessing the potential for cracking in restrained concrete in a drying environment. Recent test methods being used to study restrained shrinkage usually involve only a form of border restraint. These procedures are  1  generally only applicable to cast concrete, not shotcrete. It is difficult to produce shotcrete specimens that are small or have narrow constrictions within the formwork. During the shooting process, rebound material is trapped in the corners of the formwork that will generally spoil most small specimens. In shotcrete research, beams, cylinders and test samples are never produced directly; a large panel is first shot and the necessary samples are saw-cut or cored when the panel has obtained sufficient strength. The delay in obtaining suitable samples negates the possibility of any testing in the plastic state.  Existing test methods for restrained shrinkage either restrain or load the border of the specimen to develop or enhance shrinkage stresses. Bar and plate tests generally resemble tensile tests, but fail to model the continuous restraint produced by casting concrete over a roughened surface.  Recent research into restrained and drying shrinkage is dominated by the ring test. The ring test provides restraint via a continuous section of concrete cast around a rigid steel ring. Ring shrinkage testing is preferred because its geometry negates the effects of boundary conditions. The ring test illustrates the differences between the tensile strength of the concrete and the stresses produced by shrinkage, but does not factor in the importance of bond between the overlay and the base material. Once a crack is formed within a plain sample all restraint is lost. Continuous restraint has never before been used in shrinkage testing, but is the most commonly encountered condition observed in the field.  Much research has been done on early age plastic shrinkage, which usually involves casting large panels and subjecting them to a drying environment. Small thin slabs have been used  2  successfully to measure crack appearance. The large surface area of these panels is particularly well suited for modeling a shotcrete system.  1.3- Objectives  The objective of this study was to develop a test procedure to model restrained plastic shrinkage in dry-mix shotcrete overlays. Pre-made bases were constructed with a uniform surface roughness. A shotcrete overlay was then shot directly onto the base under laboratory conditions using industry standard equipment. The sample was immediately transferred to a drying chamber and monitored for crack formation. Once an experimental procedure was developed, it would be used to investigate the effectiveness of fibers in controlling plastic shrinkage.  3  Chapter 2 - Restrained Plastic Shrinkage of Dry-Mix Shotcrete Literature Survey and Presentation  2.1 - Introduction to Shotcrete  Shotcrete is a mortar or concrete that is placed using pressurized air to project and compact the material onto a receiving surface. In general terms, shotcrete is different from concrete in its method of placement, as well as in its material composition. As a specialized material, shotcrete is used for new construction and in repair and retrofit. All that is required for shotcrete to be placed is a receiving surface. When used for new construction of a freestanding structure, one-sided formwork is erected. If required a rebar mat or wire mesh is laid over the formwork to provide reinforcement. In tunnel linings and slope stabilization, the ground becomes the receiving surface and obviously requires no formwork. In a repair or retrofit application, the existing concrete acts as the receiving surface.  2.2 - Shotcrete Fundamentals: Wet-Mix Vs. Dry-Mix  Two distinct methods of shotcreteing exist and are referred to as the wet and dry-process. In the dry-process, cement and aggregates are batched together in a dry state and fed into a pump. The pump conveys the material through a hose to the nozzle where water is added to turn the dry material into a semi-plastic material. This semi-plastic material is then projected at a high velocity through a nozzle onto the receiving surface. After the "fresh concrete" is accelerated out of the nozzle, it is then considered to be shotcrete.  4  In the wet-process, cement, aggregates and water are batched together and fed into a pump. The pump conveys the material through a hose to the nozzle where compressed air is added to accelerate the material out of the nozzle. The "fresh" concrete within the hose is considered to be a shotcrete after it has been accelerated out of the end of the nozzle.  Both placement methods have their advantages and disadvantages and while both produce shotcrete, there are substantial differences in the final product. Dry-mix shotcrete was the primary focus of this project and was used in all but a few selected mixes.  2.3 - Dry-Mix Shotcrete 2.3.1 - Pumping Equipment  Production of dry-mix shotcrete requires the use of specialized pumping equipment. Dry-mix shotcrete is the traditional method of shotcreting and has been in use for over 80 years. Dry-mix utilizes a dry, pre-mixed material that is fed into the pump and transported in a pressurized air stream down the hose to the nozzle at a very high velocity. When the material reaches the nozzle, water is added and the material is accelerated onto the receiving surface where it is compacted.  2.3.2 - Dry-Mix Shotcrete Material  Like concrete, shotcrete contains cement, sand, coarse aggregate and water, however these basic constituents of the mix are present in different ratios leading to a material with slightly different characteristics.  5  In dry-mix shotcrete, aggregate gradation and proportion have a profound effect on the final product. The larger an aggregate particle, the slower its velocity will be when it is projected out of the nozzle. The slower a particle's velocity, the more likely it is to rebound off the receiving surface and not be incorporated into the in-place material. Thus a much higher proportion of the coarser fraction of aggregates rebounds. For this reason the coarse aggregate content in a dry-mix shotcrete is limited to a maximum aggregate size of 10 mm and a sand:coarse aggregate ratio of 3:1. Compare a 3:1 sand:coarse aggregate content in a typical shotcrete with conventional concrete having a sand:coarse aggregate ratio of 2:3, this reveals the major difference between concrete and shotcrete, see Table 2.1. The sand used in shotcrete is the same used for cast concrete and is usually specified as an ACI Gradation No.1, a typical concrete sand. Most commonly, shotcrete mixes will contain a 10 mm coarse aggregate and conform to an ACI gradation No. 2. Shotcrete containing no coarse aggregate, commonly referred to as a "gunite" contains only sand, blended to conform to an ACI gradation No. 1. Although seldom used, a shotcrete mix containing a 12 mm aggregate is referred to as an ACI gradation No. 3. Gradation limits for shotcrete aggregate are provided in Table 2.1.  6  Table 2.1 - Gradation Limits for Combined Aggregates (ACI 506R-8) Percent by Weight Passing Individual Sieves Sieve Size  Gradation No. 1  Gradation No. 2  Gradation No. 3  Typical Concrete  40 mm  100  28 mm  97-100  19 mm  100  14 mm  60-72 40-70  12 mm  100  80-95  10 mm  100  90-100  70-90  40  4.75 mm  95-100  70-85  50-70  38-40  2.4 mm  80-100  50-70  35-55  32-40  1.2 mm  50-85  35-55  20-40  20-36  600 urn  25-60  20-35  10-30  10-26  300 urn  10-30  8-20  5-17  4-14  150 urn  2-10  2-10  2-10  1-4  Several different methods of supplying dry-mix shotcrete material are used depending on available equipment and site conditions. Shotcreteing materials may be supplied in either a pre-bagged form, delivered in bulk dry form, or in a concrete ready-mix truck  The most common method of dry-mix shotcreteing in Canada uses a pre-batched bagged material. This material is supplied in either small disposable paper bags or in large bulk bin bags. The paper bags are typically 25 to 35 kg each and are transported on pallets. The large bulk bin bags typically weigh 1600 to 2000 kg. The small bags are preferred for low volume operations when starting and stopping is required, the bulk bin bags for large volume operations. Pre-bagged material is preferred for several important reasons. With prebagged material mix variability is greatly reduced. Material proportioning is taken out of the hands of the construction personnel, where site conditions are often chaotic and  7  uncontrolled, and placed under the supervision of a material supplier where factory control is regular and methodical. A supplier can easily incorporate admixtures or fibers to the mix, often impossible to add on site. Pre-bagged material also has the benefit of being easily transportable over long distances and to isolated locations. The material is easily stacked and stored for long periods of time in sometime hostile environments.  Although uncommon, site batching of dry-mix is used in some regions of North and South America. Stockpiles of aggregate and cement are stored on site and batched as required. Materials are batched volumetrically or by mass, leading to variations in the mix due to differing moisture contents and densities within the cement and aggregates.  Dry-mix can be supplied out of a ready mix truck if the concrete supplier has access to dried aggregates. If the haul distance is short, damp aggregate can be used. Cement and aggregates are simply fed into the concrete truck and transported to site.  2.3.3 - Dry-Mix Shotcreteing Production Methods  Dry-mix shotcrete is produced using several different methods depending on the type of material and equipment available, these methods include; the Bone Dry, the PreMoisturized, the Wet Sand, and the Brazilian.  2.3.3.1 - The Bone Dry Method  The Bone Dry method, as the name suggests, uses a "bone dry" material that is pumped and placed. This method requires only a minimal amount of equipment and usually uses  8  pre-bagged material. The material is conveyed through the line in its dry state, with the total amount of water being added only at the nozzle. While very flexible for start and stop applications, this method suffers from several disadvantages.  Large amounts of dust are  produced at the pump as the mix is discharged into the hopper. Poor mixing at the nozzle prevents a large amount of the cement in the mix from receiving any water, resulting in a large amount of rebound and excessive dust., Unless meticulous shooting techniques are enforced, this method tends to produce shotcrete subject to sand lenseing. Sand lenses are produced when an intermediate layer of shotcrete receives too little water resulting in a dry area. Such shotcrete has planar weakness zones that greatly reduce strength. To overcome this problem of poor mixing and dust inherent with this technique, the PreMoisturizing method was introduced.  2.3.3.2 - The Pre-Moisturizing Method  The intent of the Pre-Moisturizing method is to reduce the amount of rebound and dust produced, which is accomplished by introducing a minimal amount of water in the "bone dry" material to dampen the mix. The material is first sent through a pre-moisturizer before it is discharged into the shotcrete pump. The pre-moisturizer consists of a hopper that feeds into a rotary auger, sheathed in a flexible rubber sleeve. As the material moves up the auger path, small water jets inject moisture into the material giving it the consistency of damp earth.  By pre moisturizing the material before adding it to the pump, less dust is produced  both at the pump and at the nozzle. Furthermore, the small amount of moisture in the dry material aids in mixing at the nozzle and reduces the likelihood of wet and dry lenses in the inplace material. Difficulties arise in this method when accelerators are used in conjunction  9  with low pumping rates. If the accelerators are aluminate based, a flash set may prematurely occur within the pre-moisturizing auger.  2.3.3.3 - The Wet Sand Method  The wet sand method, popular in bulk volume jobs and in isolated areas, is similar to the pre-moisturized method. The shotcrete mix is site batched with cement and damp sand usually on a volume basis. The damp sand is kept at a moisture content of 2-4% and mixed with cement. Both materials are added to a portable mixer, mixed sufficiently, then discharged into the hopper of the shotcrete pump. Using this method, sand can be shipped in bulk and left exposed to the elements beside the batching plant, only the cement must be stored and protected.  The Wet Sand method is common in the southeastern United States; in Canada it is very uncommon. Lack of strict quality control will lead to varying sand:cement ratios and water contents. The material must be pumped and shot quickly to avoid premature set.  2.3.3.4 - The Brazilian Method  Developed in Brazil in the early 1980's, the Brazilian method is a hybrid of the Bone Dry and Pre-Moisturizing method. The basis of the system is a water injection ring located within the material line. The pre-moisturizing ring uses high-pressure water to pre-dampen the material as it travels down the hose. The material is fed into the shotcrete pump in a bonedry state and pumped down the line. Upon reaching the water ring, usually several meters before the nozzle, the material is saturated by a mist of high-pressure water. The fine spray  10  of water and turbulence within the hose allows for adequate mixing of the material. However, since the entire amount of water is added to the bone-dry material instantaneously, a small amount of paste is produced that coats the inside of the hose. Not adhering to any aggregate particles, this remaining paste tends to dribble from the end of the nozzle onto the floor and is known as "nozzle slobber". The dust produced at the nozzle with this method is comparable to the Pre-Moisturizing method, however the dust produced at the pump is equivalent to the Bone Dry method. The Brazilian method produces a shotcrete that will tend to stiffen seconds after shooting. So while a mix may be shot at an apparently wet consistency, after a very brief time it will gain the appearance of a dry consistency. This stiffening effect is due to the moisture transfer from the paste into the oven dried aggregate. This water migration may tend to decrease the water/cement ratio around the aggregate and create a better paste/aggregate bond.  The Brazilian pre-moisturizing ring is well suited for using dry powdered accelerators as no water is added into the material until a fraction of a second before the material is placed on the wall. In the other systems, the material may be damp for several minutes before it is applied, giving the accelerator time to react and lose its effectiveness.  2.3.4 - Uses of Dry-Mix Shotcrete  Shotcrete is a versatile material used in many applications. It is commonly used in tunnel linings and slope stabilizations where the use of pre-bagged material is of particular benefit in isolated locations. Dry-mix shotcrete is extensively utilized in the repair of deteriorated concrete structures where its flexibility in placing small quantities of material quickly makes it very competitive to other repair techniques.  11  2.3.4.1 - Repair and Rehabilitation  With the large number of existing concrete structures undergoing severe deterioration due to loading, physical and environmental factors, much of the focus of the concrete industry is on repair and rehabilitation.  The repair of concrete structures involves the removal and replacement of deteriorated concrete with a new material. The success of a repair will depend largely on proper surface preparation to achieve a sound bond. Shotcrete, as a material, has a very high capacity for bond. It contains a high percentage of cement, a low water cement ratio and is forcefully compacted during the shooting process.  2.3.4.2 - Substrate  In the repair rehabilitation of aging structures, shotcrete is a popular and effective technique. Damaged areas are cleaned of spalled and loose concrete by using jackhammers or highpressure water. If jackhammers are used, the area must be further cleaned using either high-pressure water or sand blasting to remove the bruised concrete and thoroughly clean any exposed reinforcing steel. As the paste fraction of the concrete is much weaker than the aggregate, exposed coarse aggregates protrude from the surface of the repair and create an uneven surface. The surface roughness enhances the bond of the shotcrete to the original concrete by providing a sound mechanical bond with a large surface area. During the shooting process, there is an abnormally high rebound during the first few seconds of application. Until a small layer of paste can form on the receiving surface almost 100% of the coarse aggregate fraction rebounds. The first layer of paste is further compacted by the  12  impinging material placed above it, and presumably forced into the pores of the repair surface. If properly applied the shotcrete bond interface should be the strongest portion of the system.  2.4 - Shrinkage  It is well-known that when concrete is subjected to a dry environment it will shrink. Drying shrinkage is a long-term problem encountered in concrete construction with the majority of the strains developing over the first year. Good design practice can alleviate problems associated with normal drying shrinkage. Drying shrinkage involves the loss of water from the interstitial spaces of the gel particles, which causes an irreversible collapse of the layered microstructure and a resulting reduction in volume. Plastic shrinkage occurs when concrete is still in the fresh state and is much more of an immediate problem. Plastic shrinkage is a simple phenomenon that involves the loss of mix water by evaporation from the fresh concrete. Although both types of shrinkage manifest themselves in shotcrete, plastic shrinkage will be the focus of this thesis.  2.4.1 - Material Related Shrinkage  Shrinkage can be a serious problem in dry-mix shotcrete for several reasons. To avoid very high rebound dry-mix shotcrete usually contains a low coarse aggregate content. A high cement content is then necessary to coat the sand fraction adequately. Upon shooting, the bulk of the rebound is the coarse aggregate fraction, which further increases the in-situ cement content. Thus while the original batched shotcrete may contain about 400 kg of cement per cubic meter of concrete, the material that is left on the wall may have well over  13  500 kg/m of cement. In cast concrete it is primarily the coarse aggregate fraction of the mix 3  that provides the dimensional stability to the material and resists the shrinkage strains of the paste. In shotcrete, not only is the maximum size of coarse aggregate small, but there is only a very small amount of it incorporated into the final material. Furthermore, to aid in placement or to provide immediate support, accelerators are often used; this may further aggravate the shrinkage problem.  2.4.2 - Placement Related Shrinkage  After the shotcrete is applied, it is often difficult to apply proper curing techniques. Like conventional concrete, shotcrete must be properly cured to achieve its maximum potential. On high slopes or in isolated locations it is often difficult to apply a curing membrane or place soaker hoses. In hot, windy environments the large surface area of shotcrete is particularly prone to excessive evaporation resulting in plastic shrinkage. Shotcrete is almost always applied as a thin membrane that is physically restrained on one side from movement. Whether the base material is rock, soil or old concrete it is almost always dimensionally stable. The strain differential between the two materials results in the weaker of the two - the shotcrete - failing. The failure that results is a tensile failure within the shotcrete and is seen as cracks that can be randomly oriented such as pattern or map cracking, or orientated parallel to the long direction. The cracks are usually the full depth of the shotcrete layer and can vary in width anywhere up to a few millimeters.  14  2.4.3 - Consequences of Shrinkage  Restrained shrinkage in unreinforced shotcrete will result in cracking, which will be of great concern in some structures. Cracks will increase the permeability of the structure to both aggressive liquids and ions, which leads to exposure of the reinforcing steel and risk of corrosion. Corroding reinforcing steel within the shotcrete reduces the structural capacity of the section. The formation of rust will cause spalling of the cover and will lead to a progressive failure of the structure. Furthermore in a tunnel, slope stabilization or any water retaining structure a high permeability is unacceptable. Lastly, cracking has the effect of making the structure look aesthetically unpleasing, which causes the public to question its strength and effectiveness.  2.4.4 - Control of Shrinkage in Shotcrete  As the consequences of ignoring shrinkage in shotcrete applications are severe, much effort has been placed into controlling the shrinkage. Various attempts at controlling shrinkagerelated cracking in shotcrete have included: fiber reinforcement, shrinkage controlling admixtures, control joints and reduction in cement through use of supplementary cementing materials. Fiber reinforcement was of particular interest in this study. Also, two admixtures that purported to reduce shrinkage were assessed.  2.4.4.1 - Fiber Reinforcement  The use of fiber reinforcement in shotcrete applications has been commonplace for the last 20 years in North America. Fiber reinforcement provides shotcrete with a measure of  15  ductility. Fibers, provide a different function when compared to reinforcing steel, and are easily added to a shotcrete mix to provide a measure of toughness.  There is evidence of preferential fiber alignment caused by the shooting process that makes fibers particularly useful in shotcrete. During the shooting process, fibers become embedded on the surface aligned perpendicular to the shooting direction and become randomly oriented in a two dimensional plane. This orientation provides maximum flexural strength in plate structures, and efficiently resists planar tensile strains developed through shrinkage.  2.4.4.2 - Shrinkage Controlling Admixtures  Shrinkage compensating admixtures are usually based on either calcium sulfoaluminate or calcium aluminate with calcium oxide. The aluminates react with calcium hydroxide in an expansive reaction. Difficulties however arise in determining the proper dosage levels required.  2.4.5 - Shrinkage Cracking  In reinforced concrete structures, inadequate curing can cause the formation of micro and macro cracks soon after placement. These cracks can widen further due to mechanical loading on the structure or due to environmentally induced stresses. Once the cracks widen to a critical width, reinforced concrete is susceptible to freeze/thaw, and chloride attack. In view of the harmful effect of these cracks on the structure's durability it is important to avoid them.  16  Any volume changes that occur in concrete from a loss of moisture are termed as shrinkage. More specifically shrinkage can be divided into three more specific categories. Plastic shrinkage refers to the evaporation of water from fresh concrete. Drying shrinkage refers to a long-term effect on hardened concrete. Autogeneous shrinkage is the self-desiccation of mature concrete and often occurs in conjunction with drying shrinkage.  Early-age shrinkage cracking can be viewed as due to one of two causes: loss of moisture, or a temperature differential.  2.4.5.1 - Crack Formation due to Moisture Loss  Cracking can occur in shotcrete due to the loss of mix water to the environment. When this occurs in the fresh concrete, it is termed plastic shrinkage, in hardened concrete it is termed drying shrinkage.  2.4.5.2 - Plastic Shrinkage  In fresh concrete some water will rise to the surface of the mix during mixing and placement. As the water reaches the free surface it evaporates off. Water loss can also occur if it is absorbed into the sub-base or formwork. If the evaporation rate is higher than the rate of bleeding, capillary menisci will form. The reduction in volume caused by the pressure exerted by the capillary menisci, forces the concrete to compress upon itself, it is this compression that manifests itself as plastic shrinkage. Plastic shrinkage cracking takes place in the paste fraction of the concrete when the capillary pressures reach a critical pressure. As seen in Figure 2.1, aggregate has a large effect on the amount of plastic  17  shrinkage that will occur. The coarse aggregate has a stabilizing effect on the paste volume of the concrete. Paste 6  j 4 -  Mortar l/^  2 -  1  Concrete  0  0.25  1  1  0.5  1.0  1  1  2  4  1 8  Time (h)  Figure 2.1 - Effect of aggregate on plastic shrinkage. (From R . G . L'Hermite, Proceedings, Fourth International Symposium on the Chemistry of Cement, Washington, D.C., 1960, Vol.2 pp. 659-694.) Plastic shrinkage, as the name suggests, will only occur during the first few hours after casting while the concrete is still in a plastic state. The study of plastic shrinkage cracking is complicated because the material properties that determine if cracking will occur are time dependent and change rapidly during the first few hours in the life of the concrete. Primarily, the amount of plastic shrinkage cracking that occurs is dependent on the rate of evaporation of surface water from the concrete. Evaporation is dependent on the wind velocity, relative humidity of the air, ambient temperature, and the concrete temperature. These variables can be used to estimate the amount of evaporation using the ACI nomograph, see Figure 2.2. Any precautions taken to reduce the previous conditions can reduce or control the amount of plastic shrinkage cracking. The threshold for formation of plastic shrinkage cracks is around 0.5 kg/m /hour, but will vary depending on the particular mix. However, not 2  18  only environmental conditions can affect the formation of plastic shrinkage cracking. The consistency of the mix at the time of casting and the rate at which the material gains tensile strength is also a major factor. In a fluid concrete that is undergoing shrinkage due to loss of water, there may be enough creep capacity in the material to resist cracking. A very stiff mix may have very little capacity to shrink. Small strains produced by a loss of water may result in relatively large tensile stresses.  To use This chart: 1. Enter with air temperature, move up to relative humidity. 2. Move right to concrete temperature. 3. Move down to wind velocity. 4. Move left: read approx. rate of evaporation.  Figure 2.2 - Nomograph to calculate the rate of evaporation from freshly placed concrete.  Plastic shrinkage cracking will only occur over a specific period of time, which will be defined in this report as the plastic crack window. Plastic shrinkage cracking window will not occur  19  until tensile stresses exceed tensile strengths. The window will close when the concrete has gained sufficient strength that exceeds the tensile stresses generated by shrinkage.  Plastic C r a c k W i n d o w  Time Figure 2.3 - Plastic Crack Window  2.4.5.3 - Cracking due to Drying  Where plastic shrinkage cracking can be avoided by proper curing techniques, drying shrinkage is inevitable. If shrinkage strains are not allowed for in concrete design, stresses in restrained members can lead to cracking or warping. Expansion joints can provide the necessary movement to prevent stresses from occurring, or to prevent random irregular cracking. If the cracking is limited to a specific section, a sealant can easily be applied to maintain integrity. Drying shrinkage is a complex process that is not fully understood. As the humidity drops there are four levels of water loss that result in drying shrinkage. Upon initial drying of a  20  concrete member, water is lost from the capillary pores of the cement paste. Upon further drying, adsorbed water from the surfaces of the C-S-H particles is lost. At very low humidities, there is a loss of water from the structure of the C - S - H . Finally, at elevated temperatures, the C-S-H particle itself will decompose.  2.4.5.4 - Cracking due to Self-Desiccation  In concrete mixes with a low w/c ratio there is a possibility for shrinkage cracking to occur even if no moisture is lost to the environment. Termed auto-geneous shrinking, this can occur if a volume reduction occurs due to the hydration reaction and the relative humidity drops below 100%. Theoretically this could occur at a w/c ratio below 0.42, however at w/c ratio's below this level additional un-reacted cement will remain as a micro-filler and can act to strengthen the concrete. It is unimportant whether shrinkage occurs because of internal or external means.  2.5 - Review of Restrained Shrinkage Testing  While free shrinkage tests can adequately assess the magnitude of strains developed in concrete specimens they cannot predict the stresses and possible failures induced by these strains. A S T M C157, Standard Test Method for Length Change of Hardened HydraulicCement Mortar and Concrete, uses a prismatic specimen that is 250mm in length with a square cross section. By making the specimen in the shape of a long rectangle, the majority of the strains are in one direction. The shrinkage is predominantly uni-directional, exhibiting a uniaxial tensile strain on the sample in the longitudinal direction. This test will satisfactorily predict the free shrinkage strains in a concrete member. Actual structures are often not  21  able to freely shrink; restraint is imposed either through the foundations or from a bond to another stiffer member. The effect of this restraint on the shrinkage is to induce stresses.  The bulk of previous testing in restrained shrinkage resembles uniaxial tensile tests. Many researchers have evaluated linear specimens of various geometries. The specimen is either "dog-boned" or secured to the testing frame through an epoxy interface. A notch is sometimes used to initiate failure. A uniaxial tension test can provide true material properties that would be independent of the specimen's geometry. Obtaining sufficient bond at the extremity of the sample can also prove challenging. If the testing apparatus is not sufficiently stiff a non-uniform load may be applied.  Another popular method of restrained shrinkage testing involves using a steel ring. In this method, concrete is cast around a heavy steel ring and allowed to set. After demolding the concrete sample shrinks. The steel ring, causing tensile stresses to form in the sample resists the imposed shrinkage strains. As the drying continues, the sample will fail in tension, producing a visible crack on the surface.  2.5.1 - Linear Specimen Testing 2.5.1.1 - Beam Specimens  Linear beam specimens were subjected to long term deformation tests on glass fiber reinforced concrete, (Pihlajavaara and Pihlman, 1978). Samples, 500 X 80 X 20 mm were restrained by epoxying the ends to a rigid steel frame, see Figure 2.4, and subjected to a relative humidity of 40% and a temperature of 20°C over a period of five months. Plain  22  specimens cracked after a period of 12 hours to 2 days, G F R C specimens at fiber dosages of 2 to 6% remained uncracked during the test.  Figure 2.4 - Restrained shrinkage sample devised by Pihlajavaara and Pihlam (1978).  2.5.1.2 - Dog-Boned Specimens  Dog boned samples, see Figure 2.5, 1.5 m in length, with a cross-sectional area of 85 X 120 mm were end restrained to produce shrinkage strains, (Paillere et al. 1989). A steel frame physically held the samples at a constant length, when the samples underwent shrinkage they were held in tension. A dynamometer was placed on the frame to measure stresses exerted on the sample. Plain concrete samples cracked after a short period of time, the addition of fibers prevented or prolonged the time to cracking.  Figure 2.5 - Restrained shrinkage devised by Paillere et al.  23  2.5.1.3 - Internal Restraint  Pan et al. (1987) used a novel form of restraint involving an internally placed section of reinforcing steel, see Figure 2.6. The ends of the bar are firmly bonded to the ends of the beam with the mid section of the reinforcing bar being encased in a rubber tube. The rubber tube prevented bonding of the steel to the concrete in the mid-section, thus providing restraint as the beam underwent shrinkage. The specimens were tested under controlled climatic conditions of 20 - 25°C and relative humidities of 40 - 55%. The effect of Fly Ash was assessed.  -«4  4 •  r© ' r ®  •  r®  J-J  Figure 2.6 - Restrained shrinkage sample used by Pan et al (1987).  2.5.2 - Plate Specimen Testing 2.5.2.1 - Continuous Edge Restraint  Kraai (1985) developed a test method to test concrete plates subjected to restrained shrinkage. Concrete plate samples 900 X 600 X 20 mm were subject to severe drying environments. Testing was carried out in a sealed room that maintained constant environmental conditions throughout the test. Temperatures ranged from 25 to 35°C, with relative humidities between 10 and 25 %. Two large fans produced wind speeds between  24  3.1 and 3.6 m/s. The samples were cast in wooden forms limited restraint provided by light gage wire mesh around the edges of the sample. In the test, two companion samples were cast simultaneously, and tested identically to assess the change in a single material property.  Shaeles and Hover (1988) used the same experimental set-up as Kraai to examine plastic shrinkage cracking. Their test setup used Plexiglas forms with expanded metal lath attached to the inside perimeter to provide the necessary edge restraint, see Figure 2.7.  A  0.75 in.  24 in.  B  0.7Sii  Figure 2.7 - Border Restraint used by Shaeles and Hover.  2.5.2.2 - Discrete Edge Restraint  Opsahl and Kvam (1982) developed a test method that used a plate sample that developed restraint by placing four stirrups directly into each end of the slab. The stirrups were anchored to a rigid steel frame. The sample was subjected to a severe drying environment over a period of many days. Samples underwent both plastic and drying shrinkage. The  25  efficiency of enlarged end fibers was assessed. Fiber volumes as low as 0.1% were found to prevent cracking, while plain concrete samples cracked from plastic and drying shrinkage.  2.5.3 - Ring Specimen Testing i The use of ring shaped samples to assess restrained shrinkage has been widely studied. Ring specimens benefit from the fact that they are easily produced: not requiring any special type of restraint. The ring test uses a small "doughnut" of concrete cast around a rigid steel ring, see Figure 2.8. As the concrete undergoes shrinkage, the restraint provided by the steel ring forces the concrete sample into tension. Drying from only the outer surface can be obtained by sealing the top and bottom surfaces to ensure uniform shrinkage.  1  254  1  305  I  T  - 4 yl  n  (all dimensions are In millimeters)  Figure 2.8 - Typical Ring Test setup.  26  The ring specimens function by internalizing stresses produced from shrinkage strains, as the concrete shrinks, the inner steel ring resists the strain, forcing the sample into tension. With the standard size of ring sample, see Figure 2.8, it can be shown that the difference between the tensile hoop stress on the inner and outer surface of the sample is about 10%. This stress differential biases crack initiation on the outer edge of the sample. Furthermore the maximum radial stress within the sample is only 20% of the maximum hoop stress. Thus the ring test can be closely compared to a uniaxial tensile test with some extraneous stresses.  Ring samples were first used as early as 1939 by Carlson and Reading, (re-published in 1988). Carlson and Reading used a 25 mm thick, 38 mm wide sample cast around a steel ring, 25 mm thick, with an outside diameter of 175 mm. Samples were subjected to drying at various humidities to assess the effect of mix design. It was found that the cement content of the mixes had little effect on the amount of cracking. Coarse aggregate type was found to have a drastic effect on the ability of the sample to resist cracking.  Restrained plastic shrinkage of samples was measured with the use of somewhat larger than standard ring specimens by Dahl (1985). Specimens, 80 mm thick, 150 mm wide cast around a steel ring with an outside diameter of 280 mm, see Figure 2.9. 12 steel ribs were cast into the outer core to strengthen the restraint. The samples were exposed to a drying environment of 40% relative humidity and a temperature of 2 0 ° C with a wind velocity of 4 m/s. Testing showed that the use of 0.1% polypropylene fibers significantly improved the plastic shrinkage resistance of the samples. Crack widths were reduced from 33 -87%.  27  This test setup suffers from several obvious problems. The presence of the steel ribs acts as a stress raiser within the sample. Along the outer diameter of the ring, where the tensile hoop stresses are already the greatest, the ribs act as a crack initiator. Secondly, the fan placement would cause differential drying within the samples. Moisture evaporating from the first sample would be blown over the second and third sample increasing the humidity  Specinens  Figure 2.9 - Modified Ring Test setup used by Dahl.  2.5.4 - Restrained Plastic Shrinkage Tests  The formation of plastic shrinkage cracking is dependent on the development of the tensile strength of the material. If the tensile strength is larger than the tensile stress in the material, no cracking will occur. Ravina and Shalon  (1971) noted that concrete slabs were less likely  to crack if cast in direct sunlight when compared to slabs cast in the shade. It is their theory that the thermal heat of the sun causes an accelerated strength gain in the concrete, enough to resist the strains caused by plastic shrinkage. They further stated that the warming of the slab surface may have also caused an expansion in the top surface of the slab that offset shrinkage strains.  28  Shaeles and Hover (1988) used sand cement mortars to study restrained plastic shrinkage cracking. The consistency and temperature of the mortar was found to have influenced the time at which cracks occurred and the length of time over which cracking occurred. The use of stiffer mixes and/or higher air temperatures resulted in the earlier appearance of cracking and a decrease in the duration of the plastic crack window. More fluid mixtures resulted in a delay in the formation of cracking and an increase in the duration of the plastic crack window. The extent of cracking was found to be less severe in stiffer mixes. The paste content of mixes was found to drastically affect the amount of plastic shrinkage cracking. Under identical conditions, a sample with a lower paste content was found to exhibit much less plastic shrinkage cracking.  Shaeles and Hover (1988) reported that finishing operations had a profound effect on the degree and direction of plastic shrinkage cracking. Samples were cast and then immediately subjected to a severe drying environment. Similar to common construction practice, the samples were finished after initial set, but before final set, when the concrete was judged to have obtained sufficient stiffness.  Finishing operations began between 80  and 110 minutes after casting. Preferential cracking was found to form perpendicular to the direction of screeding. They concluded that the screeding operation formed minute tears in the surface of the plastic mortar. Given the viscous nature of the concrete, the formation of these tears or tensile stresses would be dependent on the rate of screeding: the faster the screeding operation, the greater the extent of cracking. Samples screeded at a rate of 10m/minute exhibited substantially more cracking than ones screeded at 1m/minute.  Shaeles and Hover (1988) performed tests on unfinished mortar samples. They found that unfinished samples consistently experienced substantially more cracking than finished  29  samples. They concluded that the finishing operation closed cracks and allowed them to "heal". Weight loss measurements of the samples during testing concluded that there was an increased water loss immediately after finishing, presumably the loss of bleed water that was forced to the surface during finishing. However, after evaporation of the surface bleed water, finished samples lost much less water to evaporation than unfinished samples. They concluded that the finishing process smoothed and densified the surface layer of the mortar, making it more resistant to evaporation.  Banthia et al (1996) developed a test procedure to reflect realistic restraint conditions on an overlay subjected to early age shrinkage. The rigid base consisted of a high strength concrete measuring 1010 mm long by 100 mm wide and 40 mm deep. The base was given an enhanced roughness by manually placing 20 mm coarse aggregate within the surface. The rigid, roughened base was cured in hot water to stabilize shrinkage. A 100 mm deep overlay was cast over the sample and immediately transferred to an environmental chamber. The sample was subjected to a constant temperature of 3 8 ° C and a relative humidity of 5%. Cracking of the sample was visually monitored periodically for 48 hours after casting.  Banthia et al recorded maximum crack widths of 2.83 mm in unreinforced samples.  Steel  fibers were found to reduce the maximum crack width as well as causing multiple cracking in samples.  2.6 The Effect of Fiber Reinforcement on Restrained and Free Shrinkage  Considerable research has been carried out on the ability of fiber reinforcement to reduce the amount of cracking in restrained shrinkage test samples. There appears to be general  30  agreement that fibers, even at very low dosages, can substantially reduce the amount of cracking in restrained shrinkage tests. Fibers are generally reported to have no effect on the free shrinkage capacity of concrete samples. The ability of fibers to control shrinkage cracks depends on how well they prevent cracks from widening. In fiber reinforced concrete, fibers bridging the crack will provide resistance to crack widening, which results in a tensile stress in the uncracked portion. Additional tensile stresses in the uncracked portion can form additional cracking, but the cumulative crack width is reduced.  Kovler et al (1992) studied the effect of polypropylene fibers on free and restrained drying shrinkage. Restrained shrinkage testing was conducted with the use of a modified ring test. The polypropylene fibers were found to significantly reduce the crack widths. They found that cracking occurred at the same time for fiber reinforced mixes and plain mixes, but cracks in the unreinforced mix attained greater widths. A rougher surface texture on the fibers was found to decrease crack widths and lengths by an order of magnitude. The fibers were noted to have had no influence on the free shrinkage characteristics of the test specimens.  Sarigaphuti et al (1993) studied the effect of cellulose and polypropylene fibers on free and restrained drying shrinkage. Restrained shrinkage testing was conducted with the use of a ring test. Free shrinkage testing was based on A S T M C 157. A significant reduction in the crack widths was observed with fiber addition in the restrained shrinkage testing. The free shrinkage properties of the concrete were not changed substantially by the addition of fibers.  Balaguru (1994) studied the effect of steel, synthetic, and cellulose fibers, and fiber geometry on plastic shrinkage. Tests were conducted using 600 X 900 X 19 mm slabs using  31  various forms of edge restraint. Edge restraint of the forms consisted of wire mesh nailed to the form edge. The overlay would bond to the wire mesh and be physically restrained from moving. Balaguru found cracking started to develop about 3 to 3.5 hours after casting and ceased about 8 hours. The results of the testing program showed the following trends;  •  Fibers with the highest aspect ratio provided a better performance.  •  Increase in fiber content led to reduction in crack areas.  •  Fiber count, number of fibers per unit mass was found to influence cracking.  •  Fibers with a lower modulus of elasticity were found to provide better crack reduction.  Malhotra et al (1994) studied the effect of polypropylene fibers on free shrinkage in shotcrete applications. They found that the shrinkage strains were virtually unaffected by either the mix proportions or fiber content of the concrete.  Shah et al (1993) studied the effect of steel, polypropylene and cellulose fibers as well as steel wire mesh on free and restrained shrinkage. Restrained shrinkage was analyzed with the use of ring test. Fibers and wire mesh were found to substantially reduce the amount of cracking in restrained shrinkage test samples. The addition of steel or cellulose fibers was found to not substantially alter the drying shrinkage of test specimens.  32  Chapter 3 - Development of the Test Procedure  During the early stages of this project, much work was done in refining the test procedure. Variations of specimen size, base preparation, mix design, shooting procedure affected the test results. Modification of these parameters was studied to maximize the amount of cracking in a plain, unreinforced sample. Early work was plagued with a lack of plastic shrinkage cracking. The formation of no cracking in plain samples, even under severe conditions, left no room for improvement in plastic shrinkage cracking resistance.  3.1 - Initial Stages 3.1.1 - Roughened Base Development  The first series of test bases was produced in small quantities as required by the testing program. The bases were made with a high strength silica fume concrete similar to those used by Banthia et al (1995). Immediately after the bases were cast, selected coarse aggregate were hand placed into the surface of the plastic concrete. The resulting surface was rough, and was considered to produce a maximum bond. These bases were time consuming to make, so their use was discontinued.  The second series of bases was produced en-mass six months prior to the beginning of the test program. The test bases measured 900 mm by 300 mm by 50 mm thick and were cast using a conventional concrete, see Table 3.1. "Green-cutting", with the use of a surface retarder, was used to create the roughened surface. The crushed coarse aggregate fraction of the mix provided an excellent bond. By casting all test bases simultaneously a uniform surface roughness was achieved, thus eliminating any variability in the test program.  33  Table 3.1. - Roughened Test Base Concrete Property  Value  Strength  35MPa  Air Content  5%  Slump  80 mm  Coarse Aggregate Type  25 mm crushed  w/c  0.40  3.1.2 - Test Chamber Development  The test chamber was constructed with 17mm thick plexiglas for a cover, see Photo 6 and Figure 3.1. Three electric heaters provided the drying conditions necessary for the test. Airflow diverters were used to channel the wind and heat directly over the samples during testing. Temperature within the drying chamber was maintained at 5 0 ° C , wind velocities at 15 km/hr. Measured evaporation rates were maintained at 1.5 kg/m /hour. 2  Figure 3.1 - Test setup for studying restrained plastic shrinkage cracking in shotcrete overlays.  34  Sample placement within the environmental chamber resulted in a uni-directional drying front. Temperature within the chamber could be considered to be uniform throughout, however since the wind source always came from the same direction, there would be some variation in humidity within the chamber. As the wind blew over the sample, moisture would be drawn into the air, decreasing its drying capability. Cracks were observed to form preferentially on the side of the sample facing the heaters, where the drying effects were the greatest. A swivel base was constructed to spin the sample during testing to provide a more even drying environment. No cracking was formed during the use of the swivel base. The use of the swivel base was discontinued due to the difficulty encountered in timing its movement. The problem of the unidirectional drying front was never resolved.  3.1.3 - Formwork Development  Early in the pre-test program it was found that unless form removal occurred soon after initial set of the shotcrete mix, no cracking would appear. It is theorized that the presence of the formwork prevented enough drying from the overlay to prevent cracking. If the formwork was removed the additional surface area increased the likelihood of cracking. The formwork was designed to be removed quickly and easily without damaging the fresh concrete.  Curling of the slabs was found to be a major problem in early tests. The ends of the overlay would "curl" up and relieve the shrinkage strains. Attempts to prevent curling of the sample involved several novel approaches; non-of which were successful. To prevent desiccation of the sample edges and base and hopefully to prevent curling, barrier methods were first attempted. These attempts involved painting the base and the edges of the sample with a coal tar paint immediately after form removal. Later, plastic shrink-wrap was used in conjunction with the coal tar paint to produce a totally impermeable membrane. After painting the section, a layer of plastic shrink-wrap was wrapped over the coal tar paint. It  35  was hoped that the presence of an impermeable membrane would prevent evaporation of water, but not provide any physical restraint that would prevent cracking. This method provided no resistance to curling in the overlay.  Attempts to control curling also included placement of steel weights onto the edges of overlay. It was hoped that the weight of the steel would physically prevent the edge from curling. The samples were still found to curl even with the use of heavy weights.  To create a sample that would resist curling and prevent the roughened base from drying out concrete was cast over five sides of a roughened base. In this method, all sides of the base, except the bottom surface, would be covered with a 50 mm concrete cover. The roughened base is placed within an oversized set of forms, and filled level with concrete. When the use of the "coated edge" model of sample proved unsuccessful in cast samples, attempts were made to strengthen the border of the samples. As previous samples seem to always fail at the edge of the base an attempt was made to strengthen the edges. A small portion of the fresh concrete was set aside immediately after mixing. This material was then mixed in with steel fibers at a dosage of about 1 % by volume. The fiber reinforced concrete was cast in at the ends of the formwork, see Figure 3.2, with the plain mix being cast into the center section. This method met with limited success. The fiber reinforced concrete had a greater tensile strength and was able to transfer enough shrinkage strain to the overlay necessary for cracking.  Three companion samples  Rigid, roughened base  •  Figure 3.2 - Early restrained shrinkage sample.  36  It was not until an actual shotcrete mix was used in the "coated edge" model that the test proved successful. The superior bonding characteristics and inherent stiffness of the shotcrete allowed the overlay to obtain enough bonding strength to produce restrained plastic shrinkage cracking.  3.1.4 - Sample Development  It was originally hoped that relatively small sized samples could be produced for the testing program. Smaller samples were more economical to produce and easier to handle. Three companion samples were cast onto one roughened base. Each sample measured 75 mm in width, 850 mm long and 60 mm thick. By casting three companion samples onto one base the variability of the roughened base was not a factor. The use of three samples provided a better statistical variation. Furthermore, the increased surface area provided a larger potential for evaporation.  A bonding agent was used to improve bond when early samples failed to produce plastic shrinkage cracks. The bonding agent consisted of cement and water mixed to the consistency of a thick paint and scrubbed onto the roughened base. The mix was then cast immediately over the bonding slurry. Further improvements to this method included using hot water to mix the cement slurry and pre-warming the base to accelerate bond development. The use of a shotcrete accelerator mixed into the bonding agent had limited success. Attempts at accelerating the bonding strength at the base/overlay interface were hoped to provide enough restraint in the sample to cause cracking. If the bond strength of the overlay exceeded the internal tensile strength of the sample cracking should occur. Eventually these attempts all proved to be futile. Under the severe drying effects of the chamber, enough water evaporated from the sub-base to desiccate the overlay/base interface and prevent good bonding. As seen in previous testing, no cracking would occur if  37  the sample was not demolded, but no bond would be developed if the sample were demolded. Finally it was a change in sample geometry that proved successful. The placement of two wood blocks under the base elevated the base enough to provide a 50 mm space between the base and the form edge on all four sides, see Figure 3.3. After shooting, the base was covered with a 50 mm thick layer of shotcrete on five sides, only the bottom remained uncovered. This method is analogous to icing a cake, where the base is the cake and the shotcrete can be considered the icing. By creating the sample in this manner, restraint of the overlay is provided by both the bond to the roughened base and by physical restraint due to the perimeter restraint of the edges. The overlay is unable to curl because of the border restraint. The overlay, exposed to the severe drying environment is allowed to shrink, but prevents moisture from evaporating from the base.  Figure 3.3 - "Coated edge" test sample.  3.1.5 - Mix Design Development  Early mix designs consisted of sand/cement mortars with high water/cement ratios. Lack of any plastic shrinkage cracking may be attributed to other factors pertaining to the specimen geometry discussed previously. The high w/c ratios and the high slumps common with these early mixes, see Table 3.2, may have benefited their apparent crack resistance. After casting, these mixes tended to produce varying amounts of bleed water. The bleed water  38  acted like a curing compound; if the evaporation rate was less than the rate of production of bleed water no plastic shrinkage cracking would occur.  Table 3.2 - Initial Test Mix Material  kg/m  Cement  380  Water  220  Sand  1230  Coarse aggregate  420  3  The mix design listed is assuming the aggregate is in a saturated, surface dry state. The aggregate gradation conformed to an ACI No. 2. The cement used was a Type 10.  Heated water was used to produce a plastic mix at a temperature of 3 5 ° C . The mix was cast directly onto the roughened base that was pre-wetted to a saturated, surface dry state.  No plastic shrinkage cracks were formed during these tests, only minor surface crazing. Small cracks at the base of the samples were observed, indicating that the samples had curled. The curling of the samples removed the bond and thus the restraint required for cracking.  3.1.5.1 - Cement Content  One of the final developments of the initial test program was to increase the cement content of the mix. By increasing the cement content of the mix from 400 kg/m to 480 kg/m , the 3  amount of plastic shrinkage cracking was greatly increased.  39  3  All dry-mix shotcrete used in this study was produced by Target Products Ltd. The mix was an industry standard containing only sand, coarse aggregate and cement, bagged in 30kg paper bags. An additional amount of cement was added to the mix to promote plastic shrinkage cracking. Aggregates used were concrete sand and a pea gravel proportioned to comply with ACI gradation No. 2 gradation, see Table 2.1. In the final product Portland cement was added at a rate of 26% by mass of dry-materials (approximately 538kg/m  3  assuming a water/cement ratio of 0.40, see Table 3.3).  Table 3.3 - Standard Test Mix Material  kg/m  Cement  538  Water  215  Sand  1132  Coarse Aggregate  377  3  3.1.6 - Effect of Finishing  Finishing was found to have a significant effect on the degree of cracking on the sample. Initially, finishing of the samples was only intended to aid in crack detection and measurement.  However, as the test program advanced the finishing process became an  important test procedure. The effect of finishing on the degree of plastic shrinkage cracking was assessed with a test series of three identical mixes. All mixes were shot using the Brazilian method with bone-dry material. Finishing states included: a standard steel trowel finish, an overworked finish and a sample that was left in its "as-shot" state. A summary of the data is presented in Table 3.4.  40  Table 3.4 - Effect of Finishing on Plastic Shrinkage Cracking Restrained Plastic Shrinkage Test Results Mix ID  Type of Finish  Total Length (mm)  Area # of Cracks Crack (mm) 2  Max. Crack Width (mm)  5-9  As-Shot  440  2  404  1.00  5-16  Steel Trowel  2235  8  1257  1.10  5-7  Excessive Steel Trowel  775  Many  119  0.35  3.1.6.1 - As-Shot Finish  The unfinished sample, Mix No. 5-9, was left in its as-shot state. The surface was heavily "pock-marked" from the impact of coarse aggregate particles during shotcreting. Only two wide cracks formed in the sample, the numerous narrow cracks typical with other samples, did not appear. It may be that the presence of the narrow cracks is due to the trowelling action.  3.1.6.2 - Steel Trowel Finish  The steel trowel finish sample, Mix No. 5-16, was trimmed with a steel trowel, then finished with a wood float, followed by a quick pass with a steel trowel. The surface had a smooth texture. The large number of cracks observed in the sample was an ideal basis for the test program, furthermore the smooth surface makes crack detection and measurement easy.  3.1.6.3 - Excessive Steel Trowel Finish  An excessive steel trowel finish was produced by vigorously stroking the surface of the sample until the top 20 mm was in a semi-fluid state. The surface had a smooth texture, but  41  not the stiffness characteristic of a shotcrete. The sample appeared to more closely resemble that of cast concrete with a fluid consistency and the presence of bleed water.  Similar to cast concrete, this mix produced no almost no plastic shrinkage cracks.  3.1.6.4 Comparison of Finishing Techniques  As is clear in Figure 3.4, the sample with the steel trowel finish had the highest degree of plastic shrinkage cracking.  Effect of Finishing on Crack Area 1200  -.  1000  Unfinished  Steel Trowel  Over Finished  (average) Figure 3.4 - Effect of Finishing on Crack Area  The large amount of cracking noted in the sample with the steel trowel finish is likely due to the surface tearing of the plastic mix noted by Shaeles and Hover (1988).  42  It is interesting to note that when dry-mix shotcrete is finished in the field, trowelling is performed in a circular motion. This method would not increase the likelihood of large unidirectional plastic shrinkage cracking. Any tearing of the surface would be randomly oriented with the resulting cracks being random and narrow in width. All samples were finished immediately after shooting during the course of this work. For dry-mix, normal construction practice would leave the shotcrete in place for up to about one hour before finishing would commence. Normally, the finishing process would begin 15 minutes after shooting to avoid debonding the shotcrete from the substrate. This delayed finishing action would re-work the surface layer and disrupt crack formation. Panel surfaces were finished immediately to promote the formation of plastic shrinkage cracking.  3.2 - Development of the Final Test Procedure  The main purpose of this study was to develop a test procedure to evaluate restrained plastic shrinkage cracking in dry-mix shotcrete and to study the effects of fiber reinforcement in preventing the formation of cracks. Present testing methods do not reflect actual restraint conditions and were not felt to accurately represent real systems. The test procedure has been modeled on the standard ASTM format and is presented in Appendix C .  The testing apparatus was envisioned as an enclosed chamber where a sample could be subjected to differing temperature, humidity and wind velocities while observing and recording physical changes and cracking within the sample. Ideally, a separate room would have been used to fully contain the test. The "as built" apparatus occupied a floor area of about 9 m and was constructed at UBC. 2  The base of chamber is constructed from a two by four frame, sheathed with 19 mm plywood and measures 2 m by 2.4 m in plan at a height of 0.4 m, see Photo 6. Slots within the side of  43  the base allow a forklift to easily move the apparatus. A Plexiglas cover was constructed to house the sample, while still permitting viewing during testing. The Plexiglas cover had to be built with 19 mm thick material to prevent any sagging under the high temperatures within. Due to the heavy weight of the Plexiglas cover, a boat winch was set up on a steel tube frame to raise and lower the cover safely. Three electric heat fans provided the heat and wind source for the test. The electric heat fans are standard construction type heaters with variable heat levels. The fans are 1500 Watt and blow at a rate of 4250 Litres per minute each. A steel track was constructed to mount an optical microscope. The track allowed easy viewing of sample all over its top surface.  It was originally hoped that the in-situ strength of the overlay could be measured at the time of cracking. Attempts were made to cast companion samples, subject them to the same environmental conditions as the shrinkage sample then test them in compression during various stages of the test. No useful data were obtained due to difficulties in demolding the samples.  The testing was carried out in a controlled environment. A computer was used to monitor and control both the temperature and humidity of the chamber. Positive feedback was used to vary the heat output of the fans and maintain pre-set temperature and humidity levels. Wind speed over the samples was maintained at a constant rate.  Cracking was observed using an optical microscope with a micrometer, cracks as small as 0.01 mm could be seen with this technique. Close monitoring of the sample was impossible during testing as the removal of the Plexiglas cover would have affected the temperature and humidity levels near the sample and adversely affected the test.  44  The main purpose of the study was examining early age plastic shrinkage; no companion free shrinkage testing was carried out.  3.2.1 - Equipment  All dry-mix shotcrete was produced with the use of an Aliva 246 shotcrete pump. The shotcrete pump was fitted with a 3.6 liter, eight-pocket drum. All shotcreteing was carried out using a 50 mm inside diameter hose. A large 500-gallon air compressor located within the building provided air pressure for the pump. Air flow into the shotcrete pump was maintained at a constant flow rate of 300 cfm. Airflow was measured by a spring loaded, in-line air flowmeter, model O M E G A FL8945. The wet-mix portion of this study was performed with the use of an Aliva 262 shotcrete pump. The shotcrete pump was fitted with an 8 liter, eight-pocket drum.  All shotcreting was carried out in a semi-enclosed shooting chamber, see Photo 3.  3.3 - Comments on Test Procedure 3.3.1 - Panel Set-Up  It is standard test practice to place shotcrete panels in a position that will reflect site conditions. If the shotcrete is intended to be placed overhead, the test panel must be shot overhead; for wall repair, the panel must be placed on a vertical surface. Panel placement at an angle of 6 0 ° to the floor was chosen for several reasons. Firstly, the shrinkage sample is extremely heavy, weighing in at around 100 kg. Placement of such a sample on a vertical surface would have required special hangers and a hoist to lower the sample. Secondly, the mixes were shot to a very wet consistency to exaggerate plastic shrinkage cracking. The semi-fluid shotcrete would not have set in the forms easily.  45  3.3.2 - Panel Finishing  In practice shotcrete is seldom finished. Finishing shotcrete is highly labour intensive, time consuming and often difficult. Improper finishing techniques can often lead to additional problems involving bond and durability. However, in repair and retrofit applications, shotcrete is increasingly being specified in a finished form. A shotcrete finish is specified as either a rough or fine finish. The shotcrete surface must first be screeded to the appropriate level immediately after shooting. Steel sections or wire gauges act as finish lines with which a screed is run over. A screed is pushed upward in a zigzag motion to cut excess material away from the surface. After the shotcrete has gained a measure of strength the surface is trowelled smooth with wood floats being pushed in an upwards, circular motion. A rough finish involves only the wood float finish. If a fine finish is specified, a steel trowel is then used after the wood float to give the surface a smooth appearance.  The steel trowel finish in this study was used to aid in crack detection and to aggravate plastic shrinkage cracking.  3.4 - Standard Test Procedures 3.4.1 - Compressive Strength  Cored samples were taken from companion test panels, shot at the same time as the shrinkage sample. The shotcrete cores were tested in accordance with A S T M C-42. The shotcrete cylinders measured 89 mm in diameter and were stored in a saturated lime bath until testing. Cylinder strengths were corrected using A S T M C-42 if the samples did not have a Length/Diameter of 2. Three samples were tested for compressive strength at each of 7 and 28 days.  46  3.4.2 - Boiled Absorption and Permeable Voids  Cored shotcrete samples were tested in accordance with A S T M C-642. Two samples were tested for each mix. The results are presented in Appendix A.  3.5 - Non-Standard Test Procedures 3.5.1 - Determination of In-situ Water Content  Within 15 minutes of shooting two samples of fresh shotcrete were set aside for fresh property determination. A 1 kg sample was placed within a fry pan and heated over a Bunsen Burner for 10 minutes to remove any water from the shotcrete. Before and after masses were recorded. The difference between the before and after masses was assumed to be the moisture content of the fresh shotcrete sample. The shotcrete mix was initially batched in a bone-dry state.  3.5.2 - Determination of In-situ Cement and Aggregate Content  A 5 kg sample was used for a wash out analysis. The fresh shotcrete was placed in a bucket and agitated vigorously with copious amounts of water. Using a sieve, particles larger than the Number 100 sieve were retained in the bucket. This procedure continued until the wash water running through the sieve was considered clean. Through this procedure all particles smaller than the No. 100 sieve were washed out of the sample. The remaining aggregate was then placed within a 4 0 0 ° C oven for a period of at least 24 hours. The sample was then removed and allowed to cool. The sample was then sieved to classify the material as a sand or coarse aggregate. Sand was defined as any material passing the 2.43 mm sieve. The masses of the sand and coarse aggregate were then weighed and recorded. The 5 kg sample was taken from the middle of the core panel and assumed to be representative of the  47  entire mix. The washout test was calibrated with cast mixes to take into account the dust fraction present in the aggregate samples. Dust content correction factors were incorporated into the final analysis.  The w/c ratios listed in the tables account for the aggregate being in a saturated, surface dry state. The absorption of the sand was calculated to be 2.3%, and 1.7% for the coarse aggregate.  3.5.3 - Measurement of Crack Widths and Lengths  At present there exist no formal procedures for measuring or quantifying the extent of plastic shrinkage cracking in concrete. Concepts used in this study were similar to other current research within this area.  To simplify crack quantification, only the crack widths and lengths were measured. The cracks are labeled according to time of appearance, i.e. crack #1 is the first crack to form, two, the second, etc. A small magnifier with a built in vernier was used to measure the width of the crack to the nearest 0.05 mm. Cracks are generally not uniform in width and tend to vary by as much as +/- 0.20 mm over a distance of a few centimeters. To avoid the problem of over or under estimating the severity of cracking a more accurate approach to measuring the "crack area" was attempted. The entire crack length was observed, and then arbitrarily separated into sections of differing crack width. An estimate of the crack width over a distance of 50 mm was made and written beside the crack on the slab. Crack lengths were then measured to the nearest 5 mm. Micro cracking and cracks smaller than 0.05 mm, were not recorded as their presence was variable and would have been difficult to measure.  48  To compare samples, the concept of crack area was used. The crack area was defined as the length of the crack times the width of the crack. The average crack width is not a useful method for comparing slabs, because the presence of fine hairline cracks, which are sometimes present, would skew the results. In the future, the concept of crack density may be a better way to compare samples. The maximum width of cracks is also listed. The maximum width of a crack is defined as the maximum width occurring over a distance of at least 10 mm on the slab. The maximum width of a crack is useful for Building code criteria that may state maximum crack width tolerances.  3.6 Calibration of Test Procedure  Calibration of the test procedure was necessary to judge the variability and repeatability. Further testing involving different shotcrete equipment and water content of the material was an attempt to gain further understanding of the fundamentals of plastic shrinkage in dry-mix shotcrete.  3.6.1 Variables  All mixes in this study were prepared with the same materials. All mixes were shot using an Aliva 246 machine and pumped through 15 meters of 50 mm diameter hose with a standard Aliva dry-mix nozzle. Variables in this portion of the study consisted of:  •  •  Water Injection System •  Brazilian Water ring  •  Traditional nozzle  Material Water Content •  Bone-dry  •  Damp  49  •  Placement Technique •  Shot  •  Shot - Remixed - Cast  3.6.1.1 Water Injection System  Two types of water injection systems were compared in this test study, the Brazilian Water ring and a traditional dry-mix nozzle. The Brazilian water ring uses a standard dry-mix nozzle, but has a water ring located 3 m before the nozzle. The water ring forces a fine stream of pressurized water to saturate the mix. Turbulence within the last 3 m of the hose allows for mixing of the shotcrete.  The standard dry-mix nozzle injects low pressure water right at the nozzle.  3.6.1.2 Material Water Content  Two different material moisture contents were compared, bone dry and damp. The bone-dry material was batched and shot directly. The damp material was meant to model the effect of pre-moisturizing. To create the damp mixes, bone dry material was mixed thoroughly with a small amount of water in a drum mixer. All "damp" shotcrete mixes had a water content of  2%.  3.6.1.3 Placement Technique  To gauge the effects of the shooting process in plastic shrinkage cracking, both shot and cast samples were compared. To produce cast concrete samples with exactly the same material components as shotcrete, mixes were first shot, then re-mixed. Immediately after  50  shooting, shotcrete panels were emptied into a pan mixer, re-mixed for five minutes, then recast into test panels. Samples were consolidated with a vibrating table.  3.6.2 Results  A total of four series of test mixes were shot using the Brazilian water ring, three sets of test mixes were produced using the traditional nozzle. Each series consisted of three or four identical mixes to obtain a measure of the variation inherent in the testing procedure. A summary of the variables and corresponding mix identification numbers is presented in Table 3.5.  Table 3.5 Variables Considered in Calibration Study Mix ID 8-15 8-14 8-13 8-12 8-10 8-9 8-8 8-7  Nozzle Type  Material State  Placement Technique  Traditional  Damp  Shot  Traditional  Dry  Shot-remixed-cast  Traditional  Dry  Shot  Brazilian  Damp  Shot-remixed-cast  Brazilian  Damp  Shot  Brazilian  Dry  Shot-remixed-cast  Brazilian  Dry  Shot  8-6 7-10 7-9 7-8 7-4 7-3 7-2 6-27 6-25 6-24 6-21 6-20 6-19 6-18 6-17  51  A summary of the in-situ coarse aggregate, sand, cement and water content is presented in Table 3.6. In-situ cement, aggregate and water contents were determined as described in Section 3.5.1 and 3.5.2. In-situ material contents differed from the as-batched material contents largely due to the amount of rebound during shooting. Rebound of the coarse aggregate during shooting resulted in higher cement and sand contents and generally lower coarse aggregate contents.  Table 3.6 Summary of In-situ Cement and Aggregate Contents in Calibration Study In-Situ Material (kg/m3) Mix ID  Coarse Aggregate  Sand  Cement  Water  w/c  8-14  474  1062  529  190  0.36  8-14  450  1093  517  194  0.38  8-13  450  1094  497  213  0.43  8-12  348  1264  440  189  0.43  8-10  316  1294  447  193  0.43  8-9  336  1269  482  200  0.41  8-8  306  1295  458  194  0.42  8-7  341  1260  455  199  0.44  8-6  328  1273  459  194  0.42  7-10  187  1346  522  200  0.38  7-9  413  1177  467  199  0.43  7-8  439  1133  498  185  0.37  7-4  432  1198  444  180  0.41  7-3  418  1097  542  197  0.36  7-2  437  1151  473  194  0.41  6-27  443  1157  465  189  0.41  6-25  343  1250  452  220  0.49  6-24  351  1229  480  206  0.43  6-21  321  1267  468  209  0.45  6-20  379  1247  446  194  0.43  6-19  225  1210  548  280  0.51  6-18  291  1304  466  204  0.44  6-17  215  1294  516  240  0.47  52  It is clear in Table 3.6 that there is considerable variation in the in-situ material contents. Differences in in-situ material content may be due to several factors. It was noted early in the study that the coarse aggregate content of the mix would decrease the higher the sample was taken in the test panel. During the shooting process paste would tend to accumulate on the lower portion of the test panel. The higher paste content on the lower half of the test panel would decrease the likelihood of the coarse aggregate from rebounding from the panel. The result of this would be a stiffer shotcrete with a lower coarse aggregate content placed on the top half of the panel. In an attempt to obtain less variability in the results and a representative sample, fresh shotcrete for the material content analysis was always taken from the middle portion of a test panel.  Table 3.7 presents the results of the plastic shrinkage cracking analysis. Analysis was performed in accordance with section  3.5.3.  53  Table 3.7 Summary of Plastic Shrinkage Cracking Results of Calibration Study  Mix ID  Nozzle Type  Material State  Crack Data Placement  8-15 8-14  Traditional  Damp  Shot  8-13 8-12 8-10  Traditional  Dry  8-9  Shot+remix +cast  8-8 8-7  Traditional  Dry  Shot  8-6 7-10 7-9  Brazilian  Damp  7-8  Shot+remix +cast  7-4 7-3 7-2  Brazilian  Damp  Shot  6-27 6-25 6-24  Brazilian  Dry  6-21  Shot+remix +cast  6-20 6-19 6-18  Brazilian  Dry  Shot  6-17  Total Length (mm)  #of Cracks  Crack Area (mm)  Max. Width (mm)  410  3  542  0.45  405  3  549  0.60  315  4  691  0.80  380  2  121  0.15  320  2  168  0.35  2  440  3  158  0.20  1795  5  752  0.80  1856  7  797  0.50  1825  6  586  0.95  285  2  198  0.30  390  3  150  0.15  370  2  166  0.20  1355  4  510  0.60  1105  5  725  0.80  1195  4  650  0.60  1480  4  475  0.50  990  2  179  0.25  1270  3  151  0.20  505  2  144  0.15  1665  5  925  1.00  1885  7  1094  1.20  1565  5  915  0.80  1780  6  969  1.15  3.6.2.1 Water Injection System  When using bone dry material the Brazilian water ring produces shotcrete with an apparently wet consistency. The water ring injects water into the mix and mixes with the bone dry material for a fraction of a second before being "shot" onto the receiving surface. During the shooting process all water is free to lubricate the mix and aid in compaction and  54  consolidation of the shotcrete. After the material is placed a portion of the water is absorbed into the aggregate fraction of the mix resulting in the apparent increase of stiffness.  The use of bone dry materials with the traditional nozzle resulted in a great deal of dust produced at the nozzle. Even with the high water contents used in this study, a few sand lenses were formed with this process, the presence of which may have affected the results.  When using a damp material the Brazilian water ring and the traditional nozzle produce a very similar material. With the damp material, the Brazilian water ring produces a slightly more consistent shotcrete with less chances of creating sand lenses.  3.6.2.2 Material Water Content  The use of pre-moistened material in the shotcrete process greatly reduced the amount of dust generated, both at nozzle and pump with both the traditional nozzle and Brazilian water ring.  The use of bone dry material with the Brazilian water ring produces much less dust at the nozzle when compared with the traditional nozzle. The three meters of travel through the hose provides enough mixing to bind most cement and sand with some water to produce a consistent shotcrete material. There is some loss of paste from the mix in the form of "nozzle slobber".  55  3.6.3 Discussion of Results  The different shotcrete equipment and material states had an effect on the plastic shrinkage capacity of the samples. Figure 3.5 shows the difference in crack areas between the four shotcrete systems.  Comparison of Crack Area  Shotcrete Samples 1200  Traditional Damp  Traditional Dry  Brazilian Damp  Brazilian Dry  Figure 3.5 - Comparison of crack areas with different shotcrete methods.  It is clear that the Brazilian water ring with bone-dry material produces the largest degree of plastic shrinkage cracking. The use of a pre-dampened material apparently reduces the degree of plastic shrinkage in shotcrete samples. There is little difference between the degree of plastic shrinkage cracking in dry-mix shotcrete produced by the traditional nozzle and the Brazilian water ring.  56  Figure 3.6 compares the maximum crack width of plastic shrinkage cracks produced by the various methods.  Comparison of Maximum Crack Width  Shotcrete Samples  Traditional Traditional Damp Dry  Brazilian Damp  Brazilian Dry-  Figure 3.6 - Comparison of maximum crack widths produced with different shotcrete methods.  The trends, although less definite than the crack areas, are similar. There appears to be little difference between the traditional nozzle and the Brazilian water ring when a dampened material is used. The Brazilian water ring with bone-dry material produces the largest crack widths when compared to the other methods.  3.6.3.1 Placement Technique  Figure 3.7 compares the difference between the crack areas of cast concrete and shotcrete produced with bone dry material with a traditional nozzle. The cast concrete was produced by; first shooting a test mix, taking the shot material, remixing it, and then casting it onto a roughened base.  57  Comparison of Crack Areas Traditional Dry Method 900  Shot+remixed+Cast  Shot  Figure 3.7 - Comparison of crack areas between cast and shot samples produced using the traditional dry method.  Figure 3.8 compares the difference between the crack areas of shotcrete and cast concrete produced with damp material with a Brazilian water ring.  58  Comparison of Crack Areas Brazillian Damp Method 800  Shot+remixed+Cast  Shot  Figure 3.8 - Comparison of crack areas between cast and shot samples produced using the Brazilian damp method.  Figure 3.9 compares the difference between the crack areas of shotcrete and cast concrete produced with bone dry material with a Brazilian water ring.  59  Comparison of Crack Areas Brazillian Dry Method 1200 1100 1000 900 800 700 ro 600 <i> 500 400 u ro i_ 300 O 200 100 0 Shot+remixed+Cast  Shot  Figure 3.9 - Comparison of crack area between cast and shot samples produced with the Brazilian dry method.  The obvious differences between the degree of plastic shrinkage in the cast and shot samples is due to several factors. Presumably, the shotcreting process creates a much stronger bond between the overlay and substrate. The cast specimens were noted to have localized sections of de-bonded concrete, which likely reduced the amount of plastic shrinkage cracking. The cast mixes had a much higher apparent workability than the shotcrete samples even though both mixes were identical. The compaction caused during the shooting process gives the shotcrete an inherent stiffness. After the sample was reworked the mixes became much more fluid. This is possibly due to the entrainment of air during the mixing and placing operations. Additional air would have the effect of increasing the apparent workability of the mix. Cast samples had more of a tendency to bleed than shot samples. It is believed that during the re-mixing and placing process bleed channels are formed throughout the fresh sample that would not have formed in the shotcrete sample. The presence of bleed water would reduce the amount of plastic shrinkage cracking.  60  It is interesting to note the crack areas of the cast samples are very similar to the "Over Finished" shotcrete sample. Table 3.8 compares the plastic shrinkage crack data of the "Over Finished" shotcrete sample with the shot-remixed-cast samples. All samples were produced using the Brazilian Method and bone-dry material. The "Over Finished" sample was shot and finished vigorously with a steel trowel. The cast samples were produced by first shooting the mix, re-mixing it and then casting it into the mold.  Table 3.8 Comparison of Over Finishing with Cast Samples Crack Data Mix ID  Placement Method  6-25  6-24  Shot-remixed-cast  6-21  5-7  Shot, then finished excessively  Total Length (mm)  #of Cracks  Crack Area (mm)  Max. Width (mm)  990  2  179  0.25  1270  3  151  0.20  505  2  144  0.15  775  Multiple  119  0.35  2  3.6.3.2 Paste Content  Although all mixes were batched and shot identically, variation in the amount of rebound can affect the paste content. The paste content is defined as the mass of "cement + water + air" per cubic meter of shotcrete. As no measurement of the air content of the mixes was made, the effect of air was not considered.  61  Effect of Paste Content on Crack Area Shotcrete Samples  E  + Traditional Damp • Traditional Dry  Q>  A Brazilian Damp  o re o  x  500  600  700  800  Brazilian Dry  900  Paste Content (kg/m ) 3  Figure 3.10 - Effect of paste content on crack area.  As can be seen in Figure 3.10, the paste content of the mix appears to have some influence on the degree of plastic shrinkage cracking only for the Brazilian Dry method. The degree of plastic shrinkage cracking may be more related to the rebound of the coarse aggregate and the total amount of water in the mix.  62  Chapter 4 - Evaluation of Variables Affecting Plastic Shrinkage  4.1 Summary of Variables  The final test program involved the evaluation of five different fiber types and two different admixtures that purported to have shrinkage reducing capabilities. The fiber types included three steel types and two different lengths of a straight Polyolefin fiber. The admixtures included a drying shrinkage controlling admixture and a waterproofing admixture. Table 4.1 presents a summary of the variables considered for the final test program and the respective mix identification numbers. All mixes, with the exception of 1-15 and 2-12, were shot with the dry-mix process using the Brazilian method and bone dry material. The base shotcrete mix is presented in Table 3.3 in Chapter 3. Fiber dosage listed in Table 4.1 is listed as the as-batched dosage rate.  63  Table 4.1 Summary of Variables in the Final Test Program Mix ID  Variable*  4-3  Batched Dosage Rate** -  Control  5-16  -  5-27  9 kg/m  12-9  10 kg/m  3  6-4  18 kg/m  3  20 kg/m  3  5-23  27 kg/m  3  10-18  40 kg/m  3  10-19  60 kg/m  3  12-11  10 kg/m  3  20 kg/m  3  10-7  40 kg/m  3  12-3  20 kg/m  3  40 kg/m  3  60 kg/m  3  Hooked End Fiber  10-16  12-13  Flattened End Fiber  10-24  Crimped Fiber  10-25 5-28 5-29  Straight 50 mm Polyolefin Fiber  6-5 6-3  12-14  1.50% 3.00% 0.25%  Straight 25 mm Polyolefin Fiber  5-31 12-17  3  0.50% 1.00%  Shrinkage Controlling Admixture  1% 2%  1-15  Wet Mix Control  -  12-12  Waterproofing Admixture (Wet Mix)  1.50%  * All mixes, with the exception of mixes 1-15 and 12-12, were shot using dry mix shotcrete with the Brazilian method and bone-dry material. ** For steel fibers the value is the mass fraction in kg/m . For Polyolefin fibers the value is the volume fraction expressed as a percentage. For the admixtures, the value is the batched dosage expressed as a percentage by weight of cement. 3  4.1.1 - Presentation of Data  Fiber contents are listed as both the as-batched and in-situ volumes. In-situ fiber contents are presented as volume fractions and mass fractions and are based on densities of  64  7850 kg/m for steel and a density of 910 kg/m for Polyolefin. All shotcrete used in this 3  3  study was shot at an extremely wet consistency. The observed fiber rebound rates were extremely low compared to typical practice. For this reason no fiber rebound values have been listed as they present unrealistically low rebound rates.  Dosages of the shrinkage controlling admixture are listed as a percentage by mass of cement in the batched shotcrete. The amount of the admixture in the shotcrete after shooting could not be assessed, but it is assumed that the amount lost would be low.  The dosage of the waterproofing admixture listed is a percentage by mass of cement in the batched shotcrete. The amount of the waterproofing admixture in the shotcrete after shooting could not be assessed. However, assuming the admixture was intimately mixed within the paste, the amount of admixture per mass of cement would remain the same before and after shooting.  Table 4.2 presents the results of the in-situ material analysis. The in-situ aggregate, cement and water contents were measured using the procedures outlined in Section 3.5.1 and 3.5.2. The water content listed in Table 4.2 does not include the water absorbed by the aggregate, all aggregate is assumed to be in a saturated, surface dry condition. The as-batched drymix shotcrete material listed in the first row of Table 4.2 is a theoretical amount and is based on an assumed water/cement ratio of 0.40. The aggregate is assumed to be in a saturated, surface dry condition; water absorption of the aggregate is accounted for in these values.  65  Table 4.2 In-situ Material Content Analysis of Final Test Program  Mix ID  Variable  Variable Dosage  In-situ Material Content (kg/m) 3  Coarse Aggregate  Sand Cement Water Fiber  377  1132  538  (215)  -  (0.40)  -  462  1161  437  191  -  0.44  -  558  923  512  257  -  0.50  5-27  9  202  1254  558  245  5.8  0.44  12-9  10  245  1205  564  234  3.0  0.41  6-4  18  354  1255  437  212  13.0  0.49  20  391  1176  464  204  17.0  0.44  5-23  27  209  1247  577  211  21.2  0.37  10-18  40  466  1166  411  187  23.0  0.45  10-19  60  408  1140  464  198  33.0  0.43  10  338  1218  474  219  8.0  0.46  20  309  1254  474  205  16.5  0.43  10-7  40  337  1207  478  199  32.0  0.42  12-3  20  245  1205  564  234  8.0  0.41  40  362  1238  435  201  16.0  0.46  60  356  1242  420  196  38.0  0.47  1.50%  375  1205  475  202  7.8  0.43  3.00%  317  1277  449  211  12.0  0.47  0.25%  294  1255  502  214  1.2  0.43  0.50%  356  1230  465  213  2.0  0.46  1.00%  324  1176  532  227  4.6  0.43  As-batched Dry-Mix 4-3 5-16  10-16  Control (Brazilian Dry)  Hooked End Fiber  12-11 12-13  10-24  Flattened End Fiber  Crimped Fiber  10-25 5-28 5-29  50 mm Polyolefin  6-5 6-3  25 mm Polyolefin  5-31  1%  323  1237  488  204  -  0.42  12-14  Shrinkage Controlling  2%  271  1210  552  219  -  0.40  1-15  Wet Mix Control  -  342  1028  485  195  -  0.40  12-12  Waterproofing  1.50%  340  1030  490  198  -  0.40  12-17  The difference in the as-batched material content and the in-situ material content reflects the variability of shotcrete, even under laboratory conditions. Differences in the values could include such factors as: •  w/c  Rebound - Rebound during shooting resulted in a loss of some of the coarse aggregate.  66  •  "Nozzle slobber" - the term used for paste dripping from the end of the nozzle discussed in Section 2.3.3.4, would result in a loss of paste during shooting.  •  w/c - The actual w/c ratio listed in Table 4.3 is generally higher than the 0.40 used for the material content calculation of the as-batched dry-mix.  •  Sampling errors  4.2 - Fiber Analysis  The relationship between crack area and fiber volume is clear for all fiber types tested. For an increasing fiber content the amount of cracking is dramatically reduced. Among the fibers tested the Hooked End steel fiber appears to be the most effective fiber at controlling plastic shrinkage cracking. The Flattened End fiber is the second most effective, followed closely by the straight 25 mm Polyolefin fiber. The Crimped fiber, although not marketed as a shotcrete fiber, was used in this study to compare fiber shape efficiency. The straight 50 mm Polyolefin fiber was found to be the most inefficient fiber type tested.  It should be noted that all the presented results are for only one particular set of environmental conditions, and although relatively severe, it only represents an evaporation rate of 1.5 kg/m /hour. More extreme environmental conditions will produce more extensive 2  cracking. The results of this study should not be viewed as an absolute analysis of plastic shrinkage cracking, but more for the trends they exhibit.  67  4.2.1 - Hooked End Steel Fiber  The Hooked End fiber was shot at seven different volume fractions. The results of the plastic shrinkage crack analysis are summarized in Table 4.3.  Table 4.3 - Summary of Plastic Shrinkage Crack Analysis of the Hooked End Fiber Batched Fiber Dosage (kg/m)  In-situ Fiber Mass (kg/m)  In-situ Volume Fraction  5-27  9  5.5  12-9  10  6-4  Mix ID  Fiber Type  Plastic Shrinkage Crack Data Total Length (mm)  #of Cracks  Crack Area (mm)  Max Width (mm)  0.07%  1525  6  453  1.10  7.9  0.10%  1280  4  406  1.00  18  13.3  0.17%  905  6  130  0.50  20  17.3  0.22%  260  4  75  0.25  5-23  27  21.2  0.27%  240  2  30  0.15  10-18  40  9.4  0.12%  920  3  260  0.70  10-19  60  33.0  0.42%  230  2  11.5  0.05  3  10-16  Hooked End Fiber  3  2  Figure 4.1 illustrates the relationship between crack area and fiber volume. The fiber volume shown is the as batched volume and does not take into account fiber rebound. It is clear from Figure 4.1, that even a small addition of fibers drastically reduces the amount of plastic shrinkage cracking.  68  Hooked End Fiber Crack Area Vs Fiber Volume  1400  0.00%  0.50%  1.00%  1.50%  Fiber Volume Figure 4.1 - Hooked End Fiber; crack area vs. fiber volume.  4.2.2 - Flattened End Steel Fiber  The Flattened End fiber was shot at three different volume fractions. The results of the plastic shrinkage crack analysis are summarized in Table 4.4.  Table 4.4 - Summary of Plastic Shrinkage Crack Analysis of the Flattened End Fiber  Mix ID  Fiber Type  Batched Fiber Dosage (kg/m)  In-situ Fiber Mass (kg/m)  In-situ Volume Fraction  10  7.9  20 40  3  12-11  12-13 10-7  Flattened End  Plastic Shrinkage Crack Data Total Length (mm)  #of Cracks  Crack Area (mm)  Max Width (mm)  0.10%  1120  3  411  1.15  16.5  0.21%  850  2  205  0.50  32.2  0.41%  685  2  118  0.25  3  69  2  Figure 4.2 illustrates the relationship between crack area and fiber volume.  Flattened End Crack Area V s Fiber Volume  1400 - j 1200 l\ 1E E 1000 - \ 800 J \ < 600 400 - • o 03 200 0 -I— 0.00%  •  0.50%  1.00%  1.50%  Fiber Volume Figure 4.2 - Flattened End Fiber; crack area vs. fiber volume.  4.2.3 - Crimped Steel Fiber with Crescent Section  The Crimped fiber was shot at three different volume fractions. The Crimped fiber is not the ideal shotcrete fiber; it has a larger ratio of surface area to volume than the other fibers tested and its shape does not appear to be as aerodynamic as the other fibers tested, which may have resulted in a large rebound factor. Furthermore, it is a heavy fiber, reducing its insitu fiber count. The results of the plastic shrinkage crack analysis are presented in Table 4.5.  70  Table 4.5 - Results of Plastic Shrinkage Crack Analysis of the Crimped Fiber  Mix ID  Fiber Type  Batched Fiber Dosage (kg/m)  In-situ Fiber Mass (kg/m)  In-situ Volume Fraction  20  11.0  40  60  3  12-3  10-24  Crimped Fiber  10-25  Plastic Shrinkage Crack Data Total Length (mm)  #of Cracks  Crack Area (mm)  Max. Width (mm)  0.14%  1755  5  574  1.00  15.7  0.20%  1360  4  391  0.80  37.7  0.48%  380  2  135  0.35  3  Figure 4.3 illustrates the relationship between crack area and fiber volume.  Crimped Fiber with Crescent Section Crack Area Vs Fiber Volume  CO CD  o  CD — i  O  0.00%  0.50%  1.00% Fiber Volume  Figure 4.3 - Crimped Fiber; crack area vs. fiber volume.  71  1.50%  2  4.2.4 - Straight 25 mm Polyolefin Fiber  The Straight 25 mm Polyolefin fiber was shot at three different volume fractions. The fiber measures 25 mm in length and 0.38 mm in width. The fibers are supplied in taped bundles that measure 75 mm in diameter. The fiber bundles are referred to as "pucks" because of the resemblance to a hockey puck.  Although shot using the dry-mix process in this study, it is not recommended for use in drymix generally. The shotcrete produced in this study had a high cement content and was shot at a very wet consistency. Shotcrete applied in most situations would have a lower cement content and would be shot at a much drier consistency. This would have increased the rebound factor dramatically. Synthetic fibers, being of a much lower density than the shotcrete materials, tend to "fly away" during the shooting process. The use of synthetic fibers in wet-mix is common and does not suffer from the detrimental effects of large rebound.  The Polyolefin fibers finished with greater ease when compared to the steel fibers and were relatively easy to dispense into the mix when added at dosages below 4 kg/m (0.44% by 3  volume). The results of the plastic shrinkage crack analysis are presented in Table 4.6.  72  Table 4.6 - Results of Plastic Shrinkage Crack Analysis of the 25 mm Polyolefin Fiber  Mix ID  Fiber Type  6-5 25 mm Polyolefin  6-3  5-31  Batched Fiber Dosage  In-situ Volume Fraction  Plastic Shrinkage Crack Data Total Length (mm)  #of Cracks  Crack Area (mm)  Max Width (mm)  2  0.25%  0.13%  1550  5  502  0.85  0.50%  0.22%  780  4  207  0.95  1.00%  0.51%  700  3  72  0.30  igure 4.4 illustrates the relationship between crack area and fiber volume for the 25 mm Polyolefin fiber.  25 mm Polyolefin Fiber Crack Area V s Fiber Volume  1400  1= ro o ro L-  o 0.00%  0.50%  1.00%  1.50%  Fiber Volume Figure 4.4 - Straight 25 mm Polyolefin Fiber; crack area vs. fiber volume.  73  4.2.5 - Straight 50 mm Polyolefin Fiber  The Straight 50 mm Polyolefin fiber was shot at two different volume fractions. The Straight 50 mm Polyolefin fiber is similar to the Straight 25 mm Polyolefin fiber, but measures 50 mm in length and 0.63 mm in diameter.  This fiber was originally chosen because its mass was similar to that of a single Hooked End fiber. Its use would illustrate the effects of fiber density within the shotcrete matrix. Although this fiber is 50 mm in length it was shot through a 50 mm diameter hose. The pump was quickly plugged during this process. Table 4.7 summarizes the results of the plastic shrinkage test data.  Table 4.7 - Results of Plastic Shrinkage Crack Analysis of the 50 mm Polyolefin Fiber  Mix ID Fiber Type  Batched Fiber Dosage  In-situ Volume Fraction  Plastic Shrinkage Crack Data Total Length (mm)  5-28  5-29  #of Cracks  Crack Area (mm)  Max. Width (mm)  2  1.50%  0.86%  615  3  95  0.30  3.00%  1.32%  155  1  16  0.20  50 mm Polyolefin  Figure 4.5 illustrates the relationship between the crack area and in-situ fiber volume of the straight 50 mm Polyolefin fiber.  74  50 mm Polyolefin Fiber Crack Area V s Fiber Volume  1400 -. c<r 1200 i\ £ 1000 - \ nT 800 2> < 600 o 400 CD  6  ,  X .  200 -  o]  0.00%  .  4  0.50%  -;  1.00%  * J  1.50%  Fiber Volume Figure 4.5 - Straight 50 mm Polyolefin Fiber; crack area vs. fiber volume.  4.2.6 - Comparison of Fibers  To compare different fiber types the crack area versus fiber volume is illustrated in Figure  75  Comparison of Fibers Crack Area Vs Fiber Volume  50 mm Ftilyolefin  E E,  . . . . Crimped 25 mm Fblyolefin  ra  . _ Flattened End  o n  _ _ _ _ _ Hooked End  u O 0.00% 0.20% 0.40% 0.60% 0.80% 1.00% 1.20% 1.40% 1.60% Fiber Volume  Figure 4.6 - Comparison of Fibers; crack area vs. fiber volume.  As seen in Figure 4.6 there appears to be a clear difference in fiber performance. For a given fiber volume the Hooked End fiber outperforms other fibers significantly.  A comparison of fiber performance based on mass fraction, instead of volume fraction results in some drastic differences. Steel fibers, at a density of 7850 kg/m , are far out3  performed on a per mass basis when compared to synthetic fibers at a density of 910 kg/m . 3  As is clear in Figure 4.7. the polyolefin fibers have a much improved efficiency when compared to the steel fibers on a fiber mass basis. The fiber mass listed is based on in-situ fiber volume fraction and already accounts for rebound.  76  Comparison of Fibers Crack Area Vs Fiber Mass 1400 ,  0  5  10  15  20  25  30  35  40  45  50  55  60  Fiber Mass (kg)  Figure 4.7 - Comparison of Fibers; crack area vs. In-situ mass fraction.  Max Crack Width vs Fiber Volume  50 mm Polyolefin . . Crimped . . Rattened B i d Hooked End 25 mm RDlyolefin  Fiber Volume  Figure 4.8 - Comparison of Fibers; maximum crack width vs. In-situ volume fraction.  77  The correlation of maximum crack width with fiber volume is clear in Figure 4.8, but not as well defined as the relationship between total crack area and fiber volume. Reasons for this may in part be due to the more subjective measurement of a single crack width. The maximum crack width defines the slab by one measurement; the crack area value is composed of the sum of numerous cracks' length times their width. The larger number of measurements necessary to obtain the crack area value averages the results. A single poor measurement would not skew the results drastically in the determination of the crack area as could happen with the maximum crack width.  It can be seen that the use of even a very small amount of fibers will drastically reduce crack widths.  4.2.6.1 Fiber Geometry  The effect of fiber geometry on plastic shrinkage cracking was examined by comparing the crack area with the number of fibers present within the mix.  78  Crack Area Vs Number of Fibers  0.00  500,000.00  1,000,000.00  1,500,000.00  2,000,000.00  Number of Fibers  Figure 4.9 - Comparison of Fibers; crack area vs. number of fibers.  A small sample of each fiber type was counted and recorded. The corresponding number of fibers per unit mass is shown in Table 4.8. The crimped fiber, with a much smaller length/diameter (l/d) value, has a much smaller fiber count than the Hooked End and Flattened End fibers with equivalent lengths. The 25 mm Polyolefin fiber, with a higher l/d ratio and a density less than 8 times that of steel, has a very high fiber count when compared to the other fibers. The 50 mm Polyolefin fiber, with its long length and large diameter has a fiber count comparable to the steel fibers.  79  Table 4.8 - Unit Quantity of Fibers. Fiber Type  # of fibers per 1.000 kg  Hooked End  22 400  Flattened End  10 500  Crimped  6 800  25 mm Polyolefin  364 000  50 mm Polyolefin  40 400  For the steel fibers evaluated the number of fibers within the matrix is the controlling factor. It is clear in Figure 4.9 that when the number of fibers contained within each mix are compared, very little difference is evident between the differing steel fiber types. This observation may indicate that fiber shape is not a factor in plastic shrinkage cracking and that the mass per surface area of fiber is the most important factor. However, as no straight, undeformed steel fibers were evaluated, Figure 4.9 may indicate that only a minimum degree of deformation is required to achieve enough bond to control cracking in the plastic state. All of the steel fibers had some degree of deformation.  The stresses and strains produced by plastic shrinkage in plastic shotcrete are very low. As stresses within the fiber are very low, fiber strength is not a factor, however the effect of fiber stiffness may be a factor. Balaguru (1994) found that a lower modulus of elasticity in the fiber provided better crack control than fibers with a higher modulus. The beneficial effect of the lower modulus in the Polyolefin fibers is not readily apparent in Figure 4.9. The 25 mm Polyolefin fiber performed very poorly compared to the other fibers in terms of fiber count.  80  The 50 mm Polyolefin fibers' performance is comparable to the steel fibers. An explanation of the relatively poor performance of the 25 mm Polyolefin fiber may be its shape.  The relatively poor performance of the straight 25 mm Polyolefin fiber on a per fiber basis may suggest that some degree of fiber deformation may be required for fibers to perform efficiently as crack control even in plastic shrinkage. Much research has been carried out on the effect of fiber shape on flexural and tensile strength of concrete. It is generally agreed that fiber shape has a large effect on crack control in concrete. Kovler et al (1992) found that a rougher surface texture on the fibers was found to reduce cracking in drying shrinkage tests. This would suggest that fiber shape is an important factor in drying shrinkage crack resistance and possibly in plastic shrinkage.  Reasons for the relatively good performance of the 50 mm Polyolefin fiber are puzzling. The 50 mm Polyolefin fiber may have had enough length to provide sufficient bond to offset its lack of shape, this fiber was almost twice as long as all other fibers evaluated. The lower modulus of elasticity in the Polyolefin fibers may have somewhat offset there lack of deformation.  4.2.6.2 Fiber Rebound  To incorporate the effects of fiber rebound into the plastic shrinkage resistance; Figure 4.10 shows the relationship between the crack area and the as-batched fiber mass. The asbatched fiber volume is the volume fraction of fibers added to the mix prior to the shooting process.  81  Fiber Efficiency Crack Area Vs Batched Fiber Volume  -  r  0.00%  1  1  1.00%  2.00%  ~  — T  3.00%  Batched Fiber Volume  Figure 4.10. - Fiber Efficiency; crack area vs. batched fiber volume.  The relatively poor efficiency of the Crimped fiber is attributed partially to its high rebound factor.  4.2.6.3 Optimum Fiber Design for Plastic Shrinkage Control  Considering the results of this study, the optimum fiber for control of plastic shrinkage cracking in shotcrete would have the following criteria:  Fiber for Dry-Mix Shotcrete •  A high l/d ratio with minimum length of the size of the coarse aggregate  •  Steel  •  Some form of deformation on the fiber  82  A steel fiber is recommended for dry-mix shotcrete to minimize fiber rebound.  Fiber for Wet-Mix Shotcrete •  A high l/d ratio with a minimum length of the size of the coarse aggregate  •  Low modulus of elasticity  •  Some form of deformation on the fiber  Some form of Poly fiber is recommended for wet-mix shotcrete where fiber rebound is not as much of an issue, compared to dry-mix.  4.3 Shrinkage Controlling Admixture  A shrinkage controlling admixture meant to reduce long term drying shrinkage was evaluated. The admixture is packaged in a liquid form, making it somewhat more conducive for use in wet-mix shotcrete. The admixture was used in this study in dry-mix shotcrete for comparative purposes.  The shrinkage controlling admixture is alcohol based, having a specific gravity of about 0.95. To dose the admixture into dry-mix shotcrete it was added into the water line of the nozzle during shooting. Knowing the rate of water addition to the dry-mix shotcrete through the pump, a proportioned amount of the admixture was added to the mixing water. Due to the low specific gravity of the admixture, constant mixing was necessary to avoid separation.  The shrinkage controlling admixture provided cohesiveness to the mix during shooting, aiding in the ability of the shotcrete to be built-up. As the mix was more plastic, the panels were easier to finish than a conventional shotcrete.  83  A major concern regarding the use of this admixture is the presence of alcohol. During shooting, there was a strong odor of alcohol in the air. Through the shooting process, the fine spray of the material out of the nozzle, promoted the evaporation of the volatile component of the admixture. There would be risks in using this admixture in confined spaces, from combustion and health hazards.  The shrinkage controlling admixture was used at two dosage factors, 1% and 2% by mass of cement. The admixture was found to substantially reduce the degree of plastic shrinkage cracking in the samples, see Figure 4.11. The results of the plastic shrinkage crack analysis are presented in Table 4.9.  Table 4.9 - Results of Plastic Shrinkage Crack Analysis of the Shrinkage Controlling Admixture  Mix ID Admixture  12-17 12-14  Shrinkage Controlling  Batched Admixture Dosage  Plastic Shrinkage Crack Data Total Length (mm)  #of Cracks  1%  975  3  276  0.7  2%  755  3  91  0.3  84  Crack Area Max Width (mm) (mm) 2  Shrinkage Controlling Admixture Crack Area V s Admixture Dosage  _ 1400 -r*!= 1200 ^ E. 1000 S < •g  600 400 -  8  0  0  w  -i  1  1  j  1  0.00%  0.50%  1.00%  1.50%  2.00%  Admixture Dosage  Figure 4.11 - Shrinkage Controlling Admixture; crack area vs. admixture dosage.  4.4 Wet-Mix Study  Wet-mix shotcrete was used briefly in this study for the evaluation of an admixture that could only be used in wet-mix shotcrete. The wet-mix shotcrete was the same material used in the bulk of this study. The same material was batched as the dry-mix, but with the exception that water was added prior to adding the material to the pump.  The plastic shrinkage cracking characteristics of the plain wet-mix shotcrete were similar to the pre-dampened dry-mix shotcrete results.  85  4.4.1 Waterproofing Admixture  A cementitious based waterproofing admixture was evaluated for its benefits of controlling plastic shrinkage. The manufacturer also purports that the admixture "Inhibits shrinkage and cracking" in concrete. The admixture is typically used in wet-mix shotcrete.  4.4.2 Results of Waterproofing Admixture Study  The admixture appears to act as a water reducer and also entrains a small amount of air, thus increasing the workability of fresh concrete. The mix was pumped and shot easily due to the increased workability. Finishing the mix was also easier due to the "fattiness" of shotcrete produced by the increased air content.  The waterproofing admixture was added to a wet-mix shotcrete at a dosage rate of 1.5% by mass of cement. The results of the plastic properties, hardened analysis and crack analysis are presented in detail in Appendix A. The admixture was found to have no significant effect on the reduction of plastic shrinkage cracking in wet-mix shotcrete. The results of the plastic shrinkage crack analysis are summarized in Table 4.10. The effects of this admixture on drying shrinkage were not evaluated in this study.  Table 4.10 - Results of Plastic Shrinkage Crack Analysis of the Waterproofing Admixture  Mix ID  Variable  Batched Dosage  Plastic Shrinkage Crack Data Total Length (mm)  # of Cracks  Crack Area (mm)  Max Width (mm)  2  1-15  Wet Mix Control  -  727  4  927  1.2  12-12  Waterproofing Admixture  1.50%  645  5  1013  1.3  86  4.5 General Discussion of Test Program  Plastic shrinkage cracks observed in this study generally formed within two to five hours after shooting, this time of three hours would be considered the plastic crack window discussed in Chapter two. The widest cracks generally are the first to appear. As is clear in Figure 4.12, the longer the delay to the initiation of cracking the smaller the size of cracks that would form.  Time of Crack Appearance All Data E  1.2 1  *•>  0.8  0.6 U ro 0.4  o "ro  4»N  0.2  •  •  •  0 0  60  120  180  240  300  Time (minutes) Figure 4.12 - Time vs. final crack width.  The presence of fibers was found to delay on the onset of cracking. It is likely that the fibers improve the plastic strength of the shotcrete just enough to reduce both the duration of the plastic crack window and increase the differential between the strength of the shotcrete and the stress exerted by the evaporation.  87  Generally, the wider the crack, the greater the depth it would penetrate. Cracks with a width of greater than 0.5 mm would be full depth of the 50 mm overlay. Cracks with a width of less than 0.5 mm would be categorized surface cracks, extending partial depth into the 50 mm overlay.  In shotcrete samples cracks were generally oriented perpendicular to the long direction of the sample. This directional cracking is expected as the majority of the strains is in this direction, but may also be a result of the finishing process. Finishing was always performed parallel to the long direction of the sample. This process may have created sub-surface tears in the stiff shotcrete, promoting cracking.  Fiber reinforcement in the overlays obviously reduced crack widths and depths in this study. The fibers reduced crack widths by bridging the crack and providing dimensional restraint. Fibers would not prevent the cracks from forming on a microscopic level, but they prevent them from growing. As the cracks are surface initiated, fibers have the added benefit of preventing cracking from occurring full depth of the overlay. The fibers may have had the added affect of reducing evaporation rates from the overlay by forming a more cohesive matrix. The effect of reducing evaporation rates from the overlay was likely exhibited by the shrinkage controlling admixture. The shrinkage controlling admixture produced a more cohesive mix and may have bound water into the mix more intimately than the plain mixes. The reduction in the loss of water from the mix will prevent plastic shrinkage cracking. The ability of fibers to reduce evaporation may partially explain the lack of differences in plastic shrinkage cracking between differing fiber shapes.  88  Chapter 5 - Conclusion and Recommendations for Future Research  5.1 - Conclusion  The proposed test procedure and setup can be used to study plastic shrinkage cracking in bonded shotcrete overlays.  In dry-mix mix shotcrete the Brazilian method using bone-dry materials was found to create significantly more plastic shrinkage cracking than other dry-mix mix shotcreting methods. Furthermore, shotcrete samples were found to be more susceptible to plastic shrinkage cracking than samples that were shot, remixed and then cast.  The use of fiber reinforcement was found to significantly reduce the amount of plastic shrinkage cracking. Fiber reinforcement reduced both crack widths and the number of cracks appearing.  The shrinkage controlling admixture was found to significantly reduce the amount of plastic shrinkage cracking. The waterproofing admixture was found to have no significant effect of the amount of plastic shrinkage cracking.  5.2 - Suggestions and Recommendations for Future Research on the Subject  Future research should focus on the fundamentals of plastic shrinkage cracking in shotcrete. Testing should be carried out within an enclosed room to accurately vary environmental conditions. The effect of humidity and temperature should be studied.  89  In the future the concept of crack density may be a better way to compare samples. The crack density would be defined as the Crack Area/Slab Area and expressed in a percentage. This would be useful for comparing specimens of different sizes or to compare lab samples to field situations.  The effect of sample size should be studied using full size test panels. The use of full size test panels under real conditions would be ideal. Large panels, 2 or 3 metres square could be covered with a thin layer of shotcrete and allowed to cure for a period of time. The panels would then be covered with a thin shotcrete overlay and allowed to shrink and crack in the field. Testing could be carried out on a hot windy day to maximize cracking. Companion panels, cast at the same time in the field, could be transferred into the lab and placed under identical conditions within the test chamber. A comparative analysis could validate the test setup and procedure.  The effect of fiber stiffness and shape should be evaluated. Fibers with similar shape, but different material should be compared to ascertain the effects of modulus of elasticity. Straight and deformed fibers of the same material should be tested to investigate the effect of fiber geometry.  The ability of fibers to prevent evaporation from fresh shotcrete should be studied more closely. Restrained shrinkage test samples could be placed on scales within the test chamber to accurately measure evaporation rates during drying.  90  References American Concrete Institute, ACI 506R-90, Guide to Shotcrete, Sprayed Concrete: Properties, Design and Application Robins). McGraw-Hill, New York, 1995.  ACI, Detroit, 1990.  (Edited by S.A. Austin and P.J.  Austrian Concrete Society, Guideline on Shotcrete, Austrian Concrete Society, Vienna, 1990. Balaguru, P., Contribution of Fibers to Crack Reduction of Cement Composites During the Initial and Final Setting Period, ACI Materials Journal, V. 91, No. 3, May-June 1994, pp. 280288. Banthia, N., Azzabi, M., and Pigeon, M., Restrained Shrinkage Cracking in Fiber-Reinforced Cementitious Composites, RILEM Materials and Structures, 1993, 26, pp. 405-413. Banthia, N., Yan, C , and Mindess, S., Restrained Shrinkage Cracking in Fiber Reinforced Concrete: A Novel Test Technique, Cement and Concrete Research, Vol. 26, No. 1, pp. 914,1996. Bloom, R, and Bentur, A, Free and Restrained Shrinkage of Normal and High-Strength Concretes, ACI Materials Journal, Vol. 92, No. 2, March-April 1995, pp. 211-217. Cabrera, J . G . , Cusens, A.R. and Brookes-Wang, Y. Effect of Superplasticizers on the Plastic Shrinkage of Concrete, Magazine of Concrete Research, Vol. 44, No. 160, September 1992, pp. 149-155. Carlson, R.W., and Reading, T.J., Model Study of Shrinkage Cracking in Concrete Building Walls, ACI Structural Journal, July-August 1988, pp. 395-404. Dahl, P.A., Plastic Shrinkage and Cracking Tendency of Mortar and Concrete Containing Fibermesh, Report No. STF65 A85039, SINTEF Div. FCB, Trondheim, Norway, September 1985. Grzybowski, M., Determination of Crack Arresting Properties of Fiber Reinforced Cementitious Composites, Department of Structural Engineering, Royal Institute of Technology, Stockholm, Sweden, June 1989. Grzybowski, M. and Shah, S.P., Shrinkage Cracking of Fiber Reinforced Materials Journal, V. 87, No.2, March-April 1990.  Concrete, ACI  Khajuria, A, Balaguru, P., Plastic Shrinkage Characteristics of Fibre Reinforced Cement Composites, Fiber Reinforced Cement and Concrete, Edited by R.N. Swamy, RILEM, 1992, pp. 82-90. Kovler, K., Sikuler, J . , Bentur, Free and Restrained Shrinkage of Fiber Reinforced Concrete with Low Polypropylene Fiber Content at Early Age, Fiber Reinforced Cement and Concrete, Edited by R.N. Swamy, RILEM, 1992, pp. 91-101. Kovler, K, Testing System for Determining the Mechanical Behavior of Early Age Concrete Under Restrained and Free Uniaxial Shrinkage, RILEM Materials and Structures, 1994, XX, pp. 324-330.  91  Kovler, K., Sikuler, J . , Bentur, A., Restrained Shrinkage Tests of Fiber-Reinforced Concrete Ring Specimens: Effect of Core Thermal Expansion, RILEM Materials and Structures, 1993, 26, pp. 231-237. Kraai, P.P., A Proposed Test to Determine the Cracking Potential Due to Drying Shrinkage of Concrete, Concrete Construction, September 1985, pp. 775-778. Malhotra, V . M . , Carette, G . G . , and Bilodeau, A., Mechanical Properties and Durability of Polypropylene Fiber Reinforced High-Volume Fly Ash Concrete for Shotcrete Applications, ACI Materials Journal, V. 91, No. 5, September-October 1994, pp. 478-486. Melbye, T, Opsahl, O, and Holtmon, J , Shotcrete for Rock Support. Sprayed Concrete: Properties, Design and Application (Edited by S.A. Austin and P.J. Robins). McGraw-Hill, New York, 1995. Morgan, D.R., McAskill, N, Carette, G.G., and Malhotra, V . M . , Evaluation of Polypropylene Fiber Reinforced High-Volume Fly Ash Shotcrete, ACI Materials Journal, V. 89, No. 2, March-April 1992, pp. 169-177. Opsahl, O.A., Kvam, S . E . , Concrete with EE-Steel Fibers, Report No. S T F 65 A82036, SINTEF Div. FCB, Trondheim, Norway, June 1982. Paillere, A . M . , Buil, M., and Serrano, J.J., Effect of Fiber Addition on the Autogenous Shrinkage of Silica Fume Concrete, ACI Materials Journal, V. 86, No.2, March-April 1989, pp. 139-144. Pan, X.W., et al., A Study on Restrained Materials, 1987  Shrinkage  Cracking of Fly Ash  Cementitious  Phillajavaara, S.E., Pihlman, E . , Results of Long-Term Deformation Tests of Glass Fiber Reinforced Concrete, Nordforsk-FRC Project 1974-1976, Delrapport O Technical Research Centre of Finland, Otaniemi, 1978. Ravina, D., and Shalon, R., Tensile Stress and Strength of Fresh Mortar Subjected to Evaporation, Proceedings, RILEM International Symposium on Concrete and Reinforced Concretet in Hot Countries, Technion, Israel Institute of Technology, Haifa, 1971, pp. 275296. Sarigaphuti, M., Shah, S.P., and Vinson, K.D., Shrinkage Cracking and Durability Characteristics of Cellulose Fiber Reinforced Concrete, ACI Materials Journal, V . 90, No. 4, July-August 1993, pp. 309-318. Saucier, F., Detriche, C . H . , and Pigeon, M, Etude de la Deformabilite en Traction d'un Beton de Reparation, RILEM Materials and Structures, 1992, 25, pp. 335-346. Shaeles, C.A., Hover, K . C . , Influence of Mix Proportions and Construction Operations on Plastic Shrinkage Cracking in Thin Slabs, ACI Materials Journal, V. 85, No. 6, NovemberDecember 1988, pp. 495-504. Shah, S.P., Sarigaphuti, M., and Karaguler, M.E., Comparison of Shrinkage Cracking Performance of Different Types of Fibers and Wiremesh, Fiber Reinforced Concrete Developments and Innovations, SP-142, ACI, Detroit, 1993, pp.1-18.  92  Appendix A - Summary of Test Data  93  Table A1 - Calibration Summary: In-situ Aggregate, Cement and Water Content  Mix Id  Nozzle  Material  Placement  State  Technique  In-Situ Material CA.  Sand  kg/m Cement  Water  Paste  w/c  3  8-15 8-14 8-13  Traditional  Damp  Shot  474 450 450  1062 1093 1094  529 517 497  190 194 213  719 711 710  0.36 0.38 0.43  8-12 8-10 8-9  Traditional  Dry  Cast  348 316 336  1264 1294 1269  440 447 482  189 193 200  629 640 682  0.43 0.43 0.41  8-8 8-7 8-6  Traditional  Dry  Shot  306 341 328  1295 1260 1273  458 455 459  194 199 194  652 654 653  0.42 0.44 0.42  7-10 7-9 7-8  Brazilian  Damp  Cast  187 413 439  1346 1177 1133  522 467 498  200 199 185  722 666 683  0.38 0.43 0.37  1198 1097 1151 1157  444 542 473 465  180 197 194 189  624 739 667 654  0.41 0.36 0.41 0.41  7-4 7-3 7-2 6-27  Brazilian  Damp  Shot  432 418 437 443  6-25 6-24 6-21  Brazilian  Dry  Cast  343 351 321  1250 1229 1267  452 480 468  220 206 209  672 686 677  0.49 0.43 0.45  1247 1210 1304 1294  446 548 466 516  194 280 204 240  640 828 670 756  0.43 0.51 0.44 0.47  921 914  454 438  218 227  672 665  0.48 0.52  6-20 6-19 6-18 6-17  Brazilian  Dry  Shot  379 225 291 215  5-9 5-7  Brazilian  Dry  Unfinished Over Finished  657 671  94  Table A2 - Calibration Summary: Hardened Properties  Compressive Strengths ASTM C642 (MPa) Voids Boiled 7d 7d cast 28d 28d cast Volume Absorption  Mix Id  Nozzle  Material State  Placement Technique  8-15 8-14 8-13  Traditional  Damp  Shot  47.6 47.8 41.7  8-12 8-10 8-9  Traditional  Dry  Cast  37.6 38.2 43.1  8-8 8-7 8-6  Traditional  Dry  Shot  42.7 38.7 40.2  7-10 7-9 7-8  Brazilian  Damp  Cast  45.9 46 52.1  66.1 65.5 61.2 31.5 35.2 38.8  47.2 50.1 58.2 55.9 46 50.8  31.6 42.4 48.2  49.4 66.4 66.3  7-4 7-3 7-2 6-27  Brazilian  Damp  Shot  45.1 46.6 46.4 42.2  6-25 6-24 6-21  Brazilian  Dry  Cast  46.2 51.9 47.4  64.3 65.9 70.6 65.9 72.4 66.5  6-20 6-19 6-18 6-17  Brazilian  Dry  Shot  49.5 51.8 58.2 57.1  5-9 5-7  Brazillian  Dry  Unfinished Over Finished  56.9 55.4  95  41.3 46.8 53  40.7 48.6 60.6  57.7 61.8 60.8 55.5 38.3 49.7 48.4  65.7 58.6 62.4  59.1 64.8 62.5  11.1% 10.5% 10.8%  4.9% 4.7% 4.8%  11.3% 10.8% 11.5%  5.0% 4.8% 5.2%  13.5% 13.9% 12.5%  6.2% 6.2% 5.7%  10.0% 10.7% 9.2%  4.5% 4.8% 4.1%  13.4% 12.1% 12.2% 12.4%  6.0% 5.5% 5.4% 5.5%  13.4% 13.4% 13.7%  6.1% 6.0% 6.1%  13.5% 11.4% 14.0% 12.8%  6.1% 5.1% 6.3% 5.8%  Table A3 - Calibration Summary: Plastic Shrinkage Results  Mix Id  Nozzle  Material State  Crac k Data Total Number Crack Length of Cracks Area (mm) (mm )  Placement Technique  2  Max width (mm)  8-15 8-14 8-13  Traditional  Damp  Shot  410 405 315  3 3 4  542 549.3 690.5  0.45 0.60 0.80  8-12 8-10 8-9  Traditional  Dry  Cast  380 320 440  2 2 3  121.3 167.5 158  0.15 0.35 0.20  8-8 8-7 8-6  Traditional  Dry  Shot  1795 1856 1825  5 7 6  752.1 796.8 856.4  0.80 0.50 0.95  7-10 7-9 7-8  Brazilian  Damp  Cast  285 390 370  2 3 2  198.4 150.2 165.5  0.30 0.15 0.20  4 5 4 4  510.2 724.9 650.1 475  0.60 0.80 0.60 0.50  7-4 7-3 7-2 6-27  Brazilian  Damp  Shot  1355 1105 1195 1480  6-25 6-24 6-21  Brazilian  Dry  Cast  990 1270 505  2 3 2  178.5 150.8 143.7  0.25 0.20 0.15  5 7 5 6  924.5 1094.3 915.2 968.8  1.00 1.20 0.80 1.15  2 NA  404 119  1.00 0.35  6-20 6-19 6-18 6-17  Brazilian  Dry  Shot  1665 1885 1565 1780  5-9 5-7  Brazillian  Dry  Unfinished Over Finished  440 775  96  Table A4 - Final Test Progam: In-situ Aggregate, Cement and Water Content  Mix ID 4-3 5-16 5-27 12-9 6-4 10-16 5-23 10-18 10-19 12-11 12-13 10-7  Fiber Type  Control  Hooked End  Flattened End  Batched Dosage  CA.  In-situ Material Content kg/m Sand Cement Water Fiber Paste  w/c  462 558  1161 923  437 512  191 257  0  -  628 769  0.44 0.50  9 10 18 20 27 40 60  202 245 354 391 209 466 408  1254 1205 1255 1176 1247 1166 1140  558 564 437 464 577 411 464  245 234 212 204 211 187 198  5.8 3 13 17 21.2 23 33  803 798 649 668 788 598 662  0.44 0.41 0.49 0.44 0.37 0.45 0.43  10 20 40  338 309 337  1218 1254 1207  474 474 478  219 205 199  8 16.5 32  693 679 677  0.46 0.43 0.42  -  3  12-3 10-24 10-25  Crimped  20 40 60  245 362 356  1205 1238 1242  564 435 420  234 201 196  8 16 38  798 636 616  0.41 0.46 0.47  5-28 5-29  50 mm Poly  1.50% 3.00%  375 317  1205 1277  475 449  202 211  7.8 12  677 660  0.43 0.47  6-5 6-3 5-31  25 mm Poly  0.25% 0.50% 1.00%  294 356 324  1255 1230 1176  502 465 532  214 213 227  1.2 2 4.6  716 678 759  0.43 0.46 0.43  ? ?  692 771  0.42 0.40  ?  680 688  0.40 0.40  Shrinkage Controlling Admixture 12-17 12-14  Shrinkage Controlling  1% 2%  323 271  1237 1210  488 552  204 219  Wet Mix Control Waterproofing  -  342 340  1028 1030  485 490  195 198  Wet Mix 1-15 12-12  1.50%  97  Table A5 - Final Test Program: Hardened Properties  Mix ID  Fiber Type  Batched Dosage (kg/m ) 3  4-3 5-16 5-27 12-9 6-4 10-16 5-23 10-18 10-19 12-11 12-13 10-7  Compressive Strength 7 Day 28 Day (MPa) (MPa)  Voids Volume  Boiled Absorption  Control  -  48.0  65.4  13.0%  5.8%  Hooked End  9 10 18 20 27 40 60  59.6 43.2 44.9 47.7 56.3 43.7 47.8  69.5 57.8 50.8 65.2 69.5 55.7 65.0  13.5% 11.5% 13.0% 12.0% 13.3% 12.6% 12.6%  6.0% 5.2% 5.8% 5.4% 5.9% 5.7% 5.6%  10 20 40  47.2 42.0 46.8  61.9 54.3 64.0  13.4% 12.4% 13.9%  5.9% 5.5% 6.2%  Flattened End  12-3 10-24 10-25  Crimped  20 40 60  44.8 42.8 43.7  59.6 56.2 56.3  11.3% 14.0% 12.3%  5.1% 6.2% 5.5%  5-28 5-29  50 mm Poly  1.50% 3.00%  47.4 no  61.2  12.8%  5.7%  6-5 6-3 5-31  25 mm Poly  0.25% 0.50% 1.00%  42.0 49.3 46.7  55.3 57.2 60.6  11.9% 12.8% 12.0%  5.3% 5.7% 5.4%  Shrinkage Controlling Admixture 12-17 12-14  Shrinkage Controlling  1% 2%  45.8 43.6  61.2 58.7  11.4% 12.5%  5.1% 5.6%  Wet Mix Control Waterproofing  1.50%  38.6 42.1  52.8 51.0  4.9% 5.0%  10.7% 11.5%  Wet Mix 1-15 12-12  98  Table A6 - Final Test Program: Plastic Shrinkage Test Data  Mix ID  Fiber Type  Batched Fiber Dosage (kg/m)  In-situ Fiber Volume  -  0.00% 0.00%  2425 2235  8 8  1320 1257  1.05 1.10  9 10 18 20 27 40 60  0.07% 0.10% 0.17% 0.22% 0.27% 0.12% 0.42%  1525 1280 905 260 240 920 230  6 4 6 4 2 3 2  453 406 130 75 30 260 11.5  1.10 1.00 0.50 0.25 0.15 0.70 0.05  10 20 40  0.10% 0.21% 0.41%  1120 850 685  3 2 2  411 205 118  1.15 0.50 0.25  3  4-3 5-16 5-27 12-9 6-4 10-16 5-23 10-18 10-19 12-11 12-13 10-7  Control  Hooked End  Flattened End  Total Length (mm)  Crack Data Number Crack of Cracks Area (mm) 2  Max width (mm)  12-3 10-24 10-25  Crimped  20 40 60  0.14% 0.20% 0.48%  1755 1360 380  5 4 2  574 390.5 135  1.00 0.80 0.35  5-28 5-29  50 mm Poly  1.50% 3.00%  0.86% 1.32%  615 155  3 1  95 16  0.30 0.20  6-5 6-3 5-31  25 mm Poly  0.25% 0.50% 1.00%  0.13% 0.22% 0.51%  1550 780 700  5 3 4  502 207 72  0.85 0.95 0.30  ? ?  975 755  3 3  276 91  0.7 0.3  0.00%  727 645  4 5  927 1013  1.2 1.3  Shrinkage Controlling Admixture 12-17 12-14  Shrinkage Controlling  1% 2%  Wet Mix Control Waterproofing  1.50%  Wet Mix 1-15 12-12  99  Appendix B - Photo Log  100  1.  Typical repair surface preparation. Loose deteriorated concrete has been chipped away exposing roughened sub-base and reinforcing. The area has been sand-blasted to remove bruised concrete and corrosion products from the reinforcing steel.  2.  Aliva 246 dry mix shotcrete pump and operator. Note the air flow meter located on the pump used to regulate airflow and the water pump used to pressurize water for the water ring.  101  3. Sample set-up within the shooting chamber immediately prior to shooting. The test base is present within the formwork in a saturated, surface damp condition.  4. Initial stages of the shooting, note the large amount of stone rebounding from the panel.  102  5. Shotcreting in progress. The shotcrete has a very wet consistency, used to aggravate plastic shrinkage cracking.  6. Test set-up with the shrinkage samples in place. The three construction heaters are in place with the air flow diverter. The air flow diverter is intended to channel the hot air over the sample.  103  7.  Measuring crack widths with the optical microscope mounted on the tracks.  UBC DEPT  OF CIVIL ENG  RESTRAINED  SHRINKAGE  OF SHOTCRETE MIX NO 4-3 9.  Mix No. 4-3.  JUNE 1996  Duplicate control sample for test series.  105  , |jjiniii I | f - M RESTRAINED SHRINKAGE OF SHOTCRETE  MIXN01E-9  DEC 1996  11. Mix No. 12-9. Hooked End steel fiber with a batched fiber dosage of 10 kg/m  106  13. Mix No. 10-16. Hooked End steel fiber with a batched fiber dosage of 20 kg/m . 3  14. Mix No. 10-18. Hooked End steel fiber with a batched fiber dosage of 40 kg/m . 3  107  17. Mix No. 12-13. Flattened End steel fiber with a batched fiber dosage of 20 kg/m  UBC DEPT OF CIVIL ENG RESTRAINED  SHRINKAGE  OF SHOTCRETE MIX NO 5-29  JUNE 1996  20. Mix No. 5-29. Straight 50 mm Polyolefin fiber with a batched fiber dosage of 3.00%.  110  UBC DEPT  OF CIVIL ENG  RESTRAINED  SHRINKAGE  OF SHOTCRETE JUNE 1996  MIX NO 6-3  22. Mix No. 6-3. Straight 25 mm Polyolefin fiber with a batched fiber dosage of 0.50%.  111  UBC DEPT  OF CIVIL ENG  RESTRAINED  SHRINKAGE  OF SHOTCRETE MIX NO 5-31  JUNE 1996  23. Mix No. 5-31. Straight 25 mm Polyolefin fiber with a batched fiber dosage of 1.00%.  24. Mix No. 5-9. Unfinished shotcrete sample produced with the Brazilian method and bone dry material.  112  25. Mix No. 5-7. Over finished shotcrete sample produced with the Brazilian method and bone dry material.  113  Appendix C Restrained Plastic Shrinkage Test Procedure  114  Proposed Standard Test Method for the Evaluation of Plastic Shrinkage Capacity in Bonded Shotcrete Overlays  1.0 Scope  1.1 This test method covers determination of the plastic shrinkage capacity of bonded shotcrete overlays exposed to severe drying conditions. 1.2 The term plastic shrinkage capacity as used here is defined as the degree of cracking in the plastic shotcrete overlay. 1.3 This test method is intended to produce comparative results for the evaluation of plastic shrinkage in shotcrete.  2.0 Significance and Use  2.1 Measurements of the degree of plastic shrinkage cracking permit assessment of shrinkage controlling admixtures and fibers in bonded shotcrete overlays. This method is particularly useful for comparative evaluation of plastic shrinkage cracking in different shotcrete mixtures.  3.0 Apparatus  3.1 Molds - The molds for casting test specimens shall measure 1000 mm by 1000 mm and 100 mm in depth. The molds shall be constructed of plywood and shall use wood screws for connections. 3.2 Drying Room and Controls - A drying chamber with a suitable viewing area shall be designed for free circulation of air over the sample. Heating will be provided by three construction heaters. Temperature shall be maintained at a constant  115  rate of 5 5 ° C . Wind speed over the sample shall be maintained at a velocity of 5 m/s for the duration of the test. The evaporation rate within the test chamber shall be measured at least once per week. Evaporation  3.2.1  Rate - The evaporation rate within the chamber shall be  measured by placing a steel tray with a known mass of water within the drying chamber for a period of 15 minutes. Initial and final masses shall be divided by the area of steel tray to give an evaporation rate in kg/m /hour. 2  3.3 Micrometer-  A micrometer with a magnification rate of at least 20X and a scale  capable of measuring to the nearest 0.05 mm shall be used to measure the widths of all cracks. 3.4 Measure  Tape - A flexible tape shall be used to measure the length of ail cracks  to the nearest 5 mm.  4.0 Sampling  4.1 Sampling for the water content test and cement test shall only be from the center of the panel.  5.0 Preparation of Rigid Base  5.1 Concept - The rigid base is meant to provide dimensional restraint to the shotcrete overlay. The base is intended to closely resemble a typical reinforced concrete structure that has deteriorated and undergone surface treatment in preparation for repair. All rigid bases for a particular series should be constructed at the same time to eliminate the variability of surface roughness. All rigid bases should be moist cured for at least one week. The rigid bases should  116  not be used in the test program until at least an age of six months. At an age of six months the majority of the shrinkage strains within the bases should be relieved. 5.2 Size and shape - The rigid base shall measure 900 mm by 300 mm and 50 mm deep. 5.3 Material - The rigid base shall be constructed of a 35 MPa concrete. The mix shall contain a 60:40, coarse aggregate:sand aggregate ratio. The coarse aggregate content within the mix shall be of a crushed or angular nature; rounded aggregate should not be used. The base may be nominally reinforced with a wire mesh or reinforcing steel. 5.4 Surface Preparation  - The surface of the rigid base shall be produced by  applying a surface retarderto the sample after placement. Three hours after the concrete has been placed roughen the surface of the sample by scrubbing the surface with a stiff brush and spraying the surface with a hose. Continue scrubbing the surface until the coarse aggregate is exposed and the panel has a uniform roughness.  6.0 Test Specimens  6.1 Size and Shape - The mold for the plastic shrinkage sample shall be rectangular in dimension, 1000 mm by 400 mm and 100 mm deep. 6.2 Number-  Only one test sample shall be taken for each mix.  6.3 Molding - When shooting the sample, the bottom corners will be filled first, followed by the lower half of the mold. After sufficient material is built up in the bottom half of the mold, the nozzle shall be brought in close to the formwork to remove any trapped rebound. The form shall be filled in one continuous operation with no stopping during the shooting operation.  117  7.0 Procedure  7.1 Formwork  Preparation  - Prepare the formwork with hydraulic oil. Hydraulic oil or  similar heavy oil should be used to ease in the removal of the formwork. Other wood form release mixtures tend to bond more to the fresh shotcrete disturbing the sample and initiating cracks. 7.2 Formwork  Placement  - Place the shrinkage mold beside the compressive  strength mold; both should lye on a 6 0 ° angle to the floor. Keep the panels close together to ensure the same shooting conditions are used for both panels. 7.3 Sample Placement  - Place the rigid base within the shrinkage mold allowing it to  rest on two spacer blocks. The spacer blocks can be of any rigid inert material, but must be a 50 mm cube. Ensure the base is in a saturated, surface dry condition before placement. If the base is dry, water will be absorbed from the fresh shotcrete and diminish bond capacity. 7.4 Application  of the Shotcrete  - Air pressure, pump speed and water pressure  must be kept constant during the shooting operations. Once the desired consistency of the material is attained by moderating the amount of water input, begin shooting the companion mold and shrinkage sample. A companion mold is always produced with the shrinkage specimen for hardened analysis of the shotcrete. 7.5 Surface Preparation  - Immediately after shooting the shrinkage sample excess  materials cut away form the sample using the edge of a steel trowel. A wood float is then used to finish and smooth the surface. Finally a steel trowel is lightly passed over the surface to further smooth the surface. Care must be taken to not overly disturb the shotcrete when finishing and moving the sample, as this will  118  affect the results. The steel trowel finish makes for easy crack identification and measurement. 7.6 Transportation  of Sample to Chamber - Immediately after finishing, the sample  must be moved into the drying chamber. Special care must be taken to not overly vibrate or disturb the sample during movement. 7.7 Test Setup - The panel is placed perpendicular to the airflow, 200 mm form the front of the fans. The side of the sample that was at the bottom during shooting faces the fans. The heating fans are turned on and the Plexiglas cover is closed creating a covered environment. 7.8 Form Removal - O n e hour after shooting the Plexiglas cover is raised and the forms are carefully removed to expose all four sides of the panel. 7.9 Recording  of Data - Record and time the appearance of cracking on the slab.  After 24 hours the heaters are turned off and the panel is allowed to cool. The length and width of the cracks are recorded to an accuracy of +\- 5 mm and +\0.05mm.  8.0 Calculation  8.1 Calculate the crack area by multiplying the length of crack by it corresponding width and record to the nearest 10 mm . 2  8.2 Calculate the water/cement ratio for each mix based on the water content results and the cement content results. 8.3 Calculate the mass of cement per m based on the washout test results. 3  8.4 Calculate the mass of sand and coarse aggregate content per m based on the 3  washout test results.  119  9.0 Report  9.1 A report shall be made for each specimen tested and shall include the following information where applicable:  9.1.1  Identification number for the test sample.  9.1.2  Shotcrete process used, wet or dry.  9.1.3  Type of pump used in producing the shotcrete.  9.1.4  Batched mix design.  9.1.5  As shot mix contents.  9.1.6  Total length of cracks.  9.1.7  Number of cracks appearing.  9.1.8  Maximum crack width.  9.1.9  Total crack area.  120  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0050144/manifest

Comment

Related Items