Citation
Alternate methods for testing shear strength at a bonded concrete interface

Material Information

Title:
Alternate methods for testing shear strength at a bonded concrete interface
Creator:
Swan, Anne Elise ( author )
Place of Publication:
Denver, Colo.
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electronic file (75 pages) : ;

Thesis/Dissertation Information

Degree:
Master's ( Master of science)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Civil Engineering, CU Denver
Degree Disciplines:
Civil engineering

Subjects

Subjects / Keywords:
Shear (Mechanics) ( lcsh )
Concrete -- Testing ( lcsh )
Concrete construction -- Maintenance and repair ( lcsh )
Concrete construction -- Maintenance and repair ( fast )
Concrete -- Testing ( fast )
Shear (Mechanics) ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
Two methods for testing shear strength were explored at the University of Colorado, Denver in order to determine if those methods would be appropriate to evaluate the shear strength at a bonded concrete interface. The first type of test is a direct shear stress test. Twelve beams with three different surface treatments between two layers of concrete were tested using a modified Iosipescu loading scheme until a slipping (direct shear) failure occurred at the interface. The second type of test is a three point flexural shear stress test. Six beams with three different surface treatments between two layers of concrete were tested using flexure until a slipping (flexural shear) failure occurred at the interface. In both tests, the two layers of concrete are identical to one another apart from a difference in color. The success of the direct shear test makes that method a good candidate for a formalized standard to test the direct shear stress at a bonded concrete interface. If formalized into a standard test, this would be a relatively simple and effective way for the engineering community to test the shear strength of concrete for the application of repairs and overlays. However, the results from the flexural shear stress test suggest the test method is not effective for evaluating the shear stress at a bonded concrete interface.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: Adobe Reader.
Statement of Responsibility:
by Anne Elise Swan.

Record Information

Source Institution:
University of Colorado Denver Collections
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
987265194 ( OCLC )
ocn987265194
Classification:
LD1193.E53 2016m S93 ( lcc )

Downloads

This item has the following downloads:


Full Text
ALTERNATE METHODS FOR TESTING
SHEAR STRENGTH AT A BONDED CONCRETE INTERFACE
by
ANNE ELISE SWAN B.S., Colorado State University, 2011
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Civil Engineering Program
2016


This thesis for the Master of Science degree by Anne Elise Swan has been approved for the Civil Engineering Program by
Frederick Rutz, Chair Carnot Nogueira Peter Marxhausen
Date: 17 December 2016


Swan, Anne Elise (B.S., Civil Engineering)
Alternate Methods for Testing Shear Strength at a Bonded Concrete Interface Thesis directed by Associate Professor Frederick Rutz
ABSTRACT
Two methods for testing shear strength were explored at the University of Colorado, Denver in order to determine if those methods would be appropriate to evaluate the shear strength at a bonded concrete interface. The first type of test is a direct shear stress test. Twelve beams with three different surface treatments between two layers of concrete were tested using a modified Iosipescu loading scheme until a slipping (direct shear) failure occurred at the interface. The second type of test is a three point flexural shear stress test. Six beams with three different surface treatments between two layers of concrete were tested using flexure until a slipping (flexural shear) failure occurred at the interface. In both tests, the two layers of concrete are identical to one another apart from a difference in color. The success of the direct shear test makes that method a good candidate for a formalized standard to test the direct shear stress at a bonded concrete interface. If formalized into a standard test, this would be a relatively simple and effective way for the engineering community to test the shear strength of concrete for the application of repairs and overlays. However, the results from the flexural shear stress test suggest the test method is not effective for evaluating the shear stress at a bonded concrete interface.
The form and content of this abstract are approved. I recommend its publication.
Approved: Frederick Rutz
m


DEDICATION
This thesis is dedicated to the incredibly supportive family and friends I am fortunate enough to have in my life ... particularly Ron, Chris, and Katie Swan, Nadim Chammas, Lindsey Reynolds, Alyson Heon, and my dog Titus. Thank you for always believing in me, especially during the times when I didnt.
IV


ACKNOWLEDGEMENTS
This thesis would not have been possible without the contributions made by the following people:
Tom Thuis & Peter Sillstrop
Andy Pultorak & Christian Rosen
Geotechnical Lab Brian Volmer
Machine Shop Jac Corless & Brian Carpenter
Kris Kemp & Nadim Chammas
Most of all, Id like to thank and acknowledge Dr. Frederick Rutz, my esteemed advisor, for funding the project and assisting me every step of the way.
v


TABLE OF CONTENTS
CHAPTER
I OVERVIEW...........................................................1
1.1 Introduction................................................1
1.2 Goal........................................................3
1.3 Outline 3
II BACKGROUND.........................................................4
2.1 Introduction................................................4
2.2 Common Bond Strength Tests..................................4
2.3 Direct Shear................................................5
2.4 Flexural Shear..............................................7
III. CONTROLLING EQUATIONS & SUPPORTING CALCULATIONS...................10
3.1 Introduction 10
3.2 Direct Shear Test..........................................10
3.3 Flexural Shear Test........................................10
IV. EXPERIMENTS.......................................................12
4.1 Introducti on 12
4.2 Direct Shear Test..........................................12
4.2.1 Specimen Preparation.................................13
4.2.2 Testing..............................................15
4.3 Flexural Shear Test........................................17
4.3.1 Design 19
4.3.2 Specimen Preparation.................................20
vi


4.3.3 Testing.
27
4.3.3.1 Flexure by Compression Testing.............28
4.3.3.2 Flexure by Tension Testing,................31
4.4 Compressive Strength Testing,............................38
V RESULTS........................................................40
5.1 Introduction.............................................40
5.2 Compressive Strength Testing,............................40
5.3 Direct Shear Test........................................41
5.4 Flexural Shear Test......................................46
VI DISCUSSION.....................................................50
6.1 Introducti on............................................50
6.2 Direct Shear Test........................................50
6.3 Flexural Shear Test......................................52
VII. CONCLUSION & RECOMMENDATIONS...................................56
REFERENCES...........................................................58
APPENDIX.............................................................60
vii


LIST OF TABLES
TABLE
4.1 Flexural shear test design parameters.........................................19
4.2 Summary of beams, surface treatments, and testing methods used................24
5.1 Summary of shear stresses in psi..............................................46
5.2 Summary of t^l(fc) values....................................................46
5.3 Flexural shear test results: Beams 1-3........................................47
5.4 Flexural shear test results: Beams 4-6........................................48
6.1 Comparison of results from other researchers..................................51
viii


LIST OF FIGURES
FIGURE
4.1 Iosipescu beam specimen surface treatments..............................14
4.2 Construction of typical Iosipescu beam specimen.........................14
4.3 Iosipescu loading scheme................................................15
4.4 Photograph of specimen in the Iosipescu apparatus.......................16
4.5 Flexure induced by shear at interface combined with compression and by
shear at interface combined with tension................................18
4.6 Cross-section of typical beam specimen for flexure tests................20
4.7 Reinforcing steel used in each beam specimen............................21
4.8 Concrete form ready for first placement.................................22
4.9 Application of raked surface treatment to substrates..................23
4.10 Flexural beam specimen surface treatments...............................24
4.11 Influence areas created with the bond breaker...........................25
4.12 Form prior to second concrete placement.................................26
4.13 Design of flexure by compression test, elevation view..................29
4.14 Design of flexure by compression test, Section 1.......................30
4.15 Flexure by compression testing set-up...................................31
4.16 50 kip capacity load frame used for the flexure by tension tests........32
4.17 HSS secured to the floor used to anchor the ends of the beam specimens..33
4.18 Design for flexure by tension test, elevation view.....................35
4.19 Design for flexure by tension test, section cuts.......................36
4.20 Flexure by tension testing set-up.......................................37
4.21 Modified testing set-up for Beam 5......................................38
4.22 Compressive testing machine used........................................39
IX


5.1 Compressive strength curves from first concrete placement.......................40
5.2 Compressive strength curves from second concrete placement......................41
5.3 Iosipescu beam specimen at failure..............................................42
5.4 Shear stresses measured for the smooth interface........43
5.5 Shear stresses measured for the bush hammered interface.43
5.6 Shear stresses measured for the j ackhammered interface.........................44
5.7 Shear stresses measured for the smooth interface expressed as a function
of <(Tc)........................................................................44
5.8 Shear stresses measured for the bush hammered interface expressed as a
function of A(fc)..............................................................45
5.9 Shear stresses measured for the j ackhammered interface expressed as a
function of A(fc)..............................................................45
5.10 Maximum flexural shear stress of Beams 1-6......................................49
5.11 Maximum W(fc) values...........................................................49
6.1 Flexural shear stress failure of Beam 3.........................................52
6.2 Diagonal tension failure of Beam 1..............................................53
6.3 Failure of Beam 6...............................................................55
x


CHAPTERI
OVERVIEW
1.1 Introduction
Concrete is an extremely versatile construction material and can be used for a vast array of different types of structures. According to the World Business Council for Sustainable Development, Twice as much concrete is used in construction around the world than the total of all other building materials including wood, steel, plastic, and aluminum (World Business Council for Sustainable Development 2015). Inevitably however, concrete can deteriorate over time or become damaged and in need of repair. Simply put, repairing concrete typically entails bonding new concrete to old concrete.
Composite construction is accomplished when two or more materials are connected together so strongly that they act as a monolithic unit. Composite construction is often necessary to repair damaged structural members; the goal is to repair in a way that allows the member to act as it was originally designed. Monolithic behavior is desirable because it increases the strength and efficiency of a structure, or a member within a structure, which typically leads to a more economical design. In order for a composite unit to perform monolithically, the bonded interface must be capable of successfully transferring forces such as compression, tension, and shear. Simple and straightforward standard tests exist to measure the compressive and tensile strengths of bonds but when it comes to shear, few methods have been formalized into standard tests (Helmick et al. 2016). For this reason, engineers need an effective method to experimentally evaluate the shear strength at the bonded interface between new and old concrete for the purpose of conducting repairs or applying overlays.
1


There are three different types of shear: direct, flexural, and torsional. The tests presented in this thesis will focus on direct shear and flexural shear; torsional shear is not a part of the scope. Stress is defined as a force per unit area. Direct or general shear stress is the most basic and straightforward type and is simply the force applied divided by the cross-sectional area of the material. Flexural, often called horizontal, is more complex because it deals with the internal shear stress within a beam subject to bending.
If flexural shear properties are to be measured, a beam test specimen made up of multiple layers has to be exposed to flexure (bending). Once a perpendicularly applied force overcomes the strength of the bond between the layers, the layers will separate from one another and slip. This concept can be visualized by considering a thick phone book being bent in the middle; as the phone book goes from being flat to curved, the pages slide relative to each other.
If direct shear properties are to be measured, a test specimen has to directly experience uniformly distributed shear stress. This can be difficult to achieve because compressive or bending forces tend to be introduced in one way or another. Dr. Donald F. Adams is the founder and president of Wyoming Test Fixtures, Inc., a company that specializes in the design and fabrication of mechanical test fixtures for composite materials; he has done extensive research on shear testing and the Iosipescu method. According to Dr. Adams, There has been a greater variety of shear test methods developed and used during the past 40 years than for any other type of mechanical test of composites, including tension and compression, the two other basic tests (Adams 2009). Yet few have been formalized into standard tests. This can be attributed to the difficulty in developing a successful shear test method; many have been developed and used but
2


few have been effective enough to be formalized as standards. This is not only true of bonded materials, but monolithic concrete as well. It is difficult to apply many of the shear tests to concrete, and that difficulty is behind the lack of a formalized standard for testing the shear capacity of a bonded interface between two layers of concrete.
1.2 Goal
The goal of this research is to examine and evaluate the applicability of two different types of shear strength tests and to determine if the methods developed are appropriate to evaluate the shear strength of a bonded interface between two layers of concrete. Ideally, this would provide the engineering community with simple and effective standard tests to evaluate the shear capacity of a bonded interface.
1.3 Outline
This thesis is made up of seven chapters. The first chapter is an overview and a description of the goal.
Chapter 2 contains background on the material including a literature review. Chapter 3 presents the controlling equations and supporting calculations required. Chapter 4 explains the tests conducted.
Chapter 5 presents the results of the tests.
Chapter 6 analyzes the results of the tests.
Chapter 7 contains a summary of conclusions drawn from the test results.
3


CHAPTER II
BACKGROUND
2.1 Introduction
When dealing with plain concrete, the shear strength of a bonded interface comes mainly from adhesion and the mechanical interlock of the aggregate between a base or bottom layer, referred to herein as the substrate, and a top or second layer, referred to herein as the overlay. Because of this, it is fairly intuitive that the rougher the surface, the better this shear capacity will be. This is a phenomenon recognized by the International Concrete Repair Institute, Inc. (ICRI). There are many formalized standards to test different properties of a bond but very few for shear.
2.2 Common Bond Strength Tests
One of the most common tests performed on bonded concrete is the pull-off test, formally standardized as ASTM Cl583, where a pure tensile force acts on a bond. A core is drilled through an overlay and at least 1 inch or half the core diameter into the substrate. With the use of an epoxy, a steel puck is attached to the overlay. Once the epoxy has set, the testing apparatus is attached to the puck and an upward force is applied until failure occurs (ASTM Cl 583 2015). Other than a failure of the epoxy connection, there are three possible outcomes and each will indicate something regarding the tensile strength of the bond. This test can be done in the field and is very straightforward and easy to perform, hence its popularity. However there are not very many situations in the real world that will apply this type of force alone to a bond so other tests are necessary to truly evaluate the shear strength at a bonded interface.
4


The Arizona Slant Shear Test, commonly called the slant shear test, has been formalized as ASTM C882 to specifically determine epoxy bond strength. For this test, core samples are taken of the overlay and substrate at an angle so the bonded interface is at a diagonal. Once the ends have been cut so that the core has flat top and bottom surfaces, a compressive force is applied at each end of the sample until failure occurs (ASTM C882 2015). Again, the different possible outcomes provide information specific to the strength of the bond but the clamping action that occurs makes it difficult to evaluate the shear strength alone without the influence of compression (Rosen 2016). Because of the clamping action and the angle at which the load is being applied relative to the interface surface, this is not a good test to evaluate direct shear strength.
2.3 Direct Shear
It is difficult to design a test that measures direct shear strength alone. Some tests that have tried to do so include the push-off test and the guillotine test. A push-off test is when a concrete overlay is literally pushed-off of a substrate below with the use of a jack. Due to physical restrictions, it is practically impossible to do this without some eccentricity because the jack is placed on the overlay above the interface. This creates another situation where shear is not acting alone; bending moments are introduced which influence results. Double shear guillotine tests are performed on core samples taken perpendicular to the surface so the bonded interface is straight across the center of the sample. The core is placed horizontally into a device sometimes called a guillotine shear jig. The jig is composed of two cradle type boxes. Half of the sample (a cored specimen) lays in one cradle while the other cradle lays on top of the other half of the sample. While the bottom is held in place a compressive force is applied to the top until the sample splits
5


between the two pieces of the guillotine. Single shear guillotine tests, sometimes called Iowa Tests, have also been developed. These guillotine tests may be the closest the engineering community has achieved toward a direct and pure shear test; however this method has not been formalized as a standard.
One popular method used to test in-plane shear properties of composites was originally developed by Nicolae Iosipescu of Romania in the early 1960s and was specifically used at that time for testing metals and other isotropic materials by inducing a high shear stress at the location of interest. Over time, this method has been adapted to evaluate orthotropic and composite materials (Adams and Walrath 1986). Today, the modified method is formalized as ASTM D5379, Standard Test Method for Shear Properties of Composite Materials by the V-notched Beam Method but has not yet been formalized as a standard fortesting shear in concrete (Helmick et al. 2016).
In an effort to formalize the standard for concrete applications, a journal article titled Evaluation of Shear and Diagonal Tension in Plain Concrete published in January of 2016 describes the success of using the Iosipescu loading scheme to test the shear capacity of plain monolithic concrete (Helmick et al. 2016). The authors compared results for modified Iosipescu, flexural, split cylinder, and split prism tests and found that the modified Iosipescu test method provided consistent results. Therefore they recommended that it be considered as a candidate as a useful standard for predicting the direct shear strength of plain concrete (Helmick et al. 2016). Because they had such positive results using this method on plain monolithic concrete to determine its shear strength, this author thought it might also be applied to a two-layered concrete beam to evaluate the shear
6


strength of a bonded interface with the changing variable being differing surface treatments at the interface.
2.4 Flexural Shear
When designing a concrete-to-concrete repair, if the calculated demand for interface shear stress exceeds 60 psi, Sections 7.4.2 and 7.4.3 of ACI 562-16 titled Code Requirements for Assessment, Repair, and Rehabilitation of existing Concrete Structures and Commentary require the use of interface reinforcement to transfer the forces across the interface (ACI 562 2016). This 60 psi limit is derived from a nominal shear strength of 80 psi multiplied by a strength reduction factor of (p = 0.75. This is based on established requirements and design methodologies included in ACI 318 and which are familiar to engineers (Brewe et al. 2016). But according to research done in the past, this number is quite conservative because it gives little consideration to the horizontal shear resistance provided by the concrete interaction alone (Kovach and Naito 2008). Therefore it does not represent the actual behavior of concrete repairs.
In 1994, Loov and Patnaik presented an article entitled, Horizontal Shear Strength of Composite Concrete Beams with a Rough Interface outlining the testing they did on sixteen simply supported beams with a point load applied at the center. Although their research involved the use of stirrups, they determined that the stirrups were not stressed until a horizontal shear stress of about 220-290 psi was reached, which suggests that as the strength of the interface without reinforcement. Not until about 430 psi did the stirrups become relatively effective (Loov and Patnaik 1994). Seven years later,
Patnaik extended the research to include beams with a smooth interface and published an article that describes their behavior. The horizontal shear strength at failure ranged from
7


200-900 psi, but this research was done using interface reinforcement as before with the addition of varying clamping forces, thus the large range (Patnaik 2001).
In 2003, Gohnert outlined his research in an article titled, Horizontal Shear Transfer across a roughened surface. He suggested using a design curve that expressed horizontal shear strength as a function of the surface roughness. The curve was formed from horizontal shear strength test results from the push-off method ranging from 70-300 psi. The lower numbers represent beams that had a smooth interface surface and the higher numbers represent beams that had a rough interface surface. Gohnert then tested six additional beams bent in flexure and found the experimental predictions were conservative in every case (Gohnert 2003).
ATLSS (Advanced Technology for Large Structural Systems) Report No. 08-05 published in 2008 presents research done by Kovach and Naito regarding the Horizontal Shear Capacity of Composite Concrete Beams without Interface Ties. Five-point and two-point loading tests were conducted on multiple beam specimens and the horizontal shear stress results ranged from 475-1000 psi. To conclude the research, the report recommends using horizontal shear capacities of 435, 465, and 570 psi for composite concrete sections with a broomed, as-placed, and raked interface surface treatments, respectively (Kovach and Naito 2008).
The numbers from the above mentioned research suggest a typical horizontal shear strength capacity of a bonded concrete interface ranging from about 200-600 psi, depending on the surface treatment (well above the 60 psi code provision). Because of the use of various testing methods and differing surface treatments, it is difficult to determine the actual horizontal shear capacity of a bonded concrete interface without
8


reinforcement, but it certainly suggests a value greater than the 60 psi limit used in practice currently. The experimentation presented in this thesis is meant to build upon the research conducted by others in order to explore test methods that might lead to a standardized test.
9


CHAPTER III
CONTROLLING EQUATIONS & SUPPORTING CALCULATIONS
3.1 Introduction
In order to develop shear strength tests that may be formalized into standards, it is important to understand the concept of shear as related to the mechanics of a material. The following outlines the controlling equations and supporting calculations required for the design and evaluation of the shear strength tests conducted.
3.2 Direct Shear Test
Stress and strain are the two basic parameters used when defining the mechanics of a material. When a material is loaded with a force, it produces a stress. Strain is the response of a system to that applied stress. There are three types of stress: tensile, compressive, and shearing. Tensile and compressive stresses occur normal to a plane and are usually denoted as normal stress (a) while shearing stress occurs parallel to a plane and is usually denoted as shear stress (t). The direct shear test conducted for this thesis only requires the simple equation used to express the shear stress of a material:
V
T ~ A
Where: r = shear stress (psi, N/m2)
V = shear force (lb, N)
A = surface area parallel to the force (in2, mm2)
3.3 Flexural Shear Test
As mentioned in the introduction, flexural shear stress is more complex than direct shear stress because it involves the internal stresses of a beam caused by flexure
10


(bending). To understand flexural shear stress, first it is important to understand a term called shear flow. Shear flow describes the lateral (transverse) or longitudinal shear load applied at the interface of a built-up beam induced by an externally applied force. Shear flow (q) is defined by the following equation:
VQ
q=T
Where: q = shear flow (lb/in, N/m)
V = shear force at the section (lb, N)
Q = first moment of area (in3, mm3)
= for a rectangular beam / = moment of inertia (in4, mm4)
bh^
= for a rectangular beam
Shear stress (t) can then be defined and calculated by dividing the shear flow by the width (b) of the beam supporting the stress:
_VQ
T lb
Where: b = width of cross-section at location of interest (in, m)
This equation for shear stress can be used for narrow beams. A beam is considered narrow when the ratio of b/h is small, typically equal to or less than 0.25. If a beam is narrow and has a rectangular cross-section, the above equation can be simplified in the following way:
_VQ _V X bh2/Q 12V 3V lb bh3j^ x b Bbh. 2A
11


CHAPTER IV
EXPERIMENTS
4.1 Introduction
Two different experiments were designed to evaluate the shear strength at a bonded concrete interface. The direct shear test used a modified Iosipescu method to induce a direct shear failure without bending or axial compression while the flexural shear test induced a slipping between the two layers of concrete by bending the specimen.
4.2 Direct Shear Test
As explained in the background section of this thesis, Helmick et al. researched the applicability of a modified Iosipescu method of testing when applied to plain monolithic concrete (Helmick et al. 2016). To take this one step further, the following summarizes a similar test conducted on two-layered concrete beams rather than plain monolithic beams in order to test the shear strength at the bonded interface rather than the concretes monolithic shear strength.
The goal of this test was to understand if the Iosipescu method can be appropriately applied to concrete bond testing, and eventually lead to a formalized standard test to be used for concrete repairs and overlays. In the past, the shear strength of a bonded interface has been tested in many ways, a few of which have been discussed in the introduction. But the authors literature review revealed no previous test done in this manner to test a bonds shear strength. Additionally, the Iosipescu method has been used to test the shear stress of composite materials and plain monolithic concrete, but not for two-layered concrete beams, as far as is known. If the test proves to be successful, this
12


would be an effective and simple way to effectively measure the shear strength of a bonded concrete interface.
4.2.1 Specimen Preparation
In order to conduct the tests, first the two-layered concrete beam specimens were constructed in the Structures Lab at the University of Colorado Denver (UCD). Forms were built from plywood and 2x6 Douglas Fir boards so twelve 6x6x21 inch beams could be cast vertically rather than horizontally (each 6x6 inch compartment was 21 inches tall). Both concrete pours used the same concrete mix, a 5000 psi sack-mix concrete, but the first had color pigment added to it in order to easily see the demarcation line. The forms were filled to the halfway mark, 10.5 inches high and consolidated with a pencil vibrator. Then the first placement was allowed to cure for 28 days. Once the 28 days had passed, two different surface treatments were applied to eight of the beams. Four of the beams were left smooth, four were bush hammered, and four were jackhammered. The International Concrete Repair Institute (ICRI) has a standardized measurement for the roughness of a surface: a concrete surface profile or CSP. This gage of roughness ranges from 1 to 10 with 1 being the smoothest and 10 being the most rough. The smooth surface created for this thesis compared to a CSP of 2 or 3, the bush hammered surface compared to a CSP of 6, and the jack hammered surface was significantly rougher than the max CSP of 10. The increasingly rough surfaces are shown in Figure 4.1.
Research regarding shear bond strength between old and new concrete done at Louisiana State University concluded that saturated surface dry (SSD) conditions resulted in significantly higher bond strengths when compared to air dry conditions (Wan 2010). Immediately prior to the second placement of concrete, the surfaces were cleaned and
13


prepared. Then water was sprayed into each compartment and sponged off in order to accomplish these ideal SSD conditions.
Figure 4.1. Iosipescu beam specimen surface treatments. Smooth, bush hammered, and jackhammered surface preparations on the substrate. View is looking down into 6x6 inch forms at the prepared surfaces of the first pour. A second pour followed. Both pours were compacted by vibration.
Next, the second pour was placed, filling the forms the rest of the way to the top. Figure 4.2 shows a cross-section of a beam specimen constructed in the form.
Figure 4.2. Construction of typical Iosipescu beam specimen.
14


Like the first placement, the concrete was consolidated by vibration after the second placement. Then the beams cured for another 28 days. Once the second 28-day mark had been reached, each beam specimen was removed from the form and individually tested. Additionally, multiple cylinders were taken from each placement in order to test and confirm that the anticipated compressive strength of the concrete had been reached.
4.2.2 Testing
For each test, a beam was set-up in the Iosipescu loading scheme shown in Figure
4.3 and placed into a 200 kip capacity MTS load frame shown in Figure 4.4.
Figure 4.3. Iosipescu loading scheme.
15


Figure 4.4. Photograph of specimen in the Iosipescu apparatus. The wood blocks are for temporary support and are not part of the test.
A number of small steel parts were required for the Iosipescu test and will be described from the bottom up. First a 6x6x'/2 inch steel plate was placed on the bottom platform of the MTS load frame just to protect the equipment. Next a 1-inch diameter round rod 6 inches long was placed in the center of the plate in order to create a single point load that would act from below the beam specimen. Two lxl inch square bars 6 inches long were adhered with epoxy to a plate that was 1 inch thick, 6 inches deep, and 14 inches long. They were positioned on the plate with 8 inches measured center to center and 3 inches from each center to the edge of the plate. This assembly was balanced with the square rods facing upward on the 1 inch diameter round rod so that the centerline of the rod
16


lined up with the inside edge of one of the square rods. In order to accomplish this, a wood block was cut to the appropriate length to support the other side of the plate and keep the surface flat. Then the beam was placed onto the square rods so that the end of the beam lined up with the end of the plate below. Again, a wood block was placed under the other side of the beam for support. An identical assembly of the plate and the two square bars was then placed on top of the beam anti-symmetrically, mirroring how the bottom was placed but offset in the opposite direction with the inside of the other square rod lined up with the centerline. Next a 1-inch diameter round rod 6 inches long was placed along that same centerline to create a single point load that would act from the top with another 6x6x'/2 inch plate above to again protect the top platform of the MTS load frame.
Care was taken in order to assure everything was concentrically placed. Then a compressive force was applied under displacement control at a constant rate of travel while a load cell measured and a computer recorded three parameters: force, time, and displacement. Each beam was loaded until it failed in shear at which point the test was complete. As-built dimensions were also measured prior to each test. It should be noted that as the test was being conducted, the ram at the bottom would rise so that the set-up would no longer be supported by the wood blocks. Therefore they did not influence the test in any way; the set-up would remain in place due to the compression of the MTS load frame.
4.3 Flexural Shear Test
Six beams with two layers of concrete were constructed and tested in order to determine if the method created is appropriate to evaluate the flexural shear strength of a
17


bonded interface. Three different surface treatments were utilized between the two layers of concrete. The six beams were tested in flexure in two ways using a three point testing method. The first way was a typical simply supported beam: a support would be provided underneath both ends of a beam while a downward force was applied to the top center of the beam to induce flexure, referred to herein as flexure induced by compression. The second way was the reverse of this. Steel pipes placed strategically between the layers during concrete placement were used to anchor each end of a beam while an upward force was applied to the center of the beam to induce flexure, referred to herein as flexure induced by tension. Figure 4.5 illustrates the basic forces applied for these two methods.
/IS
Figure 4.5. Flexure induced by shear at interface combined with compression (clamping action ) (left) and by shear at interface combined with tension (unclamping action) (right).
In theory, the beam specimens tested using compression will have a higher flexural shear strength than those tested using tension; the compression will add a clamping force making it more difficult for the layers to debond from one another while the tension will help facilitate the separation between layers. The following summarizes details of the construction and testing of the beams.
18


4.3.1 Design
First, the test had to be designed. Dimensions of the beam specimens were determined with the help of an Excel spreadsheet. Setting some basic parameters was the starting point. Based on prior shear stress tests done by Rosen at UCD (Rosen, 2016), the target shear stress was set at 500 psi. For the beams to be considered narrow, the minimum span to depth ratio was set to 4. The spreadsheet was set up to calculate the flexural shear stress, the moment, the required reinforcement, and the shear strength of the beam from an input of dimensions and forces. Trial and error was used and the dimensions and forces were adjusted until optimal calculated results were reached. The goal was to get each value as close to the target value as possible. Table 4.1 shows the final values for each target value, design value, dimension, and calculation. For dimensions, L represents the length of the beam, h represents the height of the beam, d represents the distance from the top of the beam to the centroid of the reinforcement, b represents the width of the beam, and t represents the width of the contact area between the two layers of concrete.
Table 4.1. Flexural shear test design parameters.
Dimensions Span/Depth Ratio Flexural Shear Stress Predicted Forces Calculated Design Shear Strength Reinforcement
L = 7 ft Fi = 11.25 in Minimum Taraet Midspan 6.68 kips Reauired As 0.583 in2
4 500 psi 14 kips],
d = 10.5 in Applied Moment Provided As
b = 4.5 in Desian Desian Supports 0.6 in2
7.47 467 psi 7 kips) 24.5 k-ft
t = 2 in [w/ (3) #4 Bars]
19


In order to ensure that the desired type of failure would occur, a bond breaker was used at the interface to decrease the surface area of the connection. Again, the width of the surface area at the interface is represented by t in Table 4.1. Figure 4.6 shows a cross section of a beam specimen with the dimensions labeled.
Figure 4.6. Cross-section of typical beam specimen for flexure tests. The bond breaker completely separates the two layers of concrete at the ends and center of the beam (left) and only partially separates the two layers in the middle of each half to create the influence areas illustrated later in Figure 4.11.
4.3.2 Specimen Preparation
Next the beam specimens were created. Forms were constructed with a plywood base and 2x12 wood boards. Since the beams were to be 11.25 inches in height, it was convenient to use 2x12s because the actual width of a 2x12 is 11.25 inches. Seven boards 7.5 feet in length were nailed to the plywood. Because the design length was set at 7 feet, each end was extended an additional 3 inches to allow for a support to be placed under each end during testing. This made the actual length of each beam 7.5 feet. The boards were attached along their narrow edge for the length of the plywood to create six
20


compartments 11.25 inches high and 4.5 inches wide with a board across each end. As previously mentioned, it was important that the beams failed in the correct manner: flexural shear rather than diagonal tension or flexure. To make sure they didnt fail in flexure, reinforcement was utilized: three No. 4 Grade 60 reinforcing bars were to be used in each beam. Next, the rebar had to be cut and the ends bent into 180 hooks in order to achieve a suitable development length and fit into the dimensions of the beams properly. Once all the rebar had been cut and bent, they were connected together with tie wire in groups of three as seen in Figure 4.7 and placed into the forms.
Figure 4.7. Reinforcing steel used in each beam specimen.
Small strips of extra rebar were attached to the bottom of each group of three as spacers so that the rebar didnt rest on the bottom of the form, thus providing some clear space.
The three beams that were to be tested in flexure using tension required a method to anchor the two ends to the ground. As a result, steel pipe VA inch in diameter was cut into 4.5 inch segments and was wedged into the end of three of the compartments, 3 inches from each end to the pipes center, so that the top of the pipe would be flush with the top of the first layer of concrete. Further details regarding the purpose of the pipes
21


will be explained later. Once this had been done, the forms seen in Figure 4.8 were ready for the first concrete placement.
Figure 4.8. Concrete form ready for first placement.
The beams were then labeled 1-6. Beams 1-3 were to be tested in flexure using compression and Beams 4-6 (those with the steel pipe segments) were to be tested in flexure using tension. At this point, it was time to place the first layer of concrete. The test utilized a 5000-psi sack-mix concrete and color pigment was added in order to clearly see the demarcation line between the two layers of concrete. The forms were filled to the half way mark 5.625 inches high. The concrete in each compartment was then
22


consolidated using a pencil vibrator to expel any air pockets and to assure the concrete filled each form completely, particularly under the rebar and around the steel pipe segments. Before the concrete was allowed to set, a screwdriver was used to create a rake-type surface treatment to two of the six beams, one of the beams with pipes (Beam 4) and one without (Beam 1) as can be seen in Figure 4.9.
Figure 4.9. Application of raked surface treatment to substrates.
Then the concrete was allowed to cure for 28 days. After 28 days, two more beams received surface treatments. A jackhammer with a bush bit was used on one of the beams with pipes (Beam 5) and one without (Beam 2). This provided three different surface treatments to be tested: a smooth finish (Beams 3 and 6), a bush hammered finish (Beams 2 and 5), and a raked finish (Beams 1 and 4) seen in Figure 4.10. Table 4.2 outlines the surface treatment of each beam specimen and the way in which they were to be tested.
23


Figure 4.10. Flexural beam specimen surface treatments: bush hammered (top), raked (middle), and smooth (bottom).
Table 4.2. Summary of beams, surface types, and testing methods used.
BEAM NUMBER SURFACE TYPE TESTING METHOD
1 RAKED COMPRESSION
2 BUSH HAMMERED COMPRESSION
3 SMOOTH COMPRESSION
4 RAKED TENSION
5 BUSH HAMMERED TENSION
6 SMOOTH TENSION
As previously mentioned, a bond breaker was needed to encourage slippage between the two layers; this reduced the surface area of the connection between the two layers of concrete. This slippage was the flexural shear failure being tested. Plastic wall paneling 2 mm thick was cut the same length and width of each beam to create the bond breaker. Influence lengths were calculated based on using 6-inch wide supports at each end and a 4.5x4.5 inch plate in the middle of the beam under the jack, see Figure 4.11.
24


ELEVATION
PLAN
T~
4.5" * . 4 Z 4 * <
J
> / 11.625" J*- 25.5 3 < 15.75" j l- 25.5' 1 11.625" 3'
Figure 4.11. Influence areas created with the bond breaker.


Two square holes 2 inches wide (the t dimension in Table 4.1) and the influence length long were cut into each strip of the bond breaker to create the areas where the two layers of concrete would have direct contact. This area was the bonded interface being tested. The interface was then exposed to the flexural shear force created from the bending of the beams. The bond breakers were set into each compartment of the form and secured with some duct tape, as can be seen in Figure 4.12.
Figure 4.12. Form prior to second concrete placement. The white bond breaker is PVC wall paneling sheet.
26


Unfortunately, the rake marks previously created on two of the beams extended the whole width of the beam for the entire length of the beam. This meant that there were holes for the concrete to flow into which defeated the purpose of the bond breaker. To eliminate this possibility, Plaster of Paris was used to fill in the rake marks that were located directly under the plastic bond breaker. Once the bond breakers had been set and secured in the forms, it was time for the second placement of concrete.
Immediately prior to the second placement of concrete, the surfaces were cleaned and water was sprayed into each compartment and sponged off to achieve the ideal SSD conditions previously discussed for the direct shear test. Additionally, one 4.5-inch segment of the VA inch diameter steel pipe was set directly on the bond breaker in the center of Beams 4, 5, and 6. These would later be used to attach the middle of the beam to the tension testing machine. The three beams that were to be tested in tension, Beams 4, 5, and 6 needed to have rebar in the tops of the beams rather than the bottoms, so rebar was cut, bent, and tied together in the same manner as before.
The next layer of concrete was then placed without the added color pigment. The rebar was inserted into the fresh concrete on Beams 4, 5, and 6 so that the tops of the spacer bars were flush with the top of the concrete. Then the concrete was consolidated again with the pencil vibrator and allowed to cure for another 28 days. Multiple cylinders were taken from each concrete placement in order to test and confirm that the anticipated compressive strength was reached within 28 days.
4.3.3 Testing
Each beam specimen was removed from the form and as-built dimensions were measured. The goal was to induce a slipping failure between the two layers of concrete.
27


Beams 1, 2, and 3 were tested in compression while Beams 4, 5, and 6 were tested using tension, as previously described in this chapter.
4,3.3.1 Flexure by Compression Testing
The set-up for the compression testing was comprised of a simply supported beam with a single downward force acting at midspan. The lab at UCD has P/t-inch diameter round threaded metal posts of different lengths that can be secured to a strong floor by screwing them into threaded holes set 3-feet apart center-to-center in a grid throughout the lab. Two steel channels were secured together to form an I-shaped beam that was mounted on two of the round metal posts about 3 feet in the air. Then a plate was attached to the bottom of the two channels at the center of the beam with C-clamps in order to create a platform for the jack to press against. The purpose of the apparatus was to have something for the jack to brace itself against when applying the downward force onto the center of each beam; the jack would be sandwiched between the I-beam and the beam specimen and then pumped to expand. Two 3-foot long 4x6 Douglas Fir wood posts were placed on either side of the I-beam/post set-up with the 6-inch side on the ground, 7 feet apart center-to-center. These would serve as the supports for each beam and allow for deflection to occur. Figures 4.13 and 4.14 show the design used for this test.
28


to
VO
Figure 4.13. Design of flexure by compression test, elevation view.


At this point, it was time to test Beams 1, 2, and 3. To begin each test a beam specimen was placed under the center of the I-beam onto the wood supports. The jack used was a Simplex RC306C with an area of 6.44 square inches. The jack was placed between the I-beam and the beam specimen at its center. Figure 4.15 illustrates the complete set-up of the flexure by compression test.
30


Figure 4.15. Flexure by compression testing set-up.
By pumping the jack at a slow rate, a constant downward force was applied to the beams midspan until failure. The maximum pressure of the jack was recorded and the beams fracture was observed and analyzed.
4,3,3,2 Flexure by Tension Testing
The set-up for the tension test was a bit more elaborate than that for the compression test but the concept was similar: anchor the two ends of the beam and apply a single upward force to the midspan. A 50 kip capacity load frame was used, pictured in Figure 4.16.
31
UJ


Figure 4.16. 50 kip capacity load frame used for the flexure by tension tests.
First it was necessary to anchor the two ends of the beam to the ground so it would be centered under the machine. A 2-inch diameter hole was drilled through each end of a 7.5-foot long steel 4x4x'/4 inch HSS (hollow structural section) tube with 6 feet measured center-to-center. These holes allowed the HSS tube to be mounted on two short threaded metal posts screwed into two of the threaded holes in the lab 6 feet apart centered under the machine. Next, hollow concrete blocks were placed on either end of the HSS tube for each beam specimen to rest upon prior to testing as seen in Figure 4.17.
32


Figure 4.17. HSS secured to the floor used to anchor the ends of the beam specimens.
Once a beam was set into position, it had to be anchored to the HSS tube and connected to the machines arm.
The l'A-inch steel pipe segments placed in the top of each end of Beams 4, 5, and 6 during the first concrete placement were used to anchor the beam specimens to the HSS tube. Other materials used to anchor the beam ends included two 10-inch segments of a 1-inch diameter all-treaded round rod with two nuts for each, four '/2-inch eyebolts with couplers, two 10x5x3/4 inch steel plates, and extra '/2-inch all-threaded rod cut appropriately with four nuts total. On each end, the 1-inch diameter all-threaded rod was threaded through the pipe embedded in the top of the first layer of concrete. An eyebolt was put onto each end of the rod so that the beam was sandwiched in-between and
33


secured with nuts. The extra '/2-inch threaded rod was cut appropriately and attached to each eyebolt with a coupler so when they hung down they extended an inch or two below the HSS tube. Holes were drilled into the two 10x5x3/4 inch plates 5 inches apart center-to-center so the '/2-inch threaded rod could pass through them. The plates were held under the HSS tube and the threaded rod was threaded through them and secured with nuts.
This allowed each end of the beam to be anchored to the HSS tube, which was anchored to the ground.
The 1'/4-inch steel pipe segments placed in the middle of Beams 4, 5, and 6 on the bottom of the second concrete placement were used to connect the beam specimens to the arm of the machine. Other materials used to make this connection included a 10-inch segment of the same 1-inch diameter all-threaded round rod with two nuts, a 8x8x3/4 inch steel plate, four 5/8-inch diameter bolts with nuts, two 3/4-inch eyebolts with couplers, and extra 3/4-inch all-threaded rod with two nuts. The configuration at the center connecting the beam to the machine was similar to the configuration at each end. The 8x8x3/4 inch plate had six holes drilled into it; four were used to attach the plate to the machine and two were used to attach the beam specimens to the plate. The plate was attached to the machines arm with the bolts and nuts. Next the threaded rod was threaded through the pipe at the center of the beam being tested and the eyebolts were secured onto each end with nuts. The extra '/2-inch all-threaded rod was cut appropriately, connected to each eyebolt with the couplers, inserted through the two other holes on the plate and secured with nuts. Figures 4.18 and 4.19 show the design used for this test
34


13/16" 0 HOLE
Figure 4.18. Design of flexure by tension test, elevation view.


I
SECTION 1 SECTION 2 SECTION 3
Figure 4.19. Design for flexure by tension test, section cuts.


Once a beam specimen was set into place, it was time for the testing to begin. Figure 4.20 illustrates the complete set-up of the flexure by tension test.
Figure 4.20. Flexure by tension testing set-up.
The machines arm applied an upward force to the center of the beam at a slow, constant rate until failure occurred and the two layers debonded from one another. During the test a computer recorded force, displacement, and time and once the failure occurred, the fracture was observed and analyzed.
While getting the second beam specimen (Beam 5) set into place, some unanticipated separation of the layers occurred at each end early during the testing. This was addressed by modifying the way in which the beam was anchored to the HSS tube. Longer segments of threaded rod were used to connect the plate located underneath the
37


HSS tube to another additional plate placed on top of the beam so that both the HSS tube and beam were being sandwiched or clamped together with the plates and the threaded rod. This modification can be seen in Figure 4.21. However, since the bond breaker was solid at the ends, it is believed that this did not influence the results acquired for this beam.
Figure 4.21. Modified testing set-up for Beam 5.
4.4 Compressive Strength Testing
During the first and second placement of concrete, three samples were taken from each batch in order to test the compressive strength of the mix. The samples were placed into cylindrical molds 4 inches in diameter and 8 inches tall. This was done in three lifts
38


with consolidation occurring between using a metal rod. Then the cylinders were set aside and allowed to cure. Three days after placement, one cylinder from each batch was tested. Testing was done again on day seven and day twenty-eight in order to develop a compressive strength curve for each batch. For each compression test, the sample cylinder was removed from the mold, placed between two neoprene pad caps, and centered in a Forney compression testing machine as pictured in Figure 4.22. A compressive force was applied at a slow, constant rate until failure at which time the maximum force was recorded. Figures showing the compressive strength curves developed can be found in the following chapter.
Figure 4.22. Compressive testing machine used.
39


CHAPTER V
RESULTS
5.1 Introduction
Once all the testing was completed, data from the shear tests conducted for this thesis was compiled. The results of each test are presented in this chapter.
5.2 Compressive Strength Testing
Data collected from the compressive strength tests was entered into an Excel spreadsheet. A curve was fit to each set of data in order to determine the compressive strength if c) of the concrete at any specified time after placement. Figure 5.1 shows the curves associated with the first placement of concrete and Figure 5.2 shows the curves associated with the second placement.
CYLINDER TESTING FIRST PLACEMENT
---BEAM 3/4 -BEAM 2/4 -BEAM 1
8000
Figure 5.1. Compressive strength curves from first concrete placement.
40


CYLINDER TESTING SECOND PLACEMENT
--10 BEAMS -BEAM 1/2 -BEAM 4/5
6000
Figure 5.2. Compressive strength curves from second concrete placement.
From the curves, an average compressive strength was calculated for each placement. The lower of the two governs and is used to express the results of the other tests as a function of the compressive strength of the concrete; the shear strength is equal to a factor multiplied by the square root of the compressive strength of the concrete.
5.3 Direct Shear Test
As expected, each direct shear test specimen successfully fractured along the bonded surface once the force from the MTS load frame overcame the bond strength. An example of this can be seen in Figure 5.3. The beams with smooth and bush hammered surface treatments had extremely clean fracture faces between the layers while the jackhammered beams had fractures that varied a bit more.
41


Figure 5.3. Iosipescu beam specimen at failure.
Once the data had been compiled, it required little manipulation in order to extract the desired information. In other applications of this method, researchers have made shallow
notches on the top and bottom of the beam close to the centerline to induce the fracture from one notch to the other. Because of this, the fracture occurred at a slight diagonal angle between the top and bottom inside square bars in which case the surface area of the fractured face was slightly larger than the cross section of the beam. However, in this case, there already was a weakened plane at the bonded interface so notches were not necessary and the fracture occurred at the bonded cross-section of the beam. No new surface area calculations or decomposition of forces were necessary. The shear stress was calculated dividing the force parallel to the plane (the shear force from the MTS load frame) by the surface area of the failure plane (the cross-sectional area of the beam). The data is graphically represented in Figures 5.4-5.9. Each graph shows the results of the
42


four beams for each surface treatment and is presented with absolute values and values expressed as a function of the square root of the compressive strength of the concrete,
V/",.
SMOOTH INTERFACE SURFACE
-----SI -------S2 -------S3 -------S4
Figure 5.4. Shear stresses measuredfor the smooth interface.
BUSH HAMMER INTERFACE SURFACE
600
Figure 5.5. Shear stresses measuredfor the bush hammered interface.
43


JACKHAMMER INTERFACE SURFACE
J1 --------J2 ------ J3 ---------J4
Figure 5.6. Shear stresses measuredfor the jackhammered interface.
SMOOTH INTERFACE SURFACE
8 7 6 5 4
A 3 2 1 0 -1
Figure 5.7. Shear stresses measuredfor the smooth interface expressed as a function of
'i(U
44


BUSH HAMMER INTERFACE SURFACE
----B1 ------B2 -----B3 ------B4
Figure 5.8. Shear stresses measured for the bush hammered interface expressed as a function of V(fc)
JACKHAMMER INTERFACE SURFACE
8
-----J1 --------J2 --------J3 --------J4
Figure 5.9. Shear stresses measuredfor the jackhammered interface expressed as a function of V(fc)
It can be seen from the graphs that the outcomes for each surface type are fairly consistent. Finally, the maximum shear stress for each of the twelve beams was determined. The average maximum shear stress was calculated for each surface type and
45


has been summarized in Tables 5.1 and 5.2. Again, the first table presents absolute values and the second tables values are expressed as a function of the square root of the compressive strength of the concrete, V/V
Table 5.1. Summary of shear stresses in psi.
SMOOTH BUSH HAMMERED JACKHAMMERED
BEAM 1 272 467 518
BEAM 2 409 278 455
BEAM 3 281 389 274
BEAM 4 251 280 370
AVERAGE 303 354 404
COV 24% 26% 26%
Table 5.2. Summary of t \fc values.
SMOOTH BUSH HAMMERED JACKHAMMERED
BEAM 1 3.83 6.60 7.31
BEAM 2 5.78 3.93 6.42
BEAM 3 3.97 5.49 3.86
BEAM 4 3.55 3.96 5.22
AVERAGE 4.28 4.99 5.70
COV 24% 26% 26%
5.4 Flexural Shear Test
Results from the flexural shear tests are presented in Tables 5.3 and 5.4 and Figures 5.10 and 5.11. Each table and figure is presented with absolute values and values expressed as a function of the square root of the compressive strength of the concrete,
Vf c. Due to a machine error, no data was collected for Beam 4.
46


Table 5.3. Flexural shear test results: Beams 1-3.
BEAM NUMBER BEAM 1 BEAM 2 BEAM 3
SURFACE TREATMENT RAKED BUSH HAMMERED SMOOTH
TESTING METHOD COMPRESSION COMPRESSION COMPRESSION
MAX PRESSURE (psi) 2800 2400 2900
MAX FORCE (kips) 18.032 15.456 18.676
MAX SHEAR (kips) 9.016 7.728 9.338
Z b o 4.75 4.25 4.75
HH Z h W 11 11.25 11
s HH 4. Q t 2 2 2
Q = (thA2)/8 30.25 31.64 30.25
I = (thA3)/12 221.83 237.30 221.83
MAX SHEAR STRESS
t = VQ/It = 3V/(2A) (psi) 615 515 637
T/Vf'C 8.68 7.27 8.99
47


Table 5.4. Flexural shear test results: Beams 4-6.
BEAM NUMBER BEAM 4 BEAM 5 BEAM 6
SURFACE TREATMENT RAKED BUSH HAMMERED SMOOTH
TESTING METHOD TENSION TENSION TENSION
MAX PRESSURE (psi) N/A N/A N/A
MAX FORCE (kips) NO DATA COLLECTED 5.008 3.814
MAX SHEAR (kips) NO DATA COLLECTED 2.504 1.907
Z b 4.50 4.75 4.50
o
HH Z h W
11.25 11.25 11
% HH 4- Q t 2 2 2
Q = (thA2)/8 31.64 31.64 30.25
I = (thA3)/12 237.30 237.30 221.83
MAX SHEAR STRESS t = VQ/(It) = 3V/(2A) (psi) NO DATA COLLECTED 167 130
T/Vf'C NO DATA COLLECTED 2.36 1.84
48


700
MAXIMUM SHEAR STRESS (psi)
600
500
400
300
200
100
0
BEAM 1 I BEAM 2 BEAM 3 BEAM 4 BEAM 5 BEAM 6
Figure 5.10. Maximum flexural shear stress of Beams 1-6.
10.00
9.00
8.00
7.00
6.00
5.00
4.00
3.00
2.00
1.00 0.00
MAXIMUM x/V(f c) VALUE
BEAM 1 I BEAM 2 BEAM 3 BEAM 4 BEAM 5 BEAM 6
Figure 5.11. Maximum r/^(fc) value.
49


CHAPTER VI
DISCUSSION
6.1 Introduction
As discussed in the introduction and background of this thesis, the engineering community is in need of a way to evaluate the shear strength capacity of a bonded interface between two layers of concrete. The methods previously outlined were tested and the results have been presented. The following contains an analysis of those results.
6.2 Direct Shear Test
As shown in Tables 5.1 and 5.2, the jackhammered interface had the highest shear stress capacity at 404 psi or 5.70V/c, the bush hammered interface had a slightly lower capacity at 354 psi or 4.99^fc, and the smooth interface had the lowest capacity at 303 psi or 4.28V/V This gradual decrease in strength correlates well with the decrease in surface roughness, and is the first good sign that the test was a success. Repeatability among the results for each surface type is the next encouraging sign that the test is applicable. The coefficient of variation for each set of data was around 25%.
Additionally, the numerical value of each average is also in general agreement with other shear tests done on similar surface treatments by Rosen (Rosen 2016) and Pultorak (Pultorak 2016) of UCD. Table 6.1 compares the results from the testing conducted for this thesis with Rosens and Pultoraks research. Since all three researchers tested a bush hammered surface, it is appropriate to make a direct comparison of the results associated with that surface treatment. The value obtained by this author of 354 psi lies half way between Rosens value of 294 psi and Pultoraks value of 416 psi. The numbers stray from one another a bit when expressed as a factor multiplied by the square root of the
50


compressive strength of the concrete, Vfc, because there was a range in those values. If the same concrete was used for each of the tests, perhaps these factors would have been closer to each other.
Table 6.1. Comparison of results from other researchers.
RESEARCHER TESTING METHOD SURFACE TREATMENT SHEAR STRESS GOVERNING f'c SHEAR AS FUNCTION OF SQRT(f'c)
SMOOTH 303 5016 3.55
SWAN IOSIPESCU BUSH HAMMER 354 5016 3.96
JACKHAMMER 404 5016 5.22
BROOM 496 5648 6.60
ROSEN GUILLOTINE BUSH HAMMER 294 5650 5.22
RAKE 457 3172 6.08
PULTORAK GUILLOTINE BUSH HAMMER 416 3069 7.50
Although the tests appear to have been successful, there are possible sources of error. Ideal beam specimens would have equal lengths of substrate and overlay concrete, but because the forms were filled half way and then bush hammered or jackhammered, some of the beams had larger portions of overlay concrete than substrate, particularly those that had been jackhammered. In the future, it is recommended to fill the compartments to be bush hammered and j ackhammered a bit more than halfway, especially the jack hammered beams; this would provide an allowance for some material loss during surface preparation. Additionally, the mechanical surface roughening techniques used may have bruised the top layer of the substrate concrete making the bond weaker than if other techniques were utilized. Because the forms were constructed from wood, warping of the boards created beam specimens that did not have the exact
51


same dimensions and cross-sections over their full length. During the testing, possible misalignment of the Iosipescu apparatus as well as possible non-concentric load application may have occurred, despite the care taken in positioning the parts of the apparatus. Even with these possible sources of error, it can be concluded that the results obtained support the belief that the modified Iosipescu method is an appropriate test to evaluate the direct shear capacity of concrete repairs and overlays with a bonded interface.
6.3 Flexural Shear Test
Unlike the direct shear test results the results from the flexural shear stress tests have excessive variation. In the case where compression was used to induce flexure, Beam 3 was the only one that experienced a true flexural shear failure, as seen in Figure
6.1.
Figure 6.1. Flexural shear stress failure of Beam 3.
52


While reinforcing steel precluded flexural failures, the other two beams (Beam 1 and 2) experienced a diagonal tension failure prior to a flexural shear failure. The fracture line tended to start by the jack, traveled diagonally down to the interface, then across the interface for a short distance, and then diagonally down to the reinforcement. This can be seen in Figure 6.2.
Figure 6.2. Diagonal tension failure of Beam 1.
Because two of the three beams failed in this manner, it seems as though the reinforcement may have created a weak plane where the concrete tended to separate from itself at that location rather than at the interface. Although the beams did not experience the desired type of failure, it can be concluded that in these two cases (Beams 1 and 2) the flexural shear strength of the bonded interface was higher than the diagonal tensile strength of the beam. This aspect of the test is in agreement with the varying surface treatments since the beam with the smooth surface treatment (Beam 3) experienced flexural shear failure, meaning the flexural shear strength of that interface was less than
53


the diagonal tensile strength of the beam; making its capacity the lowest. This was the expected result. This is not illustrated in the bar graphs presented because Beams 1 and 2 failed in diagonal tension prior to flexural shear. Thus, the graphs are not a good representation of the shear strength of each bonded interface.
As predicted, there is a significant difference in values between the beams tested in flexure using compression versus the beams tested in flexure using tension. This can be attributed to the clamping action incurred versus the facilitating pulling action discussed earlier.
Several problems arose when tension was used to induce flexure. Issues with the testing set-up prevented the author from acquiring any usable data for the first beam (Beam 4). The early separation of the layers at Beam 5s ends required a modification of the set up. Thus, the last beam (Beam 6) was the only one where testing went as planned. However, while observing the test, it appeared as though more of a tensile failure occurred at the interface rather than any bending of the beam and the top layer was simply lifted off of the bottom layer as pictured in Figure 6.3.
Because of all the complications and variability encountered during the flexural shear strength testing process, the author has concluded that this is not the best way to evaluate the flexural shear stress capacity of a bonded concrete interface.
54


Figure 6.3. Failure of Beam 6.
55


CHAPTER VII
CONCLUSION & RECOMMENDATIONS
Based on the results obtained, the author believes that the modified Iosipescu method of testing can be appropriately used to evaluate the direct shear strength capacity of different types of bonded concrete interfaces. However, the impediments encountered with the method used to evaluate the flexural shear strength capacity of a bonded interface make that test ineffective.
The flexural shear tests proved to be unsuccessful, particularly in the case of the flexure induced by tension because of the way the top layer lifted off of the bottom layer rather than sliding away from each other from the bending. However, with some alterations it is believed that the flexure induced by compression test could still be viable if future testing was conducted. If this is done, a few recommendations can be made. The size of the beams made it extremely difficult to maneuver and manipulate them around the lab for testing, so the use of smaller beams may be beneficial. A different reinforcement plan should also be utilized to ensure the proper type of failure occurs in all beam specimens. This type of test has been successfully used in the past by other researchers and may still be a good method, however the testing conducted as a part of this thesis was not successful.
More testing can and should be done to perfect the modified Iosipescu direct shear testing method presented. To further delve into the topic, other surface roughening techniques such as sand or water blasting or acid etching may be explored. Again, because the beam specimens were fairly large and cumbersome, the tests could be conducted with smaller beams to see if a lighter specimen would be appropriate. Helmick
56


et al. recommend using 4x4 inch beams, and this author is in agreement. In order to create identical and exact specimens it might be better to use steel forms rather than wood. The wood boards were warped which created varying sizes and shapes of the beams. If a bit more experimental testing is performed on the topic with the above recommendations taken into consideration, it is believed that this could be formalized into a standard test for concrete repairs and overlays.
57


REFERENCES
ACI Committee 562, Code Requirements for Assessment, Repair, and Rehabilitation of Existing Concrete Structures (ACI 562-16) and Commentary, American Concrete Institute, Farmington Hills, MI, 2016, pp 46-48.
Adams, D.F. (2009). A Comparison of Shear Test Methods. Composites World,
(Sept. 2016).
Adams, D.F.; Walrath, D.E., Current Status of the Iosipescu Shear Test Method,
Journal of Composite Materials, V. 21, February 1986, pp 494-507.
ASTM C882 (2015), Standard Test Method for Bond Strength of Epoxy-Resin Systems Used With Concrete By Slant Shear, American Society of Testing Materials, West Conshohocken, PA.
ASTM C1583 (2015), Standard Test Method for Tensile Strength of Concrete Surfaces and the Bond Strength or Tensile Strength of Concrete Repair and Overlay Materials by Direct Tension (Pull-off Method), American Society of Testing Materials, West Conshohocken, PA.
ASTM D5379 (2015), Standard Test Method for Shear Properties of Composite Materials by the V-Notched Beam Method, American Society of Testing Materials, West Conshohocken, PA.
Brewe, J.E.; Tumialan, J.G.; Kelley, P.L., Evolution of ACI 562 Code Part 7, Concrete International, V. 38, No. 9, September 2016, pp 56-60.
Craig, R.R. Jr., (2000), Stresses in Beams Mechanics of Materials 2nd Edition, John Wiley & Sons, Inc., Hoboken, NJ, pp 391-395.
Gohnert, M., Horizontal Shear Transfer Across a Roughened Surface, Cement and Concrete Composites, V. 25, No. 3, April 2003, pp 379-385.
Helmick, C.G.; Toker-Beeson, S.; Tanner, J.E., Evaluation of Shear and Diagonal
Tension in Plain Concrete, Concrete International, V. 38, No. 1, January 2016, pp 39-46.
Kovach, J.D.; Naito, C. (2008). Horizontal Shear Capacity of Composite Concrete Beams without Interface Ties, ATLSS Report No. 08-05.
Loov, R.E.; Patnaik, A.K., Horizontal Shear Strength of Composite Concrete Beams With a Rough Interface, Precast/Prestressed Concrete Institute, V. 39, No. 1, January 1994, pp 48-69.
58


Patnaik, A.K., Behavior of Composite Concrete Beams with Smooth Interface, Journal of Structural Engineering, V. 127, No. 4, April 2001, pp 359-366.
Pultorak, A.S. (2016). The Effects of Common Surface Pretreatments on the Shear Strength of Bonded Concrete Overlays. M.S. thesis, University of Colorado, Denver, Denver, CO.
Rosen, C.J. (2016). Shear Strength at the Interface of Bonded Concrete Overlays. M.S. thesis, University of Colorado, Denver, Denver, CO.
Sprinkel, M.M., Bond Strength between Shotcrete Overlay and Reinforced Concrete Base, Concrete Repair Bulletin, V. 29, No. 1, January/February 2016, pp 8-13.
Wan, Z., (2010) Interfacial Shear Bond Strength between Old and New Concrete, M.S. thesis, Louisiana State University, Baton Rouge, LA.
World Business Council for Sustainable Development. (2015). "Recycling Concrete Executive Summary." The Cement Sustainability Initiative (CSI), (Oct. 4, 2016).
59


APPENDIX
The following photographs illustrate the outcome of the tests performed for this thesis.
60


Direct Shear Test
61
.


Bush Hammered Interface:
62


Jack Hammered Interface:
63


Flexural Shear Test
Flexure Induced by Compression:
BEAM 1
BEAM 2
BEAM 3
64


Flexure Induced by Tension:
BEAM 4
BEAM 5
BEAM 6
65


Full Text

PAGE 1

ALTERNATE METHODS FOR TESTING SHEAR STRENGTH AT A BONDED CONCRETE INTERFACE by ANNE ELISE SWAN B.S., Colorado State University, 2011 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Civil Engineering Program 2016

PAGE 2

ii This thesis for the Master of Science degree by Anne Elise Swan has been approved for the Civil Engineering Program by Frederick Rutz, Chair Carnot Nogueira Peter Marxhausen Date: 17 December 2016

PAGE 3

iii Swan, Anne Elise (B .S., Civil Engineering) Alternate Methods for Testing Shear Strength at a Bonded Concrete Interface T hesis directed by Associate Professor Frederick Rutz ABSTRACT Two methods for testing shear strength were explored at the University of Colorado Den ver in order to determine if those methods would be appropriate to evaluate the shear strength at a bonded concrete interface. The first type of test is a direct shear stress test. Twelve beams with three different surface treatments between two layers of concrete were tested using a modified Iosipescu loading schem e until a slipping (direct shear) failure occurred at the interface. The second type of test is a three point flexural shear stress test. Six beams with three different surface treatments betw een two layers of concrete were tested using flex ure until a slipping (flexural shear) failure occurred at the interface In both tests, the two layers of concrete are identic al to one another apart from a difference in color. The success of the direct shear test makes that method a good candidate for a formalized standard to test the direct shear stress at a bonded concrete interface. If formalized into a standard test, this would be a relatively simple and effective way for the engineering community to test the shear strength of concrete for the application of repairs and overlays. However, t he results from the flexural shear stress test suggest the test method i s not effective for evaluating the shear stress at a bon ded concrete interface. The form and content of this abstract are approved. I recommend its publication. Approved: Frederick Rutz

PAGE 4

iv DEDICATION This thesis is dedicated to the incredibly supportive family and friends I am fortunate enough and Katie Swan, Nadim Chammas, Lindsey Reynolds, Alyson Heon, and my dog Titus. Thank you for always

PAGE 5

v ACKNOWLEDGEMENTS This thesis would not have been possible without the contr ibutions made by the following people: Tom Thuis & Peter Sillstrop Andy Pultorak & Christian Rosen Geotechnical Lab Brian Volmer Machine Shop Jac Corless & Brian Carpenter Kris Kemp & Nadim Chammas Most of all, thank and acknowledge Dr. Fred erick Rutz, my esteemed advisor, for funding the project and assisting me every step of the way.

PAGE 6

vi TABLE OF CONTENTS CHAPTER I. OVERVIEW 1 1.1 Introduction 1 1.2 Goal 3 1.3 Outline 3 II. BACKGROUND 4 2.1 Introduction 4 2.2 Common Bond Strength Tests 4 2.3 Direct Shear 5 2.4 Flexural Shear 7 III. CONTROLLING EQU ATIONS & SUPPORTING CALCULATIONS 10 3.1 Introduction 10 3.2 Direct Shear Test 10 3.3 Flexural Shear Test 10 IV. EXPERIMENT S 12 4.1 Introduction 12 4.2 Direct Shear Test 12 4.2.1 Specimen Preparation 13 4.2.2 Testing 15 4.3 Flexural Shear Test 17 4.3.1 Design 19 4.3.2 Specimen Preparation 20

PAGE 7

vii 4.3.3 Testing 27 4.3.3.1 Flexure by Compression Testing 2 8 4.3.3.2 Flexure by Tension Testing 31 4.4 Compressive Strength Testing 38 V. RESULTS 40 5.1 Introduction 40 5.2 Compressive Strength Testing 40 5.3 Direct Shear Test 41 5.4 Flexural Shear Test 46 VI. DISCUSSION 50 6.1 Introduction 50 6.2 Direct Shear Test 50 6.3 Flexural Shear Test 52 VII. CONCLUSION & RECOMMENDATIONS 56 REFERENCES 58 APPENDIX 60

PAGE 8

viii LIST OF TABLES TABLE 4.1 Flexural shear test design parameters 19 4.2 Summary of beams, surface treatments, and testing methods used 24 5 .1 Summary of shear stresses in psi 46 5.2 Summary of c ) values 46 5.3 Flexural shear test results: Beams 1 3 47 5.4 Flexural shear test results: Beams 4 6 48 6.1 Comparison of r esults from other researchers 51

PAGE 9

ix LIST OF FIGURES FIGURE 4.1 Iosipescu beam specimen s urface treatments 14 4.2 Construction of typical Iosipescu beam specimen 14 4.3 Iosipescu loading scheme 15 4.4 Photograph of specimen in the Iosipescu apparatus 16 4.5 Flexure induced by shear at interface combined with compression and by shear at interface combined with tension 18 4.6 Cross section of typical beam specimen for flexure tests 20 4.7 Reinforcing steel used in each beam specimen 2 1 4.8 Concrete form ready for first placement 22 4.9 23 4.10 Flexural beam specimen surface treatments 24 4.11 Influence areas created with the bond breaker 25 4.12 Form prior to second concrete placement 26 4.13 Design of flexure by compression test elevation view 29 4.14 Design of flexure by compression test, Section 1 30 4.15 Flexure by compression testing set up 31 4.16 50 kip capacity load frame used for the flexure by tension tests 32 4.17 HSS secured to the floor used to anchor the ends of the beam specimens 33 4. 18 Design for flexure by tension test elevation view 35 4.19 Design for flexure by tension test, section cuts 36 4.20 Flexure by tension testing set up 37 4.21 Modified testing s et up for Beam 5 38 4.22 Compressive testing machine used 39

PAGE 10

x 5.1 Compressive strength curves from first concrete placement 40 5.2 Compressive strength curves from second concrete placement 41 5.3 Iosipescu beam specimen at failure 42 5.4 Shear stresses measured for the smooth interface 43 5.5 Shear stresses measured for the bush hammered interface 43 5.6 Shear stresses measured for the jackhammered interface 44 5.7 Shear stresses measured for the smooth interface expressed as a function of c ) 44 5.8 Shear stresses measured for the bush hammered interface expressed as a function of c ) 45 5.9 Shear stresses measured for the jackhammered interface expressed as a function of c ) 45 5.10 Maximum flexural shear stress of Beams 1 6 49 5.11 Maximu m c ) values 4 9 6.1 Flexural s hear stress failure of Beam 3 52 6.2 Diagonal tension failure of Beam 1 53 6.3 Failure of Beam 6 55

PAGE 11

1 CHAPTER I OVERVIEW 1.1 Introduction Concrete is an extremely versatile construction material and can be used for a vast array of different types of structures. According to the World Business Council for Sustainable Development worl d than the total of all other building materials including woo d, steel, plastic, and aluminum (World Business Council for Sustainable Development 2015 ). Inevitably however, concrete can deteriorate over time or become damaged and in need of repair. Simply put, repairing concrete typically entails bonding new concrete to old concrete. Composite construction is accomplished when two or more materials are connected together so str ongly that they act as a monolithic unit. Composite construction is often neces sary to repair damaged structu ral members; the goal is to repair in a way that allows the member to act as it was originally designed. Monolithic behavior is desirable because it increases the strength and efficiency of a structure, or a member within a st ructure, which typically leads to a more economical design. In order for a composite unit to perform monolithically, the bonded interface must be capable of successfully transferring forces such as compression, tension, and shear. Simple and straightforwar d standard tests exist to measure the compressive and tensile strengths of bonds but when it comes to shear, few methods have been formalized into standard tests (Helmick et al. 2016) For this reason, engineers need an effective method to experimentally e valuate the shear strength at the bonded interface between new and old concrete for th e purpose of conducting repairs or applying overlays.

PAGE 12

2 There are three different types of shear: direct, flexural and torsional. The tests presented in this thesis w il l focus on direct shear and flexural shear ; torsional shear is not a part of the scope. Stress is defined as a force per unit area. Direct or general shear stress is the most basic and straight forward type and is simply the force applied divided by the cro ss sectional area of the material. Flexural, often called horizontal is more complex because it deals with the internal shear stress within a beam subject to bending. If flexural shear properties are to be measured, a beam test specimen made up of multiple layers has to be exposed to flexure ( bending ) Once a perpendicularly applied force overcomes the strength of the bond between the layers, the layers will separate from one anot her and s l ip This concept can be visualized by considering a thick phone book being bent in the middle; as the phone book goes from being flat to curved, the pages slide re lative to each other. If direct shear properties are to be measured, a test specime n has to directly experience uniformly distributed shear stress. This can be difficult to achieve because compressive or bending forces tend to be introduced in one way or another. Dr. Don ald F. Adams is the founder a nd president of Wyoming Test Fixtures, Inc ., a company that specializes in the design and fabrication of mechanical test fixt ures for composite materials; he has done extensive research on shear testing and the Iosipescu method. According to Dr. Adams test methods developed and used during the past 40 years than for any other type of mechanical test of composites, including tension and compres sion, the two other basic tests (Adams 2009 ) Yet few have been formalized into standard tests. This can be at tributed to the difficulty in developing a successful shear test method; many have been developed and used but

PAGE 13

3 few have been effective enough to be formalized as standards. Th is is not only true of bonded materials, but monolithic concrete as well. I t is d ifficult to apply many of the shear tests to concrete, and that difficulty is behind the lack of a formalized standard for testing the shear capacity of a bonded interface between two layers of concrete. 1.2 Goal The goal of this research is to examine and evaluate the applicability of two d iffere nt types of shear strength tests and to determine if the methods developed are appropriate to evaluate the shear strength of a bonded interface between two layers of concrete. Ideally, this would provide the eng ineering community with simple and effective standard tests to evaluate the shear capacity of a bonded interface. 1.3 Outline This thesis is made up of seven chapters. The first chapter is an overview and a description of the goal. Chapter 2 contains bac kground on the material including a literature review. Chapter 3 presents the controlling equations and s upporting calculations required. Chapter 4 explains the tests conducted. Chapter 5 presents the results of the tests. Chapter 6 analyzes the results of the tests. Chapter 7 contai ns a summary of conclusions drawn from the test results.

PAGE 14

4 CHAPTER II BACKGROUND 2.1 Introduction When dealing with plain concrete, the shear strength of a bonded interface comes mainly from adhesion and the mechanical interlock of the aggregate between a base or bottom layer, referred to herein as the substrate, and a top or second layer, referred to herein as the overlay. Because of this, it i s fairly intuitive that the rougher the surfa ce, the better this shear capacity will be. This is a phenomenon recognized by the International Concrete Repair Institute, Inc. ( ICRI ) There are many formalized standards to test different properties of a bond but very few for shear 2.2 Common Bond Strength Tests One of the most common tests performed on bonded concrete is the pull off test formally standardized as ASTM C1583 where a pure tensile force acts on a bond. A core is drilled through an overlay and at least 1 inch or half the core diameter into the substrate. With the use of an epoxy, a steel puck is attached to the overlay. Once the epoxy has set, the testing apparatus is attached to the puck and an upward force is applied until failure occurs ( ASTM C1583 2015 ). Other than a failure of the epoxy connection, there are three possible outcomes and each will indicate something regarding the tensile strength of the bond. This test can be done in the field and is very straightforward and easy to perfor m, hence its popularity. However there are not very many situations in the real world that will apply this type of force alone to a bond so other tests are necessary to truly evaluate the shear strength at a bonded interface

PAGE 15

5 The Arizona Slant Shear Test, commonly called the slant shear test, has been formalized as ASTM C882 to specifically determine epoxy bond strength. For this test, core samples are taken of the overlay and substrate at an angle so the bonded interface is at a diagonal. Once the ends ha ve been cut so that the core has flat top and bottom surfaces, a compressive force is applied at each end of the sample until failure occurs (ASTM C882 2015 ) Again, the different possible outcomes provide information specific to the strength of the bond b ut the clamping action that occurs makes it difficult to evaluate the shear strength alone without the influence of compression (Rosen 2016 ) Because of the clamping action and the angle at which the load is being applied relative to the interface surface, this is not a good test to evaluate direct shear strength. 2.3 Direct Shear It is difficult to design a test that measures direct shear strength alone. Some tests that have tried to do so include the push off test and the guillotine test. A push off test is h the use of a jack. Due to physical restrictions, it is practically impossible to do this without some eccentricity because the jack is placed on the overlay above the interface. This creates another situation where shear is not acting alone; bending mome nts are introduced which influence results. Double shear g uillotine tests are performed on core samples taken perpendicular to the surface so the bonded interface is straight across the center of the sample. The core is placed horizontally into a device so metimes called a guillotine shear jig. The jig is composed of two cradle type boxes. Half of the sample (a cored specimen) lays in one cradle while the other cradle lays on top of the other half of the sample. While the bottom is held in place a compressiv e force is applied to the top until the sample splits

PAGE 16

6 between the two pieces of the guillotine. Single shear guillotine tests sometimes called Iowa Tests have also been developed. T hese guillotine tests may be the closest the engineering community has achieved toward a direct and pure shear test ; however this method has not been formalized as a standard. One popular method used to test in plane shear properties of composites was originally developed by Nicolae Iosipescu of Romania in the early 1960s and was specifically used at that time for testing metals and other iso tropic materials by inducing a high shear stress at the location of interest. Over time, this method has been adapted to evaluate orthotropic and composi te materials (Adams and Walrath 198 6 ) Today, the modified method is fo rmalized as ASTM D5379 Properties of Composite Materials by the V formalized as a standard for testing shear in concrete (Helmick et al. 2016 ) In an effort to formalize the standard for concrete applications, a journal article of 2016 describes the success of using the Iosipescu loading scheme to test the she ar capacity of plain monolithic concrete (Helmick et al. 2016). The authors compared results for modified Iosipescu, flexural, split cylinder, and split prism tests and found that the modified Iosipescu test method provide d consistent results. Therefore th ey recommended hear strength of plain concrete (Helmick et al. 2016) Because they had such positive results using this method on plain monolithic concrete to determine its shear strength, this author thought it might also be applied to a two layered concrete beam to evaluate the shear

PAGE 17

7 strength of a bonded interface with t he changing variable being differing s urf ace treatments at the interface. 2.4 Flexural Shear When designing a concrete to concrete repair, i f the calculated demand for interface sh ear stress exceeds 60 psi, Sections 7.4.2 and 7.4.3 of ACI 562 Requirements for Assessment, Repair, and Rehabilitation of existing Concrete Structures require the use of interface reinforcement to transfer t he forces across the interface (ACI 562 2016). This 60 psi limit is deri ved from a nominal shear strength of 80 psi multiplied by a strength reduction factor of 0.75. This is based on which are familiar to engineers But accordi ng to research done in the past thi s Therefore it does not represent the actual beh avior of concrete repairs. In 1994, Loov and Patnaik presented an article en Horizontal Shear Strength outlining the testing they did on sixteen simply supported beams with a point load applied at the center. Although their research involved the use of stirrups, they determined that the stirrups were not stressed until a horizontal shear stress of about 220 290 psi was reached which suggests that as the strength of the interface without reinforcement. Not until about 430 psi did the (Loov and Patnaik 1994). Seven years later, Patnaik extended the research to include be ams with a smooth interface and published an article that describes their behavior. T he horizontal shear strength at failure ranged from

PAGE 18

8 200 900 psi but this research was done using interface reinforcement as before with the addition of varying clamping f orces thus the large range (Patnaik 2001). In 2003 horizontal shear strength as a function of the surface roughness. The curve was formed from horizontal shear strength test results rangin g from 70 300 psi. The lower numbers represent beams that had a smooth interface surface and the higher numbers represent beams that had a rough interface surface. Gohnert then tested six additional beams bent in flexure and found the experimental predictions were conservative in every case (Gohnert 2003). ATLSS (Advanced Technology for Large Structural Systems) Report No. 08 05 published in Five point and two point loading tests were conducted on multiple beam specimens and the horizontal shear stress results ranged from 475 1000 ps i. To conclude the research, the report recommends using horizontal shear capacities of 435, 465, and 570 psi for composite concrete sections with a broomed, as placed, and raked interface surface treatments, respectively (Ko vach and Naito 2008). The numbers from the above mentioned research suggest a typical horizontal shear strength capacity of a bonded concrete interface ranging from about 200 600 psi, depe nding on the surface treatment ( well above the 60 psi code provision ) Because of the use of various testing methods and differing surface treatments, it is difficult to determine the actual horizontal shear capacity of a bonded concrete interface without

PAGE 19

9 reinforcement but it certainly suggests a value greater than the 60 psi limit used in practice currently. The experimentation presented in this thesis is meant to build upon the research conducted by others in order to explore test methods that might lead to a standardized test.

PAGE 20

10 CHAPTER III CONTROLLING EQUATIONS & SUPPORTING CALCULATIONS 3.1 Introduction In order to develop shear strength tests that may be formalized into standards, it is important to understand the concept of shear as related to the mechanics of a material. The following outlines the controlling equations and supporting calculations required for the design and evaluation of the shear strength tests conducted. 3.2 Direct Shear Test Stress and strain are the two basic parameters used when defining the mechanics of a material. When a material is loa ded with a force, it produces a stress. Strain is the response of a system to that applied stress. There are three types of stress: tensile, compressive, and shearing. Tensile and co mpressive stresses occur normal to a plane and are while sheari ng stress occurs parallel to a plan e and i The direct shear test conducted for this thesis only requires the simple equation used to express the shear stress of a material: Where: shear stress (psi, N/m 2 ) shear force (lb, N) surface area parallel to the force (in 2 mm 2 ) 3.3 Flexural Shear Test As mentioned in the introduction, flexural shear stress is more complex than direct shear stress because it involves the internal stresses of a beam caused by flexure

PAGE 21

11 ( bending ) To understand flexural shear stress, first it is important to understand a term describes the lateral (transverse) or longitudinal shear load applied at t beam induced by an externally applied force. Shear flow (q) is defined by the following equation: Where: shear flow (lb/in, N/m) shear force at the section (lb, N) first moment of area (in 3 m m 3 ) for a rectangular beam moment of inertia (in 4 mm 4 ) for a rectangular beam width (b) of the beam supporting the stress: Where: width of cross section at location of interest (in, m) This equation for shear stress can be used for narrow beams. A beam is considered narrow when the ratio of b/h is small, typically equal to or less than 0.25. If a beam is narrow and has a rectangular cross section the above equation can be simplified in the following way:

PAGE 22

12 CHAPTER IV EXPERIMENT S 4.1 Introduction Two different experiments were designed to evaluate the shear str ength at a bonded concrete interface. The direct shear test used a modified Iosipescu method to induce a direct shear failure without bending or axial compression while the flexural shear t est induced a slipping between the two layers of concrete by bending the specimen. 4.2 Direct Shear Test As explained in the background section of this thesis, Helmick et al. researched the applicability of a modified Iosipescu method of testing when applied to plain monolithic concrete (Helmick et al. 2016) To take this one step further, the following summarizes a similar test conducted on two layered concrete beams rather than plain monolithic beams in order to test the shear strength at the bonded interface rather than the th. The goal of this test was to understand if the Iosipescu method can be appropriately applied to concrete bond testing and eventually lead to a formalized standard test to be used for concrete repairs and overlays. In the past, the shear strength of a bonded interface has been tested in many ways, a few of which have been discussed in the introdu ction. B terature review revealed no previous test done in this Additionally, the Iosipescu method has been used to test the shear stress of composite materials and plain monolithic concrete but not for two layered concrete beams, as far as is know n If the test proves to be successful, this

PAGE 23

13 would be an effective and simple way to effectively measure the shear strength of a bonded concrete interface. 4.2.1 Specimen Preparation In order to conduct the tests, first the two lay ered concrete beam specimens were constructed in the Structures Lab at the University of Colorado Denver (UCD). Forms were built from plyw ood and 2x6 Douglas Fir boards so twelve 6x6x21 inch beams could be cast vertically rather than horizontally (each 6x6 inch compartment was 21 inches tall). Both concrete pours used the same concrete mix, a 5000 psi sack mix concrete, but the first had color pigment added to it in order to easily see the demarcation line. The forms were filled to the halfway mark, 10.5 in ches high and consolidated with a penc il vibrator. Then the first placement was allowed to cure for 28 days. Once the 28 days had passed, two different surface treatments were applied to eight of the beams. Four of the beams were left smooth, four were bus h hammered, and four were jackhammered. The International Concrete Repair Institute (ICRI) has a standardized measure ment for the roughness of a surface : a concrete surface profile or CSP This gage of roughness range s from 1 to 10 with 1 being the smoothest and 10 being the most rough The smooth surface created for this thesis compared to a CSP of 2 or 3, th e bush hammered surface compared to a CSP of 6, and the jack hammered surface was significantly rougher than the max CSP of 10. The increasingly rough s urfaces are shown in Figure 4.1 Research regarding shear bond strength between old and new concrete done at Louisiana State University concluded that saturated surface dry (SSD) conditions resulted in significantly higher bond strengths when compared to air dry conditions (Wan 2010). I mmediately prior to the second placement of concrete, the surfaces were cleaned and

PAGE 24

14 prepared. Then w ater was sprayed into each compartment and sponged off in order to accomplish the se i deal SSD conditions. Figure 4 .1. Iosipescu beam specimen s urface treatments. Smooth, bush hammered, and jackhammered surface preparations on the substra te. View is looking down into 6x6 inch forms at the prepared surfaces of the first pour. A second pour followed. Both pours were compacted by vibration. Next, the second pour was placed, filling the forms the rest of the way to the top. Figure 4.2 shows a cross section of a beam sp ecimen const ructed in the form. Figure 4. 2 Co nstruction of typical Iosipescu b eam specimen

PAGE 25

15 Like the first placement, the concrete was consolidated by vibration after the second placement. Then the beams cured for another 28 days. Once the second 28 day mark had been reached, each beam specimen was removed from the form and individually tested. Additionally, multiple cylinders were taken from each placement in order to test and confirm that the anticipated compressive strength of the concrete had been reached. 4.2 .2 Testing For each test, a beam was set up in the Iosipescu lo ading scheme shown in Figure 4.3 and plac ed into a 200 kip capacity MTS load frame shown in Figure 4.4 Figure 4. 3 Iosipescu loading s cheme.

PAGE 26

16 Figure 4. 4 Photograph of specimen in the Iosipescu apparatus. The wood blocks are for temporary support and are not part of the test. A number of small steel parts were required for the Iosipescu test and will be described fr 6x6 x inch steel plate was placed on the bo t tom platform of the MTS load frame just to protect the equipment. Next a 1 inch diameter round rod 6 inches long was placed in the center of the plate in order to create a single point load that would act from below the beam specimen. Two 1x1 inch square bars 6 inches long were adhered wit h epoxy to a plate that was 1 inch thick, 6 inches deep, and 14 inches long. They were positioned on the plate with 8 inches measured center to center and 3 inches from each center to the edge of the plate. This assembly was balanced with the square rods f acing upward on the 1 inch diameter round rod so that the centerline of the rod

PAGE 27

17 lined up with the inside edge of one of the square rods. In order to accomplish this, a wood block was cut to the appropriate length to support the other side of the plate and keep the surface flat. Then the beam was placed onto the square rods so that the end of the beam lined up with the end of the plate below. Again, a wood block was placed under the other side of the beam for support. An identical assembly of the plate and t he two square bars was then placed on top of the beam anti symmetrically, mirroring how the bottom was placed but offset in the opposite direction with the inside of the other square rod lined up with the centerline. Next a 1 inch diameter round rod 6 inch es long was placed along that same centerline to create a single point load that would ac t from the top with another 6x6 x inch plate above to again protect the top platform of the MTS load frame Care was taken in order to assure everything was concentri cally placed. Then a compressive force was applied under displacement control at a constant rate of travel while a load cell measured and a computer recorded three parameters: force, time, and displacement. Each beam was loaded until it failed in shear at which point the test was complete. As built dimensions were also measured prior to each test. It should be noted that as the test was being conducted, the ram at the bottom would rise so that the set up would no longer be supported by the wood blocks. Ther efore they did not influence the test in any way; the set up would remain in place due to th e compression of the MTS load frame 4.3 Flexural Shear Test Six beams with two layers of concrete were constructed and tested in order to determine if the method created is approp riate to evaluate the flexural shear strength of a

PAGE 28

18 bonded interface Three different surf ace treatments were utilized between the two layers of concrete The s ix beams were tested in flexure in two ways using a three point t esting method. The first way was a typical simply supported beam: a support would be prov ided underneath both ends of a beam while a downward force was applied to the top center of the beam to induce flexure referred to herein as flexure induced by compre ssion The second way was the reverse of this. Steel pipes placed strategically between the layers during concrete placement were used to anchor each end of a beam while an upward force was applied to the center of the beam to induce flexure referred to h erein as flexure induced by tension Figure 4.5 illustrates the basic forces applied for these two methods. Figure 4. 5 Flexure induced by shear at interface combined with compression (left) and by shear at interface combined with tension (right). In theory, the beam specimens tested using compression will have a higher flexural shear strength than those tested using tension; the compression will add a clamping force mak ing it more difficult for the layers to debond from one another while the tension will help facilitate the separation between layers. The following summarizes details of the construction and testing of the beams.

PAGE 29

19 4.3 .1 Design First, the test had to be desig ned. Dimensions of the beam specimens were determined with the help of an Excel spreadsheet Setting some basic parameters was the starting point. Based on prior shear stress tests done by Rosen at UCD (Rosen, 2016 ) the target shear stress was set at 500 psi. For the beams to be considered narrow the minimum span to depth ratio was set to 4. The spreadsheet was set up to calculate the flexural shear stress, the moment, the required reinforcement, and the shear stren gth of the beam from an input of dimensions and forces. Trial and error was used and the dimensions and forces were adjusted until optima l calculated results were reached. The goal was t o get each value as close to the target value as possible. Table 4.1 shows the final values for each target val ue, design value, dimension, and calculation. For dimensions, L represents the length of the beam, h represents the height of the beam, d represents the distance from the top of the beam to the centroid of the reinforcement, b represents the wi dth of the beam, and t represents the width of the contact area between the two layer s of concrete. Table 4. 1 Flexural shear test design parameters Dimensions Span/Depth Ratio Flexural Shear Stress Predicted Forces Calculated Design Shear Strength Reinforcement L = 7 ft 6.68 kips Required As 0.583 in 2 h = 11.25 in Min imum 4 Target 500 psi Midspan d = 10.5 in Applied Moment Provided As 0.6 in 2 b = 4.5 in Design 7.47 Design 467 psi Supports 24.5 k ft t = 2 in [w/ (3) #4 Bars]

PAGE 30

20 I n order to ensure that the desired type of failure woul d occur, a bond breaker was used at the interface to decrease the surface area of the connection. Again, t he width of the surface area at the interface is represented by t in Table 4.1 Figure 4.6 shows a cross section of a beam specimen with the dimensions labeled. Figure 4. 6 Cross section of typical beam specimen for flexure tests The bond breaker completely separates the two layers of concrete at the ends and center of the beam (left) and only partially separates the two layers in the middle of each half to create the influ ence areas illustrated later in Fig ure 4.11 4.3 .2 Specimen Preparation Next the beam specimens were created. Forms were constructed with a plywood base and 2x12 wood boards. Since the beams were to be 11.25 inches in height, it w as convenient to use 2x12s because the actual width of a 2x12 is 11.25 inches. Seven boards 7.5 feet in length w ere nailed to the plywood. Because the design length was set at 7 feet, each end was extended an additional 3 inches to allow for a support to be placed under each e nd during testing. This made the actual length of each beam 7.5 feet. The boards were attached along their narrow edge for the l ength of the plywood to create six

PAGE 31

21 compartments 11.25 inches high and 4.5 inches wide with a board across each end. As previousl y mentioned, it was important that the bea ms failed in the correct manner: flexural shear rather than diagonal tension or flexure. flex ure, reinforcement was utilized: three No. 4 Grade 60 reinforcing bars we re to be used i n each beam. N ext, the rebar had to be cut and the ends bent into 180 hooks in order to achiev e a suitable development length and fit into the dimensions of the beams properly. Once all the rebar had been cut and bent they were connected together with tie wire in groups of three as seen in Fig ure 4.7 and placed into the forms. Figure 4. 7 Reinforcing steel used in each beam specimen. Small strips of extra rebar were attached to the bottom of each group of three as spacers so that thus providing some clear space The three beams that were to be tested in flexure using tension required a method to anchor the two ends to the ground. As a result, steel pipe 1 inch in diameter was cut into 4.5 inch segments and was wedged into the end of three of the compartments 3 so that the top of the pipe would be flush with the top of the fi rst layer of concrete. Further details regarding the purpose of the pipes

PAGE 32

22 will be explained later. Once this had been done, the forms seen in Figure 4.8 were ready for the first concrete placement. Figure 4. 8 Concrete form ready for first placement. The beams were then labeled 1 6. Beams 1 3 were to be tested in flexure using compression and Beams 4 6 (those with the steel pipe segments ) were to be tested in flexure using tension. At this point, it was time to place the first layer of concrete. The te st utilized a 5000 psi sack mix concrete and color pigment was added in order to clearly see the demarcation line between the two layers of concrete. The forms were filled to the half way mark 5.625 inches high. The concrete in each compartment was then

PAGE 33

23 co nsolidated using a pencil v ibrator to expel any air pockets and to assure the concrete filled each form completely, particularly under the rebar and around the steel pipe segments Before the concrete was allowed to set, a screwdriver was used to create a f the six beams, one of the beams with pipes (Beam 4) and one without (Beam 1) as can b e seen in Figure 4.9 Figure 4. 9 Then the concrete was allowed to cure for 28 days. After 28 days two more beams received surface treatments. A jackhammer with a bush bit was used on one of the beams with pipes (Beam 5) and one without (Beam 2). This provided three different surface trea tments to be tested: a smooth finish (Beams 3 and 6), a bush hammered finish eams 1 and 4) seen in Figure 4.10 Table 4.2 outlines the surface treatment of each beam specimen and the way i n which they were to be test ed.

PAGE 34

24 Figure 4. 10 Flexural beam specimen s (middle), and smooth (bottom). Table 4. 2 Summary of beams, surface types, and testing methods used. BEAM NUMBER SURFACE TYPE TESTING METHOD 1 RAKED COMPRESSION 2 BUSH HAMMERED COMPRESSION 3 SMOOTH COMPRESSION 4 RAKED TENSION 5 BUSH HAMMERED TENSION 6 SMOOTH TENSION As previously mentioned a bond breaker was needed to encourage slippage between the two layers ; this reduced the surface area of the connection between the two layers of concrete. This slippage was the flexural shear failure being tested. P lastic wall paneling 2 mm thick was cut the same length and width of each beam to create the bond breaker Influence lengths were calculated based on using 6 inch wid e supports at each end and a 4.5x4.5 inch plate in the middle of the beam under the jack, see Figure 4.11

PAGE 35

25

PAGE 36

26 Two square holes 2 inches wide (the t dimension in Table 4.1 ) and the influence length long were cut into each strip of the bond breaker to create the areas where the two layers of concrete would h ave direct contact. This area was the bonded interface bei ng tested. The interface was then exposed to the flexural shear force created from the bending of the beams. The bond breakers were set into each compartment of the form and secured with some duct tape as can be seen in Figure 4.12 Figure 4. 12 Form pri or to second concrete placement. The white bond breaker is PVC wall paneling sheet.

PAGE 37

27 Unfortunately, the rake marks previously created on two of the beams extended the whole width of the beam for the entire length of the beam. This meant that there were holes for the concrete to fl ow into which defeated the purpose of the bond breaker. To eliminate this possibility, P laster of P aris was used to fill in the rake marks that were located directly under the plastic bond breaker. Once the bond breakers had been set and secured in the forms, it was time for the second pla cement of concrete. Immediately prior to the second placement of concrete, the surfaces were cleaned and water was sprayed into each compartmen t and sponged off to achieve the ideal SSD conditions previously discussed for t he direct shear test. Additionally one 4.5 inch segment of the 1 inch diameter steel pipe was set directly on the bond breaker in the center of Beams 4, 5, and 6. These would later be used to attach the middle of the beam to the tension testing machine. The three beams that w ere to be tested in tension, Beams 4, 5, and 6 needed to have rebar in the tops of the beams rather than the bottom s so rebar was cut, bent, and tied togethe r in the same manner as before. The next layer of concrete was then placed without the added color pigment. The rebar was inserted into the fresh concrete on Beams 4, 5, and 6 so that the tops of the spacer bars were flush with the top of the concrete. Then the concrete was consolidated again with the pencil vibrator and allowed to cure for another 28 days. Multiple cylinders were taken from each concrete placement in order to test and confirm that the anticipated compressive strength was reached within 28 days. 4.3 .3 Testing Each beam specimen was removed from the form and as built dimensions were measured. The goal was to induce a slipping failure between the two layers of concrete.

PAGE 38

28 Beams 1, 2, and 3 were tested in compression while Beams 4, 5, and 6 were tested using tension as previously described in this chapter. 4.3 .3.1 Flexure by Compression Testing The set up for the compression testing was comprised of a simply supported beam with a single downward force acting at midspan. The lab at UCD has 1 inch diameter round threaded metal posts of different lengths that can be secured to a strong floor by screwing them into threaded holes set 3 feet apart center to center in a grid throughout the lab. Two steel channels were secured together to form an I s haped beam that was mounted on two of the round metal posts abo ut 3 feet in the air. T hen a plate was attached to the bottom of the two channels at the center of the beam with C clamps in order to create a platform for t he jack to press against. The purpose of the apparatus was to have something for the jack to brace itself against when app lying the downward force onto the center of each beam; the jack would be sandwiched between the I beam and the beam specimen and then pumped to expa nd. Two 3 foot long 4x6 Douglas Fir wood posts were placed on either side of the I beam/post set up with the 6 inch side on the ground 7 f eet apart center to center. These would serve as the supports for each beam and allow for deflection to occur. Figures 4.13 and 4.14 show the design used for this test.

PAGE 39

29

PAGE 40

30 Figure 4. 14 Design of flexure by compression test. Section 1. At this point, it was time to test Beams 1, 2, and 3. To begin each test a beam specimen was placed under the center o f the I beam onto the wood supports The j ack used was a Simplex RC306C with an area of 6.44 square inches The jack was placed between the I beam and the beam specimen at its center. Figure 4.15 illustrates the complete set up of the flexure by compression test.

PAGE 41

31 Figure 4. 15 Flexure by compression testing set up. By pumping the jack at a slow rate, a const ant downward force was applied to the 4.3 .3.2 Flexure by Tension Testing The set up for the tension test was a bit more elaborate than that for the compression test but the concept was similar: anchor the two ends of the beam and apply a single upward force to the midspan. A 50 kip capacity load frame was used, pictured in Figu re 4.16

PAGE 42

32 Figure 4. 16 50 kip capacity load frame used for the flexure by tension tests. First it was necessary to anchor the two ends of the beam to the ground so it would be c entered under the machine. A 2 inch diameter hole was drilled through each end of a 7.5 foot long steel 4x4x inch HSS (hollow structural section) tube with 6 feet measured center to center. These holes allowed the HSS tube to be mounted on two short threaded metal posts screwed into two of the threaded holes in the lab 6 feet ap art centered under the machine. Next, h ollow concrete blocks were placed on either end of the HSS tube for each beam specimen to rest upon prior to testing as seen in Figure 4.17

PAGE 43

33 Figure 4. 17 HSS secured to the floor used to anchor the ends of the beam specimens. Once a beam was set into position, it had to be anchored to the HSS tube and connected The 1 inch steel pipe segments placed in the top of each end of Beams 4, 5, and 6 during the firs t concrete placement were used to anchor the beam specimens to the HSS tube. Other materials used to anchor the beam ends included two 10 inch segments of a 1 inch diameter all treaded round rod with two nuts for each, four inch eyebolts with couplers, t wo 10x5x inch steel plates, and extra inch all threa ded rod cut appropriately with four nuts total. On each end, the 1 inch diameter all threaded rod was threaded through the pipe embedded in the top of the first layer of concrete. An eyebolt was put onto each end of the rod so that the beam was sandwiched in between an d

PAGE 44

34 secured with nuts. The extra inch threaded rod was cut appropriately and attached to each eyebolt with a coupler so when they hung down they ex tended an inch or two below the HSS tube. Holes were drilled into the two 10 x5x inch plates 5 inches apart center to center so the inch threaded rod could pass through them. The plates were held under the HSS tube and the threaded rod was threaded throu gh them and secured with nuts. This allowed each end of the beam to be anchored to the HSS tube which was anchored to the ground. The 1 inch steel pipe segments placed in the middle of Beams 4, 5, and 6 on the bottom of the second concrete pla cement were used to connect the beam specimens to the arm of the machine. Other materials used to make this connection included a 10 inch segment of the same 1 inch diameter all thread ed round rod with two nuts, a 8x8 x inch steel plate, four 5/8 inch diam eter bolts with nuts, two inch eyebolts with couplers, and extra inch all threaded rod with two nuts. The configuration at the center connecting the beam to the machine was similar to the con figuration at each end. The 8x8 x inch plat e had six holes dr illed into it; four were used to attach the plate to the machine and two were used to attach the beam specimens to the plate. The plate was attached to the m with the bolts and nuts. Next the threaded rod was threaded through the pipe at the ce nter of the beam being tested and the eyebolts were secured onto each end with nuts. The extra inch all threaded rod was cut appropriately, connected to eac h eyebolt with the couplers, inserted through the two other holes on the plate and secured with nuts. Figures 4.18 and 4.19 sho w the design used for this test

PAGE 45

35

PAGE 46

36

PAGE 47

37 Once a beam specimen was set into place, it was time for the testing to begin. Figure 4.20 illustrates the complete set up of the flexure by tension test. Figure 4. 20 Flexure by tension testing set up. constant rate until failure occurred and the two layers debonded from one another. During the test a computer recorded force, displacement, and time and once the failure occurred, the fra cture was observed and analyzed. While getting the second beam specimen (Beam 5) set in to place, some unanticipated separation of the layers occurred at each end early during the testing. This was addressed by modifying th e way in which the beam was anchored to the HSS tube. Longer segments of threaded rod were used to connect the plate located underneath the

PAGE 48

38 HSS tube to another additional plate placed on top of the beam so that both the HSS tube and beam were being sandwiched or clamped together with the plates and the threaded rod. This modific ation can be seen in Figure 4.21 However, since the bond breaker was solid at the ends, it is believed th at this did not influence the r esults acquired for this beam. Figu re 4.21 Modified testing set up for Beam 5. 4.4 Compressive Strength Test ing During the first and second placement of concrete three samples were ta ken from each batch in order to test the compressive strength of the mix. The sample s were placed into cylindrical molds 4 inches in diameter and 8 inches tall. This was done in three lifts

PAGE 49

39 with consolidation occurring between using a metal rod. Then the cylinders were set as ide and allowed to cure. Three days after placement, one cylinder from each batch was tested. Testing was done again on day seven and day twenty eight in order to develop a compressive strength curve for each batch. For each compression test, the sample cy linder was removed from the mold, placed between two neoprene pad caps, and centered in a Forney compression testing machine as pictured in Figure 4.22 A compressive force was applied at a slow, constant rate until failure at which time the maximum force was recorded. Figures showing the compressive strength curves developed can be found in the following chapter. Figure 4.22 C ompressive testing mach ine used.

PAGE 50

40 CHAPTER V RESULTS 5.1 Introduction Once all the testing was completed, data from the shear tests conducted for this thesis was compiled. T he results of each tes t are presented in this chapter. 5.2 Com pressive Strength Test ing Data collected from the compressive strength tests was entered into an Excel spreadsheet. A curve was fit to each set of data in order to determine the compressive strength ( c ) of the concrete at any specified time after placement. Figure 5.1 shows the curves associated with the first plac ement of concrete and Figure 5.2 shows the curves associated with the second placement. Figure 5. 1 Compressive strength curves from first concrete placement. 0 1000 2000 3000 4000 5000 6000 7000 8000 0 20 40 60 80 100 COMPRESSIVE STRENGTH (PSI) DAY CYLINDER TESTING FIRST PLACEMENT BEAM 3/4 BEAM 2/4 BEAM 1

PAGE 51

41 Figure 5. 2 Compre ssive strength curves from second concrete placement. From the curves, a n average compressive strength was calculated for each placement. The lower of the two govern s and is used to express the results of the other tests as a function of the compressive strength of the concrete; the shear strength is equal to a factor multiplied by the square root of the compressive strength of the concrete. 5.3 Direct Shear Test As expected, each direct shear test specimen successfully fractured along the bonded surface onc e the force from the MTS load fra me overcame the bond strength. An example of this can be seen in Figure 5.3 The beams with smooth and bush hammered surface treatments had extremely clean fracture faces between the layers while the jackhammered beams had fractures that varied a bit more. 0 1000 2000 3000 4000 5000 6000 0 20 40 60 80 100 COMPRESSIVE STRENGTH (PSI) DAY CYLINDER TESTING SECOND PLACEMENT IO BEAMS BEAM 1/2 BEAM 4/5

PAGE 52

42 F igure 5. 3 Iosipescu beam specimen at f ailure. Once the data had been compiled, it required little manipulation in order to extract the desired information. In other applications of this method, researchers have made shallow notches on the top and bottom of the beam close to the centerline to induce the fracture from one notch to the other. Because of this, the fracture occurred at a slight diago nal angle between the top and bottom inside square bars in which case the surface area of the fractured face was slightly larger than the cross secti on of the beam. However, in this case, there already was a weakened plane at the bonded interface so notche s were not necessary and the fractu re occurred at the bonded cross section of the beam. No new surface area calculations or decomposition of forces were necessary. The s hear stress was calculated dividing the force parallel to the plane (the shear force fr om the MTS load frame ) by the surface area of the failure plane (the cross sectional area of the beam). The data is gra phi cally represented in Figures 5.4 5.9 Each graph shows the results of the

PAGE 53

43 four b eams for each surface treatment and is presented with absolute values and values expressed as a function of the square root of the compressive strength of the concrete, c Figure 5. 4 Shear stresses measured for the smooth interface. Figure 5. 5 Shear stresses measured for the bush hammered interface. -100 0 100 200 300 400 500 600 0 10 20 30 40 50 60 70 80 SHEAR STRESS (psi) TIME (s) SMOOTH INTERFACE SURFACE S1 S2 S3 S4 -100 0 100 200 300 400 500 600 0 20 40 60 80 100 120 140 SHEAR STRESS (psi) TIME (s) BUSH HAMMER INTERFACE SURFACE B1 B2 B3 B4

PAGE 54

44 Figure 5. 6 Shear stresses measured for the jackhammered interface. Figure 5.7 Shear stresses me asured for the smooth interface expressed as a function of c ). -100 0 100 200 300 400 500 600 0 20 40 60 80 100 120 140 SHEAR STRESS (psi) TIME (s) JACKHAMMER INTERFACE SURFACE J1 J2 J3 J4 -1 0 1 2 3 4 5 6 7 8 0 10 20 30 40 50 60 70 80 / f' c TIME (s) SMOOTH INTERFACE SURFACE S1 S2 S3 S4

PAGE 55

45 Figure 5. 8 Shear stresses me asured for the bush hammered interface expressed as a c ). Figure 5. 9 Shear stresses me asured for the jackhammered interface expressed as a c ). It can be seen from the graphs that the outcomes for each surface type are fairly consistent. Finally, the maximum shear stress for each of the twelve beams was determined. The average maximum shear stress was calculated for each surface type and -1 0 1 2 3 4 5 6 7 8 0 20 40 60 80 100 120 140 / f' c TIME (s) BUSH HAMMER INTERFACE SURFACE B1 B2 B3 B4 -1 0 1 2 3 4 5 6 7 8 0 20 40 60 80 100 120 140 / f' c TIME (s) JACKHAMMER INTERFACE SURFACE J1 J2 J3 J4

PAGE 56

46 has been summarized in Tables 5.1 and 5.2 Again, the first table presents absolute values square root of the compressive strength of the concrete, c Table 5. 1 Summary of shear stresses in psi. SMOOTH BUSH HAMMERED JACKHAMMERED BEAM 1 272 467 518 BEAM 2 409 278 455 BEAM 3 281 389 274 BEAM 4 251 280 370 AVERAGE 303 354 404 COV 24% 26% 26% Table 5. 2 Summary of c values. SMOOTH BUSH HAMMERED JACKHAMMERED BEAM 1 3.83 6.60 7.31 BEAM 2 5.78 3.93 6.42 BEAM 3 3.97 5.49 3.86 BEAM 4 3.55 3.96 5.22 AVERAGE 4.28 4.99 5.70 COV 24% 26% 26% 5.4 Flexural Shear Test Results from the flexural shear test s are presented in Table s 5.3 and 5.4 and F igure s 5.10 and 5.11 Each table and figure is presented with absolute values and values expressed as a function of the square root of the compressive strength of the concrete, c Due to a machine error, no data was collected for Beam 4.

PAGE 57

47 Table 5. 3 Flexural shear test results: Beams 1 3.

PAGE 58

48 Table 5. 4 Flexural shear test results: Beams 4 6.

PAGE 59

49 Figure 5. 10 Maximum flexural shear stress of Beams 1 6. Figure 5.11 Maximum c ) value 0 100 200 300 400 500 600 700 MAXIMUM SHEAR STRESS (psi) BEAM 1 BEAM 2 BEAM 3 BEAM 4 BEAM 5 BEAM 6 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 MAXIMUM BEAM 1 BEAM 2 BEAM 3 BEAM 4 BEAM 5 BEAM 6

PAGE 60

50 CHAPTER VI DISCUSSION 6.1 Introduction As discussed in the introduction and background of this thesis, the engineering community is in need of a way to evaluate the shear strength capacity of a bonded interface between two layers of concrete. The methods previously outlined were tested and the results have been presented. The following contains an analysis of those results. 6.2 Direct Shear Test As shown in Tables 5.1 and 5.2 the jackhammered interface had the highest shear stress capacity at 404 psi or 5.70 c the bush hammered interface had a slightly lower capacity at 354 psi or 4.99 c and the smooth interface had the lowest capacity at 303 psi or 4.28 c This gradu al decrease in strength correlates well with the decrease in surface roughnes s, and is the first good sign that the test was a success. Repeatability among the results fo r each surface type is the next encouraging sign that the test is applicable. The coefficient of variation for each set of data was around 25% Additionally, t he numerical value of each average is also in general agreement with other shear tests done on similar surface treatments by Rosen (Rosen 2016 ) and Pultorak (Pultorak 2016) of UCD Table 6.1 compares the results from the testing conducted for this thesis with Rosen and Pultora s research. Since all three researchers tested a bush hammered surface, it is appropriate to make a direct comparison of the results associated with that surface treatment. The value obtained by this author of 354 psi lies half way from one another a bit when expressed as a factor multiplied by the square root of the

PAGE 61

51 compressive strength of the concrete, c because there was a range in those values. If the same concrete was used for each of the tests, perhap s these factors would have been closer to each other. Table 6.1 Comparison of results from other researchers. RESEARCHER TESTING METHOD SURFACE TREATMENT SHEAR STRESS GOVERNING f'c SHEAR AS FUNCTION OF SQRT(f'c) SWAN IOSIPESCU SMOOTH 303 5016 3.55 BUSH HAMMER 354 5016 3.96 JACKHAMMER 404 5016 5.22 ROSEN GUILLOTINE BROOM 496 5648 6.60 BUSH HAMMER 294 5650 5.22 RAKE 457 3172 6.08 PULTORAK GUILLOTINE BUSH HAMMER 416 3069 7.50 Although the tests appear to have be en successful, there are possible sources of error. Ideal beam specimens w ould have equal lengths of substrate and overlay concrete but because the forms were filled half way and then bush hammered or jackhammered, some of the be ams had larger portions of overlay concrete than substrate, particularly those that had been jackhammered In the future, it is recommended to fill the compar tments to be bush hammered and jackhammered a bit more than halfway, e specially the jack hammered beams; this would provide an allowance for some material loss during surface preparation. Additionally, the mechanical surface roughening techniques used may layer of the substrate concrete making the bond weaker than if other techniques were utilized. Because the forms were constructed from wood, warping of the boards created beam specimens that did not have the exact

PAGE 62

52 same d imensions and cross sections over their full leng th. During the testing possible misalignment of the Iosipescu apparatus as well as possible non concentric load application may have occurred despite the care taken in positioning the parts of the apparatus. Even with these possible sources of error, it can be concluded that the results obtained support the belief that the modified Iosipescu method is an appropriate test to evaluate the direct shear capacity of concrete repairs and overlays with a bonded interface. 6.3 Flexural Shear Test Unlike the direct shear test results t he results from the flexural shear stress tests have excessive variation. In the case where compression was used to induce flexure, Beam 3 was the only one that experienced a true flexural shear fa ilure as seen in Figure 6.1. Figure 6 .1. Flexural shear stress failure of Beam 3.

PAGE 63

53 While reinforcing steel precluded flexural failures, the other two beams (Beam 1 and 2) experienced a diagonal tension failure prior to a flexural shear failure. The fracture line tended to start by the jack traveled diagonally down to the interface, then across the interface for a short distance, and then diagonally down to the reinforcement This can be seen in Figure 6.2 Figure 6.2 Diagonal tension failure of Beam 1. Because two of the three beams failed in this manner, it seems as though the reinforcement may have created a weak plane where the concrete tended to separate from itself at that location rather than at the interface. Although the beams did no t experience the desired type of failure, it can be concluded that i n these two cases (Beams 1 and 2) the flexural shear strength of the bonded interface was higher than the diagonal tensile strength of the beam This aspect of the test is in agreement wit h the varying surface treatments since the beam with the smooth surface treatment (Beam 3) experienced flexural shear failure, meaning the flexural shear strength of that interface was less than

PAGE 64

54 the diagon al tensile strength of the beam; making its capacit y the lowest. T his was the expected result This is not illustrated in the bar graphs presented because Beams 1 and 2 failed in diagonal tension prior to flexural shear. Thus, t he graphs are not a good representation of the shear strength of each bonded interface. As predicted, there is a significant difference in values between the beams tested in flexure using compression versus the beams tested in flexure using tension. This can be attributed to the clamping action incurred versus the facilitating pul ling action discussed earlier. Several problems arose when ten sion was used to induce flexure. Issues with the testing set up prevented the author from acquiring any usable data for the first beam ds required a modification of the set up. Thus, the last beam (Beam 6) was the only one where testing went as planned. However, while observing the test it appeared as though more of a tensile failure occurred at the interface rather than any bending of t he beam and the top layer was simply lifted off of the bottom layer as pictured in Figure 6.3 Because of all the complications and variability encountered during the flexural shear strength testing process, the author has concluded that this is not the b est way to evaluate the flexural shear stress capacity of a bonded concrete interface.

PAGE 65

55 Figure 6.3 Failure of Beam 6.

PAGE 66

56 CHAPTER VII CONCLUSION & RECOMMENDATIONS B ased on the results obtained, the author believe s that the modified Iosipescu method of testing can be appropriately used to evaluate the direct shear strength capacity of different ty pes of bonded concrete interface s. However, the impediments encountered with the method used to evaluate the flexural shear streng th capacity of a bonded interface make that test ineffective The flexural shear tests proved to be unsuccessful, particularly in the case of the flexure induced by tension because of the way the top layer lifted off of the bottom layer rather than sliding away from each other from the bending. However, with some alterations it is believed that the flexure induced by compression test could still be viable if future testing was conducted. If this is done, a few rec ommendations can be made. The size of the beams made it extremely difficult to maneuver and manipulate the m around the lab for testing, so the use of smaller beams may be beneficial. A different reinforcement plan should also be utilized to ensure the prop er type of failure occurs in all beam specimens. Thi s type of test has been successfully used in the past by other researchers and m ay still be a good method however the testing conducted as a part of this thesis was not successfu l More testing can and s hould be done to perfect the modified Iosipescu direct shear testing method presented To further delve into the topic, other surface roughening techniques such as sand or water blasting or acid etching may be explored. Again, b ecause the beam specimens we re fairly large and cumbersome, the tests could be conducted with smaller beams to see if a lighter specimen would be appropriate Helmick

PAGE 67

57 et al. recommend using 4x4 inch beams and this author is in agreement In order to create identical and exact specim ens it might be better to use steel forms rather than wood. The wood boards were warped which created varying sizes and shapes of the beams. If a bit more experimental testing is performed on the topic with the above recommendations taken into consideratio n, it is believed that this could be formalized into a sta ndard test for concrete repairs and overlays.

PAGE 68

58 REFERENCES Existing Concrete Structures (ACI 562 16) Concrete Institute, Farmington Hills, MI, 2016, pp 46 48. (Sept. 2016). Journal of Composite Materials, V. 21, February 1986, pp 494 507. Resin Systems Used With Concrete By Slant West Conshohocken, PA. and the Bond Strength or Tensile Strength of Concrete Repair and Overlay Materials by Direct Tension (P ull Materials, West Conshohocken, PA. Materials by the V Materials, West Conshohocken, PA Concrete International, V. 38, No. 9, September 2016, pp 56 60. Craig, R.R. Jr., (2000), Mechanics of Materials 2 nd Edition, John Wiley & Sons, Inc., Ho boken, NJ, pp 391 395. Concrete Composites, V. 25, No. 3, April 2003, pp 379 385. Helmick, C.G.; Toker Tension in Pl No. 1, January 2016, pp 39 46. 05. r Strength of Composite Concrete Beams January 1994, pp 48 69.

PAGE 69

59 of Structural Engineering, V 127, No. 4, April 2001, pp 359 366. Denver, Denver, CO. thesis, University of Colorado, Denver, Denver, CO. between Shotcrete Overlay and Reinforced Concrete 1 3. Wan thesis, Louisiana State University, Baton Rouge, LA. World Business Council for Sustainable Development. (2015). "Recycling Concrete Executive Summary." The Cement Sustainability Initiative (CSI), (Oct. 4, 2016).

PAGE 70

60 APPENDIX The following photograph s illustrate the outcome of the tests performed for this thesis.

PAGE 71

61 Direct Shear Test Smooth Interface:

PAGE 72

62 Bush Hammered Interface:

PAGE 73

63 Jack Hammered Interface:

PAGE 74

64 Flexural Shear Test Flexure Induced by Compression: BEAM 1 BEAM 2 BEAM 3

PAGE 75

65 Flexure Induced by Tension: BEAM 4 BEAM 5 BEAM 6