Bond Interface Strength between Ultra High Performance Concrete and Normal Concrete

Transcription

1 Bond Interface Strength between Ultra High Performance Concrete and Normal Concrete Presented by: Mariah Safritt MATS UTC Undergraduate Summer Research Program University of Virginia Virginia Center for Transportation Innovation and Research July 2015 Total Words: 4,665 + (11*250) = 7,415 Total Figures: 11

2 Mariah Safritt 2 ABSTRACT The crumbling infrastructure in the United States has become a major problem, and needs to be replaced throughout the country. There is need for a more sustainable and cost effective approach to maintain bridges and other structures. Ultra high performance concrete (UHPC) is a much stronger material than normal concrete (NC), so it can be used in thin layers as a repair material. It uses fewer raw materials (industrial byproducts and recycled materials can be used instead of raw materials like cement, sand, and other aggregates), it has a very high strength (usually around 22 ksi), and it is more expensive to produce at first but will theoretically have a longer life cycle and therefore a lower total life cycle cost. The main issue faced currently is its compatibility with existing normal strength concrete structures, and the achievable bond strength between the two materials. In this paper, the highest achieved bond strength was 7,635 psi from Mix 1 with a sandblasted surface preparation. The UHPC mix with the highest compressive strength was Mix 2 (with silica fume and fly ash) with a strength of 11,289 psi. MOTIVATION Repairing damaged and deteriorating infrastructure is very expensive and time consuming, and uses a lot of resources (such as sand, cement, aggregates, water, and energy). Replacing existing concrete structures with the same normal strength concrete is not an efficient or sustainable way of managing the aging infrastructure, because the same problems will keep occurring in the future if a new approach is not taken. A more sustainable option would be to use Ultra High Performance Concrete (UHPC) as a repair material; it is a much stronger material, so it can be used in smaller quantities, like thin layers, and still achieve higher overall strengths for the structure than with using normal strength concrete. Since UHPC started being developed around the world over the past few decades, the main question facing researchers and infrastructure managers alike is whether UHPC can be used in thin layers as a repair material for existing concrete structures such as bridges. In order to determine this, it needs to be investigated whether UHPC and normal concrete (NC) are compatible, and what the bond interface strength between the two materials is. If the bond strength between the two materials is less than either the strength of the UHPC or the NC, then UHPC will not be suitable for use as a repair material as the bond strength between the two materials will weaken the overall structure. INTRODUCTION UHPC has a very high strength compared to conventional normal concrete, which has a compressive strength of about 4 to 8 ksi (28 to 55 MPa) (3). UHPC is typically characterized by its higher compressive strength, its higher tensile strength and ductility (usually due to the addition of fibers), increased durability, and higher initial unit cost (4). In terms of sustainability, UHPC is seen as a much more sustainable product than regular concrete. The energy needed to produce UHPC is about 74% of that needed to produce NC, and the raw material consumption of UHPC is about 58% that of NC (4). UHPC is a very dense material, due to the removal of coarse aggregate and the use of very fine particles such as fine sand, silica fume, fly ash, and sometimes ground quartz powder. These fine particles fill in most

3 Mariah Safritt 3 of the air voids left in the mix, causing the mix to be very dense. Due to this dense mix with very few voids, UHPC has an extremely low permeability. Since water and other particles cannot infiltrate down through the UHPC, this causes it to have greater frost and deicing salt resistance, a lower rate of carbonation, and greater chloride resistance than NC, which leads to lower maintenance and repair costs in the long term (4). The initial unit cost of UHPC is much higher than NC, due to the cost of materials like fly ash and ground quartz as well as the labor costs for the special mixing procedures, but the 100 year life cycle cost is much less than NC (4). A few studies have been done in using Ultra High Performance Fiber Reinforced Concrete (UHP-FRC) composites in thin product applications, such as panels, decks, and cladding, but it can also be used as a stand-alone material (3). Since UHPC is stronger, it can be used in fewer quantities to make smaller more lightweight structures, allowing for easier transport and installation; this can be useful for precast and prestressed concrete applications (5). Other benefits of using UHPC besides its extremely low permeability include self-consolidation and rapid strength gain, which make it an idea material to use as an overlay material for bridge deck repair; it will disrupt traffic for shorter periods of time and be more durable to environmental stresses (5). LITERATURE REVIEW Background Ultra High Performance Concrete, according to the Federal Highway Administration (FHWA), is defined as a hydraulic cement-based concrete with a compressive strength at least equal to 150 MPa (22 ksi) (3). UHPC can also be defined as a cementitious-based composite material with discontinuous fiber reinforcement, compressive strengths above 21.7 ksi (150 MPa), pre- and post-cracking tensile strengths above 0.72 ksi (5 MPa), and enhanced durability via their discontinuous pore structure (4). UHPC has been used around the world for the past few decades in various bridge structures and other applications, but it has only been available commercially in the United States since 2000 (4). Some of the countries that currently use UHPC in their structures include Canada, Germany, France, Australia, Austria, Croatia, Italy, Japan, Malaysia, the Netherlands, New Zealand, Slovenia, South Korea, and Switzerland (4). Many of these countries have also participated in the three International Symposiums on UHPC in 2004, 2008, and 2012 in Germany, publishing their research and findings through the University of Kassel, where the symposiums took place. According to the FHWA, UHPC has been used so far in bridge applications in the following materials: precast, prestressed girders; precast waffle panels for bridge decks; jointing material between precast concrete deck panels and girders; and jointing material between flanges of adjacent girders (4). History There have been numerous advances in concrete technology and design over the past 50 years as science uncovers new ways of making concrete that is stronger and has other desirable properties such as early strength gain, low permeability, higher ductility, and low cost. Some of these advances affected the following aspects of concrete design: fiber reinforcement, cementitious matrix, interface bond between fiber and matrix, fundamental understanding of the mechanics of

4 Mariah Safritt 4 the composite, and improved cost-effectiveness (3). In the 1970s, high compressive strengths up to 510 MPa were achieved by changing the curing conditions (curing with a vacuum, heat, and pressure) and adding fibers to the concrete mix (3). The 1980s brought the addition of polymers, very low water/cement ratios (which led to compressive strengths greater than 200 MPa without need for pressure or heat curing), chemical additives, micro fibers, fly ash, and silica fume (3). In the 1990s, superplasticizers, supplementary cementitious materials, synthetic fibers, and high packing density fine particles began to be used, and the self-consolidating properties of UHPC were discovered (3). In the 2000s, ultra high strength steel fibers, carbon nano-tubes, and carbon nano-fibers started being added to increase the strength of concrete greatly (3). So far in the 2010s, carbon nano-fibers, graphene, and nanotechnology have been emphasized in developing new types of concrete (3). Concrete has undergone many changes over the past 50 years, as can be seen, and there will probably be yet more changes to come in the future. But more important than coming up with new materials is coming up with a way to integrate these new materials into the existing infrastructure in a way that will benefit not only the current generation, but generations to come. Properties of UHPC Since becoming more popular in building applications, there have been many demands expected of UHPC. UHPC is expected to have high strength, toughness, durability, ductility, energy absorption, axial and bending resistance, shear resistance, rotational capacity, and small combined plastic shear and plastic bending deformations (3). It also needs to have high blast and impact resistance, stiffness, freeze-thaw and corrosion resistance, fire resistance, stability, tightness, construct-ability, and an affordable cost (3). Despite all of these conflicting property demands, there are a few property attributes UHPC has that make it the ideal material to use for certain applications. UHPC has negligible permeability and electrical conductivity, resistance to chloride penetration (which is helpful for road and bridge applications that get frequently covered with salt during the winter each year), and volume stability (resistance to shrinkage/expansion during the curing process) (3). Applications to Date In North America, UHPC has been used in box and tee beams, girders, deck panels, connectors, joints, precast post-tensioned sections, joint fill, and panel joints (4). In Europe, UHPC has been used in arch bridges, footbridges, pi-shaped beams, precast and post-tensioned beams, U-shape beams, deck panels, box girders, waterproofing layers, trusses, repair/widening and rehabilitation of bridge decks, and protective surface layers (4). In Asia and Australia, UHPC has been used in precast pre-tensioned I-beams, noise barrier panels, box girders, U-shaped girders, pre-tensioned slabs, composite I-girders, pi-beams, curved saddles for cable stays, and cable-stayed bridges (4). However, uses of UHPC are not limited to bridge applications. UHPC can also be used in drill bits, sewer pipes, precast columns and poles, barrier walls, thin-bonded overlays, cable-stayed bridge superstructures, bridge bearings, precast tunnel segments, and seismic retrofit of columns (4). Mix Design Recommendations Based on many different research projects on how to design the ideal UHPC mix, a few key points will be highlighted here. Because of the nature of UHPC, specific mixing procedures must

5 Mariah Safritt 5 be followed to ensure high quality results. Part of what gives UHPC such high strength is its low water/cement ratio; the lower the water/cement ratio, the higher the strength is (as long as it s above the minimum ratio necessary to complete the hydration process of the cement). However, with a low water/cement ratio, the concrete mix becomes less workable and more difficult to place in forms and molds, so there is a tradeoff. Usually the water/cement (w/c) ratio is kept between 0.16 and 0.27, with the optimum being about 0.22 (6). In order to reduce the water demand, cement with a moderate fineness and low C 3 A content should be used (6). The cement particles should not be too small, because the smaller the particles are, the larger surface area they will have, and they will require more water in order to hydrate. In order to be able to use less water and still have a mix that is workable, High Range Water Reducers (HRWRs) can be used to achieve a more workable mixture without losing strength. Usually HRWRs are polycarboxylate ether based, and 1.4 to 2.4% of cement by weight is most typical (6). In order to have a well graded mixture of dry components, a sand-to-cement ratio of 1.4 with a maximum sand grain size of 0.8 mm (No. 20 sieve) should be used (6). If silica fume is to be used in the mix, it should have a low carbon content, measure 25% of cement by weight, and have a median particle size of 1.2 μm (instead of the usual 0.5 μm) in order to reduce the surface area and therefore water demand (6). In general, in order to achieve a more desirable UHPC mix, all coarse aggregates should be removed, the amount of water used should be reduced, highly refined silica fume and steel fibers should be added, and the silica fume should be 1/100 the size of the cement particles (to fill interstitial voids) (5). METHODOLOGY Experiment Design For this project, 18 different concrete specimens were prepared and tested for bond strength between UHPC and a normal strength concrete substrate. Two different tests were performed, a splitting tensile test (using rectangular prisms instead of cylinders), and a slant shear test. Digital image correlation (DIC) was utilized during all of the tests to obtain more information about how the specimens failed, although that wasn t part of the original project. Three different mix designs of UHPC were used along with three different surface preparations for the interface between the two types of concrete. The three surface preparations used were normal, sandblasted (with an abrasive), and etched with hydrochloric acid. The first UHPC mix design consisted of steel fibers, silica fume, a cement/sand ratio greater than one, and wet sand (saturated surface dry condition). The second mix design consisted of silica fume, fly ash, a cement/sand ratio less than one, and dry sand (oven dried condition). The third mix design consisted of fly ash, a cement/sand ratio less than one, and wet sand (saturated surface dry condition). The full mix design for each of the three mixes can be found in Table 1. The normal concrete was made from an American Concrete Institute (ACI) A4 standard concrete mix, using fly ash and an air content of 7%. The full mix design can be found in Table 2.

6 Mariah Safritt 6 TABLE 1. UHPC mix designs Material Mix 1: Total lbs Mix 2: Total lbs Mix 3: Total lbs Portland cement Fine sand Silica fume Superplasticizer 12 ml 25 ml 12 ml Steel Fibers Water Fly Ash TABLE 2. ACI A4 normal concrete mix design Material Weight (lbs) Portland cement 18.8 Silica Fume 0 Fly Ash 4.7 Water 10.6 Air 7.00% Coarse Agg 54.3 Fine Agg here is not applicable Mix Design When coming up with the different mix designs for the UHPC, several variables that affect the strength and other properties of the UHPC were taken into account: use of fly ash, silica fume, and steel fibers, water to cement ratio, and sand to cement ratio. Fly ash and silica fume are usually used to reduce permeability, reduce chloride diffusivity, and increase the resistivity of a concrete mix (5). Steel fibers are added to concrete mixes to increase the tensile strength and ductility by helping the concrete stay together after cracks begin to form during loading. The water to cement ratio is usually kept lower for UHPC mixes to increase strength, but in order to be classified as low slump dense UHPC, the w/c ratio should be no greater than 0.3 (5), which is the w/c ratio used for all three UHPC mix designs. The other materials used in the mixes were sand, cement, and a HRWR (also called superplasticizer). For the sand, the two factors taken into account were fineness and moisture content. The fineness used was 75 μm to 1.2 mm, which is smaller than a No. 16 sieve and larger than a No. 200 sieve (2). For the moisture content, wet sand (saturated surface dry or SSD condition) from the Virginia Center for Transportation Innovation and Research (VCTIR) was used in mixes one and three, and oven dried sand from the University of Virginia (UVA) was used in mix two. Extra HRWR was used in mix two to compensate for the absorption of water due to the dryness of the sand. The cement type used was type I/II, which is the most common type of cement used for general construction purposes. The HRWR used was Glenium Experiment Procedure After deciding which tests would be performed and how many specimens were needed, the molds for the cylinders and prisms were obtained. The mold sizes were based on ASTM standards for ASTM C39 (7), ASTM C882 (8), and ASTM C496 (9). For the splitting tensile test, rectangular prisms that were 4 by 3 by 16 inches were used, and brass molds of this size were borrowed from VCTIR. The prisms were divided into two sections by using pieces of wood to fill half the width of the prism while one side was being filled with one type of concrete. Therefore, each prism was made up of one section of NC that was 4 by 1.5 by 16 inches, and one

7 Mariah Safritt 7 section of UHPC that was 4 by 1.5 by 16 inches. Before testing, these were cut up into sections that were 3 inches long, as can be seen in Figure 1 (5). The cylinders were 4 inches in diameter and 8 inches long. They were divided in half along an angle of 30 from the vertical, leaving a lip of 0.5 inches on each end. These dimensions were adapted from ASTM C882 (8), which calls for 3 by 6 cylinders, and changed to 4 by 8 inches due to the availability of 4 by 8 molds at UVA. A diagram for the cylinder specimen setup can be seen in Figure 2 (5). In order to create half a section of concrete at a time, it was necessary to make dummy sections to put in the cylinder so that only half a cylinder of concrete would be made at a time. To create these dummy sections, pieces of foam were glued together to form a cylinder shape, and then each one was cut in half along a 30 angle and placed in the forms. Figure 1. Splitting Tensile Prism Diagram Figure 2. Slant Shear Cylinder Diagram The normal concrete sections were made first, so that they would have a longer time to cure and gain strength. Then, after curing for 7 days in a moist curing room at VCTIR, extra 4 by 8 cylinders of the concrete (made for compressive strength testing purposes) were tested and found to have a strength of 4,668 psi which was deemed sufficient to move on with the experiment process. The surfaces of the half cylinders and prisms that would be in contact with the UHPC were then prepared by sandblasting three cylinders and three prisms, and etching three cylinders and three prisms with hydrochloric acid (HCl). The purpose of this step was to roughen the surface of the concrete, to etch away some of the cement paste and expose the aggregate so the surface would be rough enough to form a mechanical bond with the UHPC when it was placed on the NC. After this step was completed, the UHPC was ready to be mixed and placed in the forms. The mixing procedure used was as follows. First all the cementitious materials (cement, silica fume, and the pozzolanic fly ash) were mixed together dry to avoid clumping. Then the sand was added and mixed thoroughly with the other dry materials. Next the HRWR was added to the water and stirred, and then half of the water was added to the mixture and mixed for 1-2 minutes. Then the remaining water was added and mixed thoroughly for 3-4 minutes or until all the cement became hydrated. At the very end of the mixing procedure the steel fibers were added for Mix 1 and mixed until the fibers were evenly distributed throughout the mix. The concrete mix was then placed in the remaining half of each of the forms. The completed specimens were then placed in the moist curing room at VCTIR for about two weeks. Images of the specimens can be seen in Figures 3 and 4.

8 Mariah Safritt 8 Figure 3. Splitting tensile cube Figure 4. Slant shear cylinder Three different sets of UHPC cubes that were 2 by 2 by 2 inches were tested for compressive strength at 2 days, 7 days, and 14 days. Once the UHPC was deemed strong enough for testing, the prisms were cut up into 3 inch long sections using a water-cooled saw at VCTIR, and then all of the specimens were brought back to UVA for testing. The front faces of each of the specimens (all 9 cylinders, and 1 representative cube from each of the prisms) were spray painted white and speckled using a black permanent marker to create a random pattern of dots to be used for DIC imaging and analysis. Once these specimens were painted and ready to go, they were tested according to ASTM C882 (8) and C39 (7) for the slant shear test, and ASTM C496 (9) for the splitting tensile test. Three dimensional DIC was performed on all of the specimens in order to gather more data and information from the rounded cylinders. Once the tests were completed, the data was then post-processed and analyzed using Vic-3D software to determine the strain contours and to analyze how each specimen failed as it was loaded. The specimen failures were characterized, the strengths recorded, and images taken of the broken bond surfaces. RESULTS The UHPC cube compressive strengths from 2 days, 7 days, and 14 days can be found in Table 3 below. The results from the slant shear test and splitting tensile test can be found in Table 4 and Table 5 respectively. The overall strength and failure characterizations for each specimen are listed. TABLE 3. UHPC cube compressive strengths Batch 2 Day Strength (psi) 7 Day Strength (psi) 14 Day Strength (psi)

9 Mariah Safritt 9 TABLE 4. Slant Shear test results Specimen Surface Prep Strength (psi) Break Characterization 1J HCl etched 2013 Bond interface 1K Sand blasted 2490 Bond interface 1L Normal 2153 Bond interface 2J Normal 1975 UHPC failure 2K HCl etched 1519 UHPC failure 2L Sand blasted 2841 Bond interface 3J Normal 1734 Bond interface 3K HCl etched 1724 Bond interface 3L Sand blasted 2129 Bond interface TABLE 5. Splitting Tensile test Specimen Surface Prep Strength (psi) Break Characterization 1A Normal 4200 Bond interface 1B Sand blasted 7635 Substrate failure 1C HCl etched 4124 Bond interface 2A Sand blasted 6973 Substrate failure 2B Normal 3486 Bond interface 2C HCl etched 2524 Bond interface 3A Normal 3034 Bond interface 3B Sand blasted 5183 Bond interface 3C HCl etched 1386 Bond interface Images from the DIC analysis are shown below in Figures 5, 6, and 7. Although DIC was not originally a part of this project, data from this experiment will be used for future work in analyzing the deformation and strain of each specimen during testing. Figure 5. Strain contour of cube Figure 6. Strain contour of cylinder at failure

10 Mariah Safritt 10 DISCUSSION As can be seen in Table 3 above, Mix 2 (with silica fume and fly ash) was the strongest overall UHPC mix because it had the highest 14 day compressive strength of 11,289 psi. However, for the slant shear test Mix 1 was the strongest with a strength of 7,635 psi, and for the splitting tensile test Mix 2 was the strongest with a strength of 2,841 psi, as seen in Tables 4 and 5 above. It can also be seen from Tables 4 and 5 that the surface preparation that resulted in the highest bond strengths was the sand blasting for each mix design. This is very useful to know, because in real world applications it would be unrealistic to try to use hydrochloric acid as an etchant on an existing concrete structure such as a bridge to prepare the surface for repair by UHPC. Sand blasting is already a practice in use on construction sites, so these results are very helpful in moving forward with the real world applications of this project. Tables 4 and 5 also list the break characterization for each specimen upon failure, and almost all of the specimens broke along the bond interface between the NC and the UHPC. However, two specimens failed in the substrate material (NC) during the slant shear test and two specimens failed in the UHPC during the splitting tensile test. For the slant shear test, this means that those two specimens have bond strengths even higher than the strengths listed in Table 4 because the bonds did not even fail during the test. For the splitting tensile test, this means that the strength of the UHPC in tension was lower than the bond strength between the two materials. It can be also be shown, by comparing Tables 4 and 5, that the specimens had higher strengths in resisting combined shear and compressive stresses (from the slant shear test) than in resisting indirect tensile stresses (from the splitting tensile test). CONCLUSION In conclusion, the UHCP mix that had the highest overall compressive strength was Mix 2, which had silica fume and fly ash in it. For the slant shear test, Mix 1 had the highest strength and for the splitting tensile test Mix 2 had the highest strength. The surface preparation that resulted in the highest bond strengths between the normal concrete and the UHPC was sand blasting. The concrete specimens had higher compressive strengths than tensile strengths, because the strengths from the slant shear tests were much higher than the strengths from the splitting tensile tests. In the future, more research will hopefully be done into being able to use UHPC in thin layer applications, and as a repair material for aging normal strength concrete structures. These tests were very specific and idealized, and in order to determine whether UHPC will be suited to real world applications it will be necessary to do more testing in the field to mimic actual situations. More options for future work could involve using various UHPC mixes that include different types of fibers or more nanoparticles in the UHPC mix.

11 Mariah Safritt 11 ACKNOWLEDGMENTS First of all I would like to thank Professor Devin Harris, who was my faculty advisor and mentor, for all his help and guidance on this project. Dr. Andrei Ramniceanu, who is the lab manager for the Civil Engineering department at UVA, also helped me a lot with preparing and testing my samples and getting everything that I needed to do this experiment. Dr. Amir Gheitasi also was a huge help to me; he set up the DIC equipment for me, helped me with testing, and showed me how to use the software to post-process and analyze my data. I would also like to recognize graduate students Evelina Khakimova, Muhammad Sherif, and Sherif Daghash for their assistance throughout my project. Thanks also go to VCTIR for allowing me to use their space and resources; Mike Burton, Ken, Sam, and many other VCTIR employees were a huge help to me in mixing up my batches of concrete. Finally, I would like to thank Dr. Emily Parkany for her guidance and support throughout this research project, and the MATS UTC Undergraduate Summer Research Program for giving me the opportunity to participate in this research program. REFERENCES 1) Carbonell Muñoz, Miguel Ángel. "Compatibility of Ultra High Performance Concrete as Repair Material: Bond Characterization with Concrete under Different Loading Scenarios." Michigan Technological University, ) Graybeal, Ben. "Development of Non-Proprietary Ultra-High Performance Concrete for Use in the Highway Bridge Sector." Turner-Fairbank Highway Research Center: Federal Highway Administration, ) Naaman, Antoine E., and Kay Wille. "The Path to Ultra-High Performance Fiber Reinforced Concrete (UHP-FRC): Five Decades of Progress." UHPC International Symposium, ) Russell, Henry G., and Benjamin A. Graybeal. Ultra-High Performance Concrete: A State-of-the-Art Report for the Bridge Community: Federal Highway Administration, ) Sarkar, Jayeeta. "Characterization of the Bond Strength between Ultra High Performance Concrete Bridge Deck Overlays and Concrete Substrates." Michigan Technological University, ) Wille, Kay, Antoine Naaman, and Gustavo Parra-Montesinos. Ultra-High Performance Concrete with Compressive Strength Exceeding 150 Mpa (22 Ksi): A Simpler Way, ACI Materials Journal, ) ASTM C39, 2015, Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens, ASTM International, DOI: /C0039_C0039M-15A, 8) ASTM C882, 2013, Standard Test Method for Bond Strength of Epoxy-Resin Systems Used with Concrete by Slant Shear, ASTM International, DOI: /C0882_C0882M-13a, 9) ASTM C496, 2011, Standard Test Method for Splitting Tensile Strength of Cylindrical Concrete Specimens, ASTM International, DOI: /C0496_C0496M-11,