Restored Strength of Cracked Concrete Beam Repaired by Epoxy and Polymethyl Methacrylate

INTRODUCTION

In the last decades, due to the issue of environmental protection, more and more innovation projects were replaced by rehabilitation projects in the U.S. and many developed countries. The importance of concrete repair in the field of concrete construction leads to increasing demands for repair materials and techniques. In 1999, after the Ji-ji earthquake in Taiwan (with a magnitude of 7.3 on the Richter scale), thousands of buildings were demolished or damaged and needed to be rehabilitated. Numerous damaged concrete structures with various types of cracks were repaired. For economical concern, some shearing wall structures and footings several meters long and 10 mm (0.394 in.) wide cracks were rehabilitated by epoxy mortar. It was considered valid and necessary to investigate the mechanical properties of those materials on repairing flexural and shearing cracks of concrete.

Epoxy resin,1 a thermal setting material, is the material most commonly used to repair concrete cracks. Acrylicbased materials are now increasingly being used because they resist the penetration of chloride and carbonation.2-4 One such acrylic-based material, polymethyl methacrylate (PMMA), is a thermal plastic material, which performs well and is durable. This material was used to repair cracked concrete and compared with epoxy mortar in this study.

Assessment of repaired cracked concrete

The restoration of the strength of cracked concrete was measured for various conditions of the concrete surfaces and was investigated by Hindo.5 A comparison among the repair surfaces, including smooth, wire-brushed, needle-punched, and hand-crushed concrete surfaces, was investigated by Abu-Tair et al.6 It was concluded that hand-crushed concrete surfaces turn out the highest adhesion as repaired by epoxy, followed by those of needle-punched, wired-brushed, and smooth. The real crack surface repair, however, seems to be more close to the real case than those aforementioned based on this research. Small7 tested the strength of a specimen cored from cracked concrete that had been repaired using epoxy. Little research, however, is available on the bond strength of repair material and the cracked surface of the concrete. A real cracked concrete beam was used in this study to determine its bond strength after it is repaired using resin materials.

Fracture mechanics of concrete

Cracks always occur before a concrete structure deteriorates or fails. The study of cracks in concrete is referred to as concrete fracture mechanics, which was begun by Kaplan.8 From the fracture mechanics point of view, cracks can be classified as three patterns-opening mode (or Mode I), shearing mode (or Mode II), and tearing mode (or Mode III). Actual cracks, however, are a complex combination of all three modes. A crack in concrete is thought to develop when the fracture toughness, which is always defined quantitatively in terms of the critical stress intensity factor KC or the fracture energy GC, has been exceeded. A number of testing methods have recently been suggested to RILEM,9 such as Hillerburg's10 fictitious model, Jenq and Shah's11 two-parameter model, and Karihaloo and Nalathambi's12 effective crack model, all of which provide methods for determining the fracture properties of concrete. These works all used single-edge notch (SEN) concrete beams, in which crack propagation in the Mode I fracture was investigated. In this study, Mode I or Mode II cracks in concrete are considered. This work examines only the load strengths of the concrete beam before cracking and after repair. The critical stress intensity factor KC and the fracture energy GC will be discussed elsewhere.

Analysis of concrete fracture

Concrete is a brittle material that has a failure type different from that of a ductile material such as a metal. Concrete has a low tensile strength and exhibits strainsoftening behavior, which displays nonlinear stress distribution along the crack path. Thus, conventional theories are valid only when concrete is stressed within its linear stage. In the literature, the crack propagation pattern was simulated by finite element programs with several approaches, such as the lattice model13-14 and discrete crack approach.15-18 For the nonlinear part, the region of strain localization should be modeled using other approaches, such as the fictitious crack model.10 Taha,19 however, applied the fictitious crack model on a finite element code to predict the crack path on the beam. The result turned out virtually identical to those obtained by using linear elastic fracture mechanics (LEFM).

A finite element program18 was used in the analysis. It is an interactive program for two-dimensional analysis of structures. Its capacities include elastic and elastic-plastic material response, and simulation of linear elastic and elastic-plastic crack growth (using the crack tip opening angle approach). It can be developed to dynamically model the propagation of a crack, a structural response, and failure, using interactive computer graphics and an automated remeshing scheme. The analyzed results obtained similar crack patterns as those determined from the experiment. This indirectly verified the previously mentioned aspect that the crack paths are similar for those using LEFM or fictitious crack models on concrete.

In a practical manner, Mode I and Mode II cracks dominated most of the failures of concrete structures. In consequence, the finite element code was employed in this study to predict the path of a crack and to designate the testing configuration to form a unique crack in Mode I or Mode II. This work presents the results of testing cracked concrete that had been repaired using epoxy and PMMA mortars with various fractions of fine sand for cracks of different types and widths.

RESEARCH SIGNIFICANCE

Investigating the restoring efficiency of concrete with a real crack repaired using resin-type material can be challenging. A pair of fractured surfaces of a concrete specimen being created from the fracture mechanics test makes the evaluation of feasibility of various repair materials become possible. This paper demonstrates the results of bond performance of epoxy mortar and PMMA mortar, which include various contents of fine sand in a flexural crack and a shearing crack with different widths. It provides good and practical information for engineers to use these types of materials for concrete repair on site. This can also lead to an innovation of the device to use those materials in the near future.

TESTING PROGRAM

The testing program consists of three portions. There are concrete tests to obtain the mechanical properties of the concrete, repair materials tests to determine the mechanical properties of the materials, and beam tests to investigate the performance of the cracked concrete that had been repaired using each type of repair material.

Properties of repair materials

Epoxy and PMMA were used as repair materials in this testing program. The former is composed of an epoxy resin and a curing agent in the ratio of 2:1. Its setting time is approximately 60 minutes. The storage, handling, mixing, surface evaluation, and preparation as well as its inspection and quality control all conform to ASTM C881, Type IV, Grade 1, Class C. The PMMA used herein is referred to as a methylacrylic monomer system, which contains a polymerization catalyst-dibenzoyl peroxide (BPO) and an activator to initiate a free radical reaction that results in the crosslinking of polymer chains. It will polymerize to become a tough, strong, and durable plastic with greatly enhanced mechanical properties and adhesion to concrete surfaces. The amount of BPO determines the curing duration. Two percent by weight of acrylic resin was used herein, which corresponds to approximately 30 to 40 minutes of working time. PMMA is a liquid with a low viscosity, which soaks into dry concrete, filling the cracks for structural repair. It also exhibits some volatility, toxicity, and flammability, however, which should be carefully employed. Cracks in concrete were repaired using those resins with various sand contents (passing No. 16 sieve)-0, 10, 20, and 40%-and the sand used in the epoxy mortar and the PMMA mortar had a maximum particle size of 1.18 mm (0.047 in.) that was sieved from regular sand.

The fundamental mechanical properties of the repair materials were determined using various test specimens that met the ASTM specifications.20-22 Cubic specimens20 of repair materials were tested to determine their compressive strength. A mold was used to fabricate specimens whose tensile strengths were measured.21 The rupture modulus and the elastic modulus were determined by performing a threepoint bending flexural test on prisms22 with dimensions of 40 x 40 x 160 mm (1.57 x 1.57 x 6.30 in.). All specimens were air-cured for 7 days after they had been cast for 1 day in the mold, under the same conditions as those under which the cracked concrete had been repaired.

Preparation and properties of concrete

In this study, crushed limestone was adopted as coarse aggregate, which had a maximum particle size of 19 mm (3/4 in.) and a fineness modulus of 6.78. Its specific gravity was 2.62 measured from the test. Regular sand with a maximum particle size of 4.75 mm (0.19 in.) and a fineness modulus of 2.43 was used to fabricate all the concrete specimens. Type I portland cement was used, following ASTM C150 or ASTM C595 for conventional concrete.

A unique mixture was adopted throughout the entire testing program. Quartzite was used as aggregates with the maximum grain size of 19 mm (3/4 in.). The water-cement ratio (w/c) was 0.6. Table 1 presents the mixture proportions. The mechanical properties at 28 days were as follows: a compressive strength of 30.5 MPa (4450 psi), a rupture modulus of 3.5 MPa (510 psi), a splitting tensile strength of 2.2 MPa (320 psi), and a modulus of elasticity of 23.7 GPa (3450 ksi). All specimens were designed as a single-edge notch (SEN) beam with the dimensions of length of 350 mm (13.8 in.), a span of 300 mm (11.8 in.), a depth of 100 mm (3.94 in.), and a width of 100 mm (3.94 in.). For the beams in a flexural test, a notch of 2 mm (0.079 in.) wide and 10 mm (0.39 in.) deep was cast in the middle where the cracks are initiated; whereas for those in the shearing test, a notch of 2 mm (0.079 in.) wide and 20 mm (0.79 in.) deep was cast. All beam specimens were cast in plywood forms and kept for 24 hours with a plastic sheet placed over the exposed surface. Companions of ö100 mm x 200 mm (ö3.94 in. x 7.88 in.) cylinders were made to determine their fundamental mechanical properties. All specimens were then demolded and cured in a 100% relative humidity environment for 28 days. Then, each beam was broken and repaired immediately. A 1 mm (0.04 in.) crack in such a beam was sealed, and then filled with epoxy or PMMA resin by low-pressure injection method. A 5 or 10 mm (0.20 or 0.39 in.) crack was filled with epoxy or PMMA mortar with slight vibration. They were cured in the air for 7 days before being tested.

Test methods for concrete beams

All tests were conducted using a hydraulic loading system operating with load control. The beam tests were loaded in a three-point bending or four-point bending manner to generate a flexural crack or a shearing crack. The test configuration was thus designed according to a simulation analysis made using the finite element code. Figure 1(a) depicts a typical loading configuration for a flexural test specimen, and Fig. 1(b) and (c), respectively, show the specimen after it had been loaded analytically and experimentally to elucidate the Mode I cracking of concrete. Similarly, Fig. 2(a) to (c) show the loading configuration and the crack patterns determined from analysis and experiment for a shearing test of concrete specimen. The cracks were repaired later by epoxy or PMMA mortar to examine the efficiency of restoration of cracked concrete.

Repair procedure for concrete crack

Each beam was completely broken into two pieces and then reassembled as the original beam to compare the restoring efficiency for various crack types and widths. Accordingly, a specific crack was ready to be repaired. The repair procedures for a 1 mm (0.04 in.) wide crack differed from those for a 5 or 10 mm (0.2 or 0.39 in.) wide crack. They were kept open at a desired crack width, and then all of the surface cracks were sealed with sealer. A 100 x 20 x 3 mm (3.9 x 0.8 x 0.12 in.) plate attached with two plastic sheets on both sides (easy to peel out later) was stuffed into the notch of the beam with a 1 mm (0.04 in.) wide crack before sealing, such that the crack width was adjusted to 1 mm (0.04 in.) wide, as presented in Fig. 3(a). Meanwhile, two ports were also placed at the corners of the middle of the beam. The resin was injected using a syringe from the bottom port and output from the outlet port to ensure that the crack region could be fully filled with the resin.

The region along the surface with 5 or 10 mm (0.2 or 0.39 in.) wide cracks on the three sides of the beams (but not on the notched side) were attached with plastic plates after the crack width on the three sides had been measured. Then, repair resin mortar was poured into the crack region with little vibration until the level of resin or mortar rose to the bottom of the notch. Thus, all of the fractured surfaces were filled with repair material, as shown schematically in Fig. 3(b).

Practically, a fine crack was sealed and injected by resintype material. For a wide crack, resin mortar does have difficulty immigrating into the crack. The repair technique used in this study was for research purposes only. A device for practical use can be innovated to drive the resin mortar into a wide crack in the concrete.

The sealing of each beam took approximately 2 days. After the surface sealer had set, the repair resin was injected into the fractured region and air-cured for 7 days. The surface sealer had to be removed using a grinder or a chisel and hammer to clean up the beam surface around the sealing area before the test to eliminate the effect of the surface sealer.

TESTING RESULTS

The testing program mainly involves the mechanical properties of repair materials and the evaluation of the restored strength of cracked concrete after repairing have been made. Tables 2 and 3 present the results and are discussed in the following.

Fundamental mechanical properties of repair materials

The mechanical properties of repair materials (epoxy mortar and PMMA mortar) were tested similarly to the ASTM Specifications20-22 and are shown in Table 4. It is noted that the compressive strengths of both repair materials increase with the growing amount of sand content, as displayed in Fig. 4. Similarly, the tensile strengths of PMMA mortar also increase with the growing amount of sand content, as shown in Fig. 5. It is noticed that pure epoxy results in the highest tensile strength. This can be attributed to the inclusion of sand that alters the failure type of the material. The change of its constituent from purity to complexity can induce stress concentration. Accordingly, the tensile test of both repair materials reveals that the specimens yield under pure resin, but the mortars fracture in their ultimate stage.

The three-point bending flexural test showed that the rupture moduli of both materials also increase with the increase of sand content, as shown in Fig. 6. The failure of the pure resins, but not the mortars, appears to be by yielding rather than fracturing, which are in the same manner in the tensile test. Epoxy mortar, however, has a higher rupture modulus than PMMA mortar for a given sand content. Meanwhile, the elastic moduli of those repair materials were also determined in the three-point bending test and from the recorded load and the data from a strain gauge mounted on the middle of the bottom of the prism. The moduli of elasticity of both materials also increased with the increase of sand content, as shown in Fig. 7. PMMA mortar with 20% sand inclusions had a higher modulus of elasticity than that with 40% sand inclusions.

Restoration of strength of repaired concrete

All of the beams were broken into two pieces before repairing to obtain two fractured surfaces. Accordingly, the areas of the fractured surfaces of the beams with the same type of crack are supposed to be equal. The quality of the test results, however, was controlled using three specimens for each test, including the original beam and the repaired beam. The results in Tables 2 and 3 indicate a maximum coefficient of variation of 8.3%, revealing good consistency among the results.

Tables 2 and 3 show the test results of the restored strength of concrete with a flexural crack and a shearing crack, respectively. The restored strength of concrete with a flexural crack of various widths increased with the sand content in the epoxy mortar, as displayed in Fig. 8. The restored strength of a concrete beam with a crack of 1 mm (0.04 in.) wide was approximately 102% of that before the beam cracked; however, the corresponding values for concretes with cracks of 5 and 10 mm (0.20 and 0.39 in.) wide were 52% and 73%, respectively. The latter two values were increased to 112% and 159%, respectively, as the sand content in the mortar was increased to 40% by volume. The apparent growth in the repair efficiency of the SEN beams were attributed to the repair materials that enhanced the crack resistance around the crack tip as well as their interlock effect with the fractured concrete surfaces. The repair materials, which conduct a ductile behavior and perform higher rupture modulus than concrete, may blunt the stress and form a plastic zone around the crack tip near notch bottom. This may explain the increase of the load capacity of the SEN beam after repair.

The effect of sand content in the PMMA mortar used to fill cracks in concrete is similar to that in the epoxy mortar. Figure 9 reveals that the restored strength of concrete beam with a crack of 1 mm (0.04 in.) wide is approximately 107% of that before the beam cracked; however, the corresponding values for cracks of 5 and 10 mm (0.20 and 0.39 in.) wide are 93% and 104%, respectively. These latter values can be increased to 144% and 153% as the sand fraction in the mortar increases to 40%.

Figure 10 compares the restoring efficiency for both repair materials in a flexural concrete crack. Notably, the restoring efficiency of epoxy resin is close to that of PMMA resin when used to repair concrete with a crack of 1 mm (0.04 in.) wide, but the restoring efficiency obtained with PMMA resin is higher for concrete with cracks of 5 and 10 mm (0.20 and 0.39 in.) wide. PMMA mortars provide a higher restoring efficiency than epoxy mortar when used to repair a crack 5 mm (0.20 in.) wide in concrete for a given sand content. The restoring efficiencies are similar for both mortars when used to repair the concrete crack of 10 mm (0.39 in.) wide.

Figure 11 demonstrates that the restored strengths of concretes with a shearing crack of various widths increase with the sand content in the mortar. The restored strength of the concrete with a shearing crack of 1 mm (0.04 in.) wide is approximately 66% of that before the concrete beam cracks; however, the restoring efficiencies are only 30% and 28% for concretes with cracks of 5 and 10 mm (0.20 and 0.39 in.) wide, respectively. The latter values can be raised up to 57% and 73%, respectively, by increasing the sand content in the mortar to 40%.

Figure 12 indicates that when a shearing crack is repaired using PMMA, the restored strength of concrete with a crack of 1 mm (0.04 in.) wide is only 71% of that before the beam cracking. These values, however, are 44% and 55% for concrete with a crack of 5 mm (0.20 in.) wide and that of 10 mm (0.39 in.) wide, respectively. They can be increased to 79 and 106% by increasing the sand content in the mortar to 40% by volume. The restoring efficiencies of both materials when used to repair a flexural crack exceed those when used to repair a shearing crack.

Figure 13 compares the restoration efficiencies of both repair materials in concrete with a shearing crack and reveals that the restoring efficiency obtained when PMMA resin or mortar is used to repair concrete with a 1, 5, and 10 mm (0.04, 0.20, and 0.39 in.) crack all exceed that repaired by epoxy mortars for a given sand content. Except in the case of the PMMA mortar with 40% sand content, the restoring efficiency of all repair materials appear to be less than 100% when they are used to repair a shearing crack, implying that those repair materials have poorer restoration performance for shearing cracks than for flexural cracks.

Fractured surface of repaired concrete

The method of repairing concrete cracks that is used herein ensures good quality control as the repair material completely fills the fractured surface. The effectiveness of the repair method can be observed from the fractured cross sections of the repaired concrete beams. The fractured surface of cracked concrete with cracks of various widths that were repaired by resin-type repair materials was examined. Figure 14 shows that the sections obtained from the concrete with repaired flexural cracks of 1 mm (0.04 in.) wide. Both cracked concretes being repaired by epoxy (left) and PMMA (right) resins all exhibit newly-cracked surfaces. This finding reveals a good bond quality for both epoxy and PMMA resin in a fine flexural concrete crack. In repairing concrete with a 1 mm (0.04 in.) wide shearing crack, as displayed in Fig. 15, the crack propagates through the epoxy resin on the left, but the crack propagates through the concrete to produce a new fractured surface for repairing PMMA resin on the right of the figure.

Flexural cracks that were 5 and 10 mm (0.20 and 0.39 in.) wide in the repaired concretes are similar, as displayed in Fig. 16: a new crack propagates along the interface between the concrete and epoxy in concrete that has been repaired by epoxy resin. The proportion of new fracturing area in the cracked section increases with the sand content. A new fractured surface in concrete is formed when old concrete cracks were repaired using epoxy mortar with 40% sand content. The newly fractured surfaces of repaired shearing cracks still turned out a shearing-type crack and were formed near the first shearing crack, as displayed in Fig. 17. The feature of those fractured surfaces was similar to those obtained from flexural cracks.

Figure 18 depicts the flexural crack surfaces of concrete with a crack with a width of 5 mm (0.20 in.) or 10 mm (0.39 in.), repaired with PMMA resin or mortars. Notably, a crack propagates through pure PMMA resin; the newly fractured area in concrete becomes larger as the sand content in the mortar increases. Figure 19 depicts the fractured surfaces of concrete with a shearing crack that was repaired using PMMA mortar. Their fractured surfaces still turned out a shearing-type crack, which is close to the previous crack. The feature of those fractured surfaces appears similar to those obtained from flexural cracks.

In this study, the concrete crack with 1 mm (0.039 in.) wide is defined as a narrow crack and designed to investigate the restoration efficiency for actual fine cracks in concrete. For reinforced concrete structures, corrosion protection is the main concern for crack repair on concrete structures. Nevertheless, this research has made an attempt to figure out the recovery of mechanical property on concrete after the crack being repaired, which may help engineers to understand more for the rehabilitation of concrete structures.

CONCLUSIONS

From the test results in this study, several conclusions can be drawn as follows:

1. Based on sand particle sizes passing No. 16 sieve included in resin mortar for wider crack (5 or 10 mm [0.12 or 0.39 in.]), the compressive strength, the tensile strength, the modulus of rupture, and the modulus of elasticity of epoxy and PMMA mortars generally increase with the sand content. In addition, the strengths of epoxy mortars exceed those of PMMA mortars. The failure of the pure epoxy resin and PMMA resin was by yielding rather than fracturing;

2. The restoring efficiency of both epoxy mortars and PMMA mortars used to repair concrete with a flexural crack exceed that with a shearing crack. Moreover, PMMA provides a similar restored strength as epoxy when used to repair a flexural crack, but a higher restored strength than epoxy when used to repair a shearing crack; and

3. Epoxy and PMMA resins, when used to repair a concrete crack 1 mm (0.04 in.) wide, has the highest restored strength than for those crack 5 or 10 mm (0.20 or 0.39 in.) wide. For the concrete with a crack of 5 or 10 mm (0.20 or 0.39 in.) wide, increasing the sand inclusion content in the mortar increases the restoring efficiency, allowing good workability to be achieved.

ACKNOWLEDGMENTS

The authors would like to thank the National Science Council of the Republic of China, Taiwan, for financially supporting this research under Contract No. NSC-91-2211-E-324-020.