Bond between Reinforcement and Concrete Influence of Steel Corrosion

Transcription

1 Bond between Reinforcement and Concrete Influence of Steel Corrosion Mohammed Sonebi 1 Richard Davidson 2 David Cleland 3 ABSTRACT The investigation into the deterioration of the bond between concrete and reinforcement was carried out using a series of beam-type pullout specimens. The specimens were designed to allow investigation into cover depths of 20 mm and 50 mm and bar diameters of 12 mm and 20 mm. The beam-type specimens also allowed the relative bond strength developed at top and bottom bars to be examined. The compressive strength of the concrete was kept constant throughout the study. Factors that cause the bond strength to be lower include; reduction in cover depth, increasing bar diameter and location near the top of a section. Comparisons were drawn between the design bond stress used for anchorage length in the Eurocodes and the results of this study. Applied anodic current was used to accelerate the corrosion. This method resulted in uniform corrosion along the rebar. Three levels of corrosion were investigated and as the corrosion levels increased cracks developed and were measured in each sample. Based on the results of this study the bond strength decreased as corrosion level increased. At severe corrosion level, the improvement of bond strength with higher cover was less significant. The Eurocodes underestimate the bond strength where no corrosion is present by 20%, while samples with severely corroded reinforcement exhibited bond strength only 60% of the design bond strength. In general, the bond strength was lower for bars of larger diameter but this was less significant for severe corrosion. KEYWORDS Bond strength, Bar diameter, Corrosion, Cover level, Top-bar effect. 1 School of Planning, Architecture and Civil Engineering, Queen s University, Belfast, BT7 1NN 2 School of Planning, Architecture and Civil Engineering, Queen s University, Belfast, BT7 1NN 3 School of Planning, Architecture and Civil Engineering, Queen s University, Belfast, BT7 1NN

2 Mohamed Sonebi, Richard Davidson and David Cleland 1 INTRODUCTION Among the pathological manifestations that accelerate the performance loss of reinforced structures, rebar corrosion is one the most important. Steel corrosion of reinforced bars leads to cracking, reduction of bond strength, reduction of steel cross section and loss of serviceability life of a concrete structure [Mangat and Elgarf 1999]. The factors which alter the rate of corrosion include; availability of oxygen, humidity level, permeability of the concrete, chloride content, temperature, and depth of carbonation [Vennesland et al. 2007]. These factors if controlled can limit the damage caused by corrosion. Regions such as the Arabian Gulf is one of the most aggressive environments where some of the structures need significant repairs in just years [Almusallam et al. 1996]. Pullout of the reinforcing bars associated with bond characteristics is one of the main factors affecting the ultimate behaviour and failure of reinforced concrete members. It is known that bond strength between concrete and reinforced steel bars is usually much lower for reinforcement located near the top of a cast section than for those near the bottom, which is known as the top-bar effect (ACI 318). The top bar effect is assumed to be due to a weakened interfacial transition zone (ITZ) caused by bleeding and settlement in the fresh concrete mix. The objective of this investigation is to study the effect of corrosion on the bond strength of reinforced bars. Three levels of corrosion have been used and the effect of the cover level was examined together with different bar diameters on the bond strength of reinforced steel bars. In this paper, a total of 24 specimens of concrete with reinforcing steel bars were cast with differing cover level and different bar diameters using a set-up suggested by Ghana [1996]. The specimens, measuring 300 mm deep by 200 mm wide by 300mm long were subjected to three levels of corrosion, severe, moderate and without corrosion. The specimens without corrosion were used as control specimens. 2 EXPERIMENTAL PROGRAM 2.1 Materials All mixes were prepared using ordinary Portland cement (CEM II) with a water/cement ratio of The total cement content was kept constant at 425 kg/m 3. The fine aggregate was natural siliceous sand with a fineness modulus of 2.9, a saturated surface dry specific gravity of 2.60 and water absorption of 1.5%. The total content of sand was 730 kg/m 3. Crushed basalt with a maximum nominal size of 10 and 20 mm, a saturated surface dry specific gravity of 2.65 and water absorption of 0.8%was also used. The dosages of 10 mm and 20 mm aggregates were 505 and 505 kg/m 3, respectively. The NaCl solution was used in mix at 10% of cement content (42.5 kg/m 3 ). 2.2 Method of Acceleration of Corrosion The corrosion in the samples was accelerated by setting up a potential difference between the reinforced steel bars and a nearby cathode. The electrolyte in the cell is the concrete pore solution containing sodium chloride (NaCl), in this case added at the mixing stage. The Cl- ions from the sodium chloride are then attracted to the anode to accelerate the corrosion. The circuit setup can be seen in Fig XII DBMC, Porto, PORTUGAL, 2011

3 Bond between Reinforcement and Concrete Figure 1. Circuit setup for accelerated corrosion. The potential difference was applied first to the specimens which were to be moderately corroded. The level of corrosion was deemed to have been achieved when a longitudinal crack of width 1.5 mm developed. This took three weeks at one volt and a further four weeks at three volts. The other eight specimens were subjected to a potential difference of eight volts for a period of four weeks to achieve severe corrosion and this led to longitudinal cracks of 8 mm in width. The method used to measured corrosion was by weighting the loss of the bar after it had been removed from the sample. 2.3 Method of Pullout Test of Specimen The pullout specimens were prepared with the horizontal reinforced bars protruding from the vertical surface on the side of the specimens. The bond properties between reinforced bar and concrete were studied by conducting direct pullout test of reinforced bars embedded in the specimens. During the experimental study, two different covers were used (20 and 50 mm). The positions of reinforced bars at the time of casting and the direction of concrete casting are shown in Fig. 2. There were eight moulds with the dimensions of 300 mm 200 mm 300 mm and six holes for each mould to insert the reinforced bars of which three holes at the top and three at the bottom. Four moulds were manufactured for the 12 mm diameter bars and another four for the 20 mm diameter bars; two of each set of four moulds had 20 mm and two had 50mm cover spacing. Deformed reinforcing steel bars having 15 mm nominal diameter and 400 MPa nominal yield strength were used for pullout tests. Steel bars were cleaned with a wire brush to remove any rust from the surface just before casting the concrete specimens. After casting specimens were sealed with wet burlap and stored in a temperature controlled environment. All specimens were demoulded after 24 h and transferred to a laboratory environment until they were tested. The test method used in the study was similar to that proposed by Chana [1990] where more detailed information on the pullout test setup can be found. The direct pullout test setup is shown in Fig. 2 and single reinforced bar pullout tests were conducted. A load cell, a hydraulic jack, a mounting system, and a linear variable differential transducer (LVDT) were assembled as shown in Fig. 2. The hydraulic jacking system was adopted to apply a pullout force to the reinforcing bar. When bond slip occurs, the displacement of the reinforced bar can be obtained from the measurement of the LVDT attached to the reinforced bar. During testing, the pullout load and reinforced bar slip displacement were measured and recorded automatically. The rate of loading used was increased in increments of 2 kn/s for the specimens with moderate and no corrosion. It was considered more reliable to take results every 1 kn/s with the severely corroded specimens as their failure loads were considerably smaller. XII DBMC, Porto, PORTUGAL,

4 Mohamed Sonebi, Richard Davidson and David Cleland Cover 20/50 mm Cover 20/50 mm Bar crip Hydraulic jack Bar Diameter 12/20mm 150 mm 300 mm Cathode 25mm 200 mm Figure 2. Experimental setup used for pullout test. 3 TEST RESULTS AND DISSCUSSION 3.1 Effect of Top-bar Figure 3 presents the variation of the bond strength with the top-bar effect and level of corrosion. As expected, the bond strength reduced as the corrosion level increased. In concrete without accelerated corrosion, there was a 25% reduction in bond strength due to the top-bar effect [Soylev et al. 2003]. Chana [1990] compared the top and bottom cast bars using a similar setup of specimen and concluded a comparable reduction of the bond strength for concrete having the same compressive strength. The corrosion at moderate level (between 0.5 and 1.5% weight loss) did not show a significant difference in bond strength for different positions of the bars. Whilst the specimens undergone a severe corrosion levels (3 to 4% weight loss) a top bar bond strength 30% lower than the bottom bar. It can be concluded that the top-bar effect is highly influenced by the level of the corrosion of reinforced bars. The phenomenon of top bar effect facilitates the penetration of oxygen, water and chlorides to the interface between the concrete and steel reinforcement leading to an increase of the rate of corrosion. 4 XII DBMC, Porto, PORTUGAL, 2011

5 Bond between Reinforcement and Concrete 4.5 Bond Strength (MPa) Top Bottom 0.0 No Corrosion Moderate Corrosion Severe Corrosion Figure 3. Variation of bond strength between top and bottom reinforcement. 3.2 Effect of Cover and Level of Corrosion on Bond Strength The results illustrated in Fig. 4 show consistently that increasing the cover from 20 mm to 50 mm had a positive impact on the bond strength at all levels of corrosion. The 50 mm of cover exhibited an improvement of bond strength by 30% for no corrosion level, 66% for moderate corrosion, and 25% for severe corrosion level, respectively, compared to 20 mm cover. The failure mechanism remained split failure due to the longitudinal corrosion cracks, irrespective of the cover level. During testing the hoop tension forces in the concrete surrounding the reinforced bar was best resisted by the greater confining 50 mm cover. It was interested to note that the bond strength difference between cover levels was lower where the corrosion had reached severe level (3 4%). This is due to the corrosion cracks in the specimen which result in the loss of the confinement effect of the cover concrete Bond strength (MPa) C = 20 mm C = 50 mm 0.0 No Corrosion Moderate Corrosion Severe Corrosion Figure 4. Effect of cover on bond strength of three levels of corrosion. XII DBMC, Porto, PORTUGAL,

6 Mohamed Sonebi, Richard Davidson and David Cleland 3.3 Effect of Diameter on Bond Strength with Different Levels of Corrosion It is illustrated in Fig. 5 that as the corrosion level was increased, the bond stress decreased as expected but the difference of bond strength between the two bar diameters was reduced significantly. The increase in diameter of the reinforced bar has a significant impact on the bond strength. The specimens with no corrosion had strength for 20 mm bars 45% lower than the strength for 12 mm bars. These bond strength results can be compared with the guidelines given in Eurocode 2 which, for the set of conditions considered, recommend ultimate bond strength 1.79 MPa, and this value remains valid for bar diameters up to 32 mm. In the specimens containing 12 mm diameter bars with no corrosion, the Eurocode underestimated the experimental bond strength by a factor of almost three. Even when the samples of reinforced concrete were corroded to a moderate level ( % weight loss), the 12 mm diameter reinforced bars had a bond strength greater than the value calculated by the Eurocode [BSI Eurocode ], while the bond strength with 20 mm diameter reinforced bar was below the design value recommended by the Eurocodes. Hence the results illustrate how the durability of structures can be influenced by the choice of diameter of reinforcing bar Bond strength (MPa) D = 12mm D = 20mm 0.0 No Corrosion Moderate Corrosion Severe Corrosion Figure 5. Effect of bar diameter on bond strength of three levels of corrosion. 3.4 Effect of Crack Width on Bond Strength with Different Levels of Corrosion The results in Fig. 6 show the relationship between the actual corrosion level and crack width. The actual corrosion level was obtained using the mass loss method similar to the research reported by Ouglova et al. [2008]. The corrosion level was then calculated as a percentage of the weight of corroded material per unit length of bond [Fang et al. 2006]. Figure 6 demonstrates that as the corrosion level increased the crack width increased. This finding confirms some results reported in literature which suggest a linear correlation between the loss of cross section area and the corrosion level [Apostolopoulos et al. 2006]. This is due to corrosion products resulting in an increase in volume and thus causing a strain within the concrete [Cabrera 1996]. The most obvious cracks are the longitudinal cracks parallel to the reinforcement. 6 XII DBMC, Porto, PORTUGAL, 2011

7 Bond between Reinforcement and Concrete y = x R² = 0.91 Crack width (mm) Actual Corrosion level % Figure 6. Relationship between crack width and increasing corrosion level. 4 CONCLUSIONS This study discusses an experimental program of bond strength of reinforcing bars embedded in normal concrete and subjected to corrosion. The effects of corrosion, two cover depths and two bar diameters were studied. Based on the results of this paper the following conclusions can be drawn: As expected, the increase of level of corrosion led to a reduction in the bond strength of steel reinforcing bars. The bond strength of top bars was 25% lower than bottom bars for specimens without any corrosion, and 30% for bars with severe corrosion. However, in case of moderate corrosion, there was no signifcant difference of bond strength between top and bottom bars. The increase of the cover of concrete from 20 to 50 mm resulted in an improvement of the bond strength for the three levels of corrosion considered in this study. However the difference of bond strength between both cover was less pronounced for the severe corrosion level. The increase of the diameter of bars led to a reduction of the bond strength. This reduction was highly significant for specimens without corrosion, followed by specimen having moderate corrosion. However, in case of severe corrosion, the effect of diameter on bond strength was only marginal. Corrosion cracking is critical for the bond strength. Very little corrosion after cracking is required to lead to a reduction of bond strength to an unacceptable level. REFERENCES Almusallam, A., Al-Gahtani, A., Aziz, A. Rasheeduzzafart, Effect of reinforcement corrosion on bond strength, Construction and Building Materials, 10, Apostolopoulos, CA., Papadopoulos, MP., Pantelakis, SG. 2006, Tensile behavior of corroded reinforcing steel bars BSt500s, Construction and Building Materials, 20, ACI Committee 318, Building code requirements for structural concrete (ACI ), American Concrete Institute, Farmington Hills, Michigan, 151 p. XII DBMC, Porto, PORTUGAL,

8 Mohamed Sonebi, Richard Davidson and David Cleland British Standards Institution. 2002, Eurocode 2, Bond stress. BSI: London, BS EN Cabrera, J.G., 1996, Deterioration of concrete due to reinforcement corrosion. Cement and Concrete Composites, 18, Chana, P.S., 1990, A test method to establish realistic bond stress, Magazine of Concrete Research, 42, Fang, C., Lundgren, K., Plos, M., Gylltoft, K., 2006, Bond behaviour of corroded reinforcing steel bars in concrete, Cement and Concrete Research, 36, Mangat, P.S., Elgarf, S.M., 1999, Bond characteristics of corroding reinforcement in concrete beams, Materials and Structures, 32, Ouglova, A., Berthaud, Y., Foct, F., François, M., Ragueneau, F., Petre-Lazar, I., 2008, The influence of corrosion on bond properties between concrete and reinforcement in concrete structures, Materials and Structures, 41, Soylev, T.A., François, R., 2003, Quality of steel-concrete interface and corrosion of reinforcing steel, Cement and Concrete Research, 33, Vennesland, Ø., Raupach, M., Recommendation of RILEM TC 154-EMC: Electrochemical techniques for measuring corrosion in concrete- measurements with embedded probes, Materials and Structures, 40[8], XII DBMC, Porto, PORTUGAL, 2011