Pullout tests are a fundamental method used to evaluate the bond strength between reinforcing bars and concrete. In a pullout test, a steel bar is embedded in a concrete specimen, and a tensile force is applied to the bar until it pulls out from the concrete. This test provides crucial information about the bond characteristics.
3.1 Specimen Preparation
The pullout specimens were prepared following established standards, with traditional cylindrical specimens measuring 150 mm in diameter and 300 mm in height. The steel reinforcement was embedded into the cylinder up to a depth of 200 mm, with an additional 200 mm extending outside the cylinder. While IS2770 (Part 1) − 1967 (Reaffirmed 2007) recommends using cube specimens for pullout tests, it limits the maximum embedding depth to 150 mm. Recognizing this constraint, the decision was made to utilize cylindrical specimens to achieve deeper embedment depths, thus allowing for a more comprehensive evaluation of bond strength and behavior. This approach ensures adherence to testing standards while maximizing the depth of embedding, as illustrated in Fig. 1.
Four different bundling configurations were examined in the pullout tests conducted as part of the study. The configurations included single bundling with a single 16 mm diameter rod, bundling with one 12 mm rod and one 10 mm rod, bundling with two 10 mm rods and one 8 mm rod, and bundling with four 8 mm rods. These variations were chosen to investigate the influence of the number and diameter of rods within the bundle on pullout strength and behavior.
Table 1 Reinforcement Configuration Details
All the bundling configurations shown in Table 1 were designed to maintain approximately the same cross-sectional area, ensuring consistency in the total amount of reinforcement across the different specimens. A total of 12 specimens were cast, with three specimens for each bundling configuration. These specimens were designed to provide a comprehensive understanding of the pullout behavior of bundled bars under different configurations. The arrangement of these specimens is depicted in Fig. 2.
3.3 Pull-out Test Results
The load-slip curves depicted in Fig. 4 reveal several insights into the behavior of the bundled bar specimens during pullout testing. It is evident that the maximum slip occurs at approximately 3 mm for all specimens. Furthermore, all specimens exhibit a similar trend with slight variations. The consistent trend observed across all specimens implies that the bundling configurations tested have a similar influence on the load-slip behavior during pullout testing. This suggests that variations in bundling, such as the number and diameter of bars, do not significantly affect the development length or bond behavior in the pullout test setup under the given experimental conditions.
The analysis of bundled bars, categorized into configurations 1#, 2#, 3#, and 4#, reveals significant variations in their mechanical performance. Configurations with a higher number of bars in the bundle generally exhibit superior peak strength, with configuration 4# demonstrating the highest peak strength of 60.15 kN, followed closely by configuration 2# at 57.9 kN. Conversely, configurations with fewer bars, such as 1# and 3#, display comparatively lower peak strengths of 54.8 kN and 46.35 kN, respectively. However, this trend is accompanied by differences in peak slip, where configurations with fewer bars tend to experience lower peak slip values. For instance, configuration 2# exhibits the lowest peak slip at 2.868 mm, while configuration 3# demonstrates the highest peak slip at 3.59 mm as shown in Table 2. These findings suggest that while increasing the number of bars in a bundle enhances peak strength, it may also result in higher displacement before failure.
Table 2
Analysis Results of Pull-out Test
Analysis Parameters | 1# | 2# | 3# | 4# |
Peak Strength (kN) | 54.80 | 57.90 | 46.35 | 60.15 |
Peak Slip (mm) | 3.19 | 2.87 | 3.59 | 3.11 |
Secant Stiffness (kN/mm) | 16.96 | 21.58 | 13.5 | 19.92 |
Total Energy Absorbed (kN-mm) | 72 | 82 | 81 | 102 |
The analysis of secant stiffness provides insights into the structural stiffness of bundled bars at different load levels. Secant stiffness, calculated at 80% of the peak load. Configurations with higher secant stiffness values, such as 2# and 4#, demonstrate greater, suggesting their ability to maintain structural integrity under high load levels. Conversely, configurations with lower secant stiffness values 3# may exhibit more significant deformations or displacements for the same increase in load, indicating reduced stiffness and potentially compromising structural stability. On the other hand, the analysis of total energy absorbed provides insights into the ductility and energy dissipation characteristics of bundled bars during loading. Configurations with higher total energy absorbed values, such as 4#, demonstrate enhanced ductility or toughness, indicating their ability to absorb greater amounts of energy before failure occurs. The normalized bar chart is shown in Fig. 5 for better understanding.
The findings from this study challenge the conventional perception prescribed by IS456:2000 regarding the increase in development length for bundled bars. According to Cl. 26.2.1.2 of the code, the development length can be increased by 10% for two bars in contact, 20% for three bars in contact, and 33% for four bars in contact. However, the results of this study suggest that such increases in length may not be necessary and could potentially be counterproductive. The study reveals that bundled bars actually outperform single bars in terms of development length, indicating that the additional bars in the bundle contribute positively to bond strength rather than diminishing it. Specifically, the findings show that configurations with bundled bars exhibit similar or even improved development lengths compared to single bars, despite having multiple bars in contact. This suggests that the presence of bundled bars enhances bond strength between the steel reinforcement and the surrounding concrete, leading to more efficient load transfer and improved structural performance. Bundling indeed increases the surface area of contact between the bundled bars and the surrounding concrete. This increased contact area allows for more effective transfer of forces between the reinforcement and the concrete, enhancing the bond strength. With more bars bundled together, there are more points of contact with the concrete, resulting in a denser and more interconnected interface. As a result, the mechanical interlock between the steel reinforcement and the concrete is strengthened, leading to improved load transfer and greater resistance to slippage or pullout.
The comparison of predicted to experimental ratios for bond strength predicting equations as shown in Table 3, including those from ACI Committee 408, IS456:2000, Orangun et al. (1977), Darwin et al. (1977), and Zuo et al. (2000), reveals notable insights into their predictive accuracy. The ACI equation demonstrates a moderate overestimation (predicted ratio: 1.44), while the IS456:2000 equation significantly underestimates bond strength (predicted ratio: 0.84). Conversely, the Orangun et al. equation shows a considerable overestimation (predicted ratio: 1.55), indicating potential conservatism. The Darwin et al. equation moderately overestimates (predicted ratio: 1.37), and the Zuo et al. equation exhibits similar behavior (predicted ratio: 1.42). These discrepancies highlight the necessity of validation against experimental data and potential adjustments to enhance the accuracy of bond strength predicting equations.
Table 3
S. No. | Code/Literature | Bond Strength (Tb) | Predicted (kN) | \(\:\frac{\varvec{P}\varvec{r}\varvec{e}\varvec{d}\varvec{i}\varvec{c}\varvec{t}\varvec{e}\varvec{d}}{\varvec{E}\varvec{x}\varvec{p}\varvec{e}\varvec{r}\varvec{i}\varvec{m}\varvec{e}\varvec{n}\varvec{t}\varvec{a}\varvec{l}}\) |
1 | ACI Committee 408 (FPS)[2][3] | \(\:{f}_{c}^{1/4}\left[59.9{l}_{d}\left({C}_{min}+0.5{d}_{b}\right)+2400{A}_{b}\right]\left(0.1\frac{{C}_{max}}{{C}_{min}}+0.9\right)\) | 78.94 | 1.44 |
2 | IS456:2000[8] | \(\:{\tau\:}_{bd}{l}_{d}\pi\:{d}_{b}\) | 45.84 | 0.84 |
3 | Orangun et al. (1977) (FPS)[11] | \(\:{f}_{c}^{1/2}\left[3\pi\:{l}_{d}\left(3{C}_{min}+0.4{d}_{b}\right)+200{A}_{b}\right]\) | 85.146 | 1.55 |
4 | Darwin et al. (1996) (FPS)[5] | \(\:{f}_{c}^{1/2}\left[6.67{l}_{d}({c}_{min}+0.5{d}_{b})\left(0.08\frac{{C}_{max}}{{C}_{min}}+0.92\right)+300{A}_{b}\right]\) | 75.4 | 1.37 |
5 | Zuo et al. (2000) (FPS)[15] | \(\:{f}_{c}^{1/4}\left[58.8{l}_{d}\left({C}_{min}+0.5{d}_{b}\right)+2350{A}_{b}\right]\left(0.1\frac{{C}_{max}}{{C}_{min}}+0.9\right)\) | 78.366 | 1.42 |
In the experimental calculations, the design bond stress specified by IS456:2000, Clause 26.2.1.1 was utilized, with an additional factor of 1.5 applied. In accordance with IS456:2000, the design bond stress is provided with an inherent safety margin already incorporated. Therefore, when conducting experimental calculations, this safety factor is taken into consideration by directly using the specified design bond stress without any additional multiplication factor.