Closure to “experimental Investigation of Pullout Behavior of Fiber-Reinforced Polymer Reinforcements in Sand” by Cheng-Cheng Zhang, Hong-Hu Zhu, Bin Shi, Fang-Dong Wu, and Jian-Hua Yin
Cheng-Cheng Zhang,Hong-Hu Zhu,Bin Shi,Fang-Dong Wu,Jian-Hua Yin
DOI: https://doi.org/10.1061/(asce)cc.1943-5614.0000593
2015-01-01
Journal of Composites for Construction
Abstract:The discussed paper presents a comprehensive experimental investigation of the pullout behavior of fiber-reinforced polymer (FRP) reinforcements buried in sand. A series of pullout tests were performed on glass-FRP (GFRP) and carbon-FRP (CFRP) bars and strips under low normal pressure. The authors should be complimented for providing a detailed paper that proposes a simple model based on an elastoplastic interface shear stress-strain relationship to simulate the progressive pullout failure. This is of interest for the discusser, who would like to thank the authors for that, and to offer some comments and questions for their consideration and response, mainly about the theoretical interpretation of the progressive interface failure between FRP reinforcements and sand and the derivation of the simple pullout model. Fig. 5 (all cited figures and equations in this discussion refer to the original paper) shows the idealized elastoplastic shear stressstrain relationship for the FRP-sand interface. As depicted, the shear stress (τ ) first increases linearly with the shear strain (γ) until the shear resistance (τmax) is reached for a certain interface shear strain (γ1), and then it remains constant for γ > γ1. This ideal elastoplastic constitutive relationship was used by the authors as the basis of a simple model that considered three consecutive stages to describe the pullout behavior of an FRP reinforcement in sand: an initial pure elastic stage (Stage I), a transitional elastoplastic stage (Stage II), and a final pure plastic stage (Stage III). However, the discusser is confused about the theoretical interpretation of the authors as detailed in the following. For Stage I, the authors assumed that no debonding or slippage occurs, and then the displacement of the FRP reinforcement is a result of the shear deformation of a thin layer of sand (i.e., shearing band) at the FRP-sand interface. The discusser agrees with that and believes that at this stage, only a low-level interface shear stress is mobilized along a certain embedded length, which should offer reduced pullout force and pullout displacement. However, the Stage I defined by the authors offered more than 50% of the pullout force, as shown in Fig. 7(a), and shear stresses were considered along the entire embedded length in all cases regardless of the pullout force, as shown in Fig. 7(b). Therefore, in accordance with the elastoplastic τ − γ relationship depicted in Fig. 5, the discusser notes than Stage I should include all cases with τ ≤ τmax along the corresponding length of the elastic zone (LE) from the reinforcement head. For τ 1⁄4 τmax, LE would be the maximum length (LEM) with pure elastic behavior, and the subsequent pullout force would coincide with the tensile force at the transition point TðFTÞ between the elastic and plastic zones, as considered by the authors within Stage II. Following the theoretical interpretation by the discusser, once the LEM is activated, the pullout process progresses by propagating LEM from the reinforcement head to the tail as a wave. As a result, new shear stresses are mobilized along the embedded length, which includes LEM and a length corresponding to the activated plastic zone (LP). This elastoplastic stage finishes when LEM reaches the reinforcement tail, τ being τmax in almost the entire embedded length (L). Therefore, as a uniform shear stress distribution implies a linear variation in pullout force, this process covers linear pullout force increases from FT to the maximum pullout force (Pmax). As in the discusser’s interpretation, Pmax is reached at the end of Stage II [as defined by the authors and depicted in Fig. 7(a)], whereas there are differences concerning the beginning of Stage II and on how τ should be computed in Stages I and II. If the discusser is right, the different assumptions with respect to the length of the elastic zone (L or L − LP for the authors, and LE or LEM for the discusser) would lead to the following changes in the derivation of the pullout model for consistent formulation with the discusser’s interpretation: (1) in Eqs. (10)–(14), L should be replaced with LEM, and P with FT ; (2) the condition of [LP ≤ x ≤ L] for the elastic zone in Stage II should be replaced with [LP ≤ x ≤ ðLP þ LEMÞ ≤ L]; and (3) L should be replaced with LEM þ LP in Eqs. (15)–(18) and (21). In addition, the discusser notes that the authors’ statement on Eq. (23), which shows a nonlinear relationship between the pullout force and the pullout displacement when LP increases, does not seem to agree with the bilinear pullout behavior shown in predicted curves (as depicted in Fig. 4). There are no differences with respect to Stage III, which in the discusser’s interpretation is preceded by a transitional phase (within Stage II) from the case of LEM þ LP 1⁄4 L to the case of LP 1⁄4 L, with a pullout force increase equal to 2WτmaxLEM − FT (W = diameter D of bar or width b of strip). Finally, the discusser notes that perhaps (1) the shear coefficient of FRP-sand interface (G ) has been computed in Eq. (8) as G=h instead of 2G=h (as defined by the authors); and (2) an active reinforcement perimeter of πD offers a better option than 2D to represent axisymmetric effects in the W parameter included in Eq. (1).