Theoretical modification of the laboratory-determined tensile stress–strain relationship of strain hardening cementitious composites (SHCCs) for large-scale specimens

Abstract

Strain hardening cementitious composites (SHCCs) are a type of cementitious material reinforced with randomly distributed short fibers that exhibit tensile strain hardening behavior and prominent crack resistance. The tensile behavior of SHCCs could be significantly affected by the distribution of the fiber orientation. In large-scale specimens, the distribution of the fiber orientation is close to ideal three-dimensional uniform random distribution. However, fiber orientation distributions in laboratory SHCC specimens are influenced by the specimen dimensions (width and thickness) and casting process, resulting in neither ideal two-dimensional nor three-dimensional uniform random distribution. Therefore, the performance of large-scale SHCC specimens may be overestimated in analysis when directly using the tensile properties of SHCCs obtained from uniaxial tensile tests of small specimens. Since the uniaxial tensile tests of large-scale SHCC specimens are difficult to be conducted, theoretical evaluation of the tensile properties of mass SHCCs in large-scale specimens were conducted by modifying the uniaxial tensile test results of small specimens. In this paper, closed-form probability distribution models for fiber orientations that consider the influence of the specimen width and thickness were theoretically derived. The effects of the specimen width and thickness on the tensile properties of SHCCs were studied by applying the proposed method to the uniaxial tensile numerical model of SHCCs, while uniaxial tensile specimens with two different geometries and dimensions were tested for verification. Both the analytical and experimental results showed that, within a certain range, the width and thickness of the specimens significantly influenced their ultimate tensile strength and ultimate tensile strain. To relate the uniaxial tensile test results of laboratory-sized specimens to that of mass SHCC in large scale specimens, reduction factors for the stress–strain relationship of SHCCs obtained by traditional tensile test specimens were proposed to account for the specimen size effect based on the model analysis. The reduced tensile stress–strain relationship was applied to a finite element model of SHCC beams under four-point bending tests. It was found that the finite element analysis results obtained by using the reduced tension stress–strain relationship are in better agreement with the test results.