A methodology for parameter identification and calibration of the cohesive element based meso-scale concrete model

Abstract

Meso-scale analysis techniques have demonstrated superiority in analyzing structural cracking and crack propagation compared to conventional methods. The cohesive element based meso-scale analysis technique has gained widespread acceptance due to its clear analytical concepts and the ability to simulate complex fracture modes. However, the cohesive zone model (CZM) requires numerous parameters for different fractures patterns, and the impact of each cohesive element parameter on the mechanical properties of the meso-scale model remains uncertain. This limitation restricts the applicability of CZM models in concrete structure analysis. Even experienced researchers often rely on trial-and-error methodologies or empirical methods to determine cohesive element parameters. The present study aims to investigate the effects of cohesive element parameters in the CZM concrete model under static axial loading and propose a feasible method for determining these parameters through a parametric analysis. Initially, a Python script was developed to generate random aggregate models, modify polygonal aggregates, and insert cohesive elements. Subsequently, numerical simulations were conducted on two-dimensional standard concrete specimens under axial compression and tension. Based on the results of the parametric analysis, a novel prediction formula and an inverse method for cohesive element parameters were introduced. Additionally, the accuracy of the novel method was validated in accordance with the Chinese Code. The results indicate that the proposed method exhibits exceptional capabilities in predicting initial stiffness, uniaxial strength, and corresponding strain. However, the prediction of the descending segment of the stress–strain curve lacks precision. In summary, the methods presented in this study demonstrate high accuracy, which can provide a reference basis for researchers and promote the application of CZM meso analysis method.