The numerical simulation of standard concrete tests and steel reinforcement using force flux peridynamics

Artikel i vetenskaplig tidskrift, 2023

Peridynamics is a numerical particle-based solid mechanics method that enables the simulation of brittle and quasi-brittle materials, as well as ductile materials. It allows cracking to appear spontaneously in the arms joining the particles and can therefore be used to simulate progressive fracture. In this article, we apply our version of peridynamics, which we call force flux peridynamics, to the simulation of concrete where the appearance of cracks plays an important role in the global mechanical properties. It is not difficult to modify the material parameters in peridynamics to achieve a given tensile strength or a given compressive strength. However, it is much more difficult to choose parameters which will model all the strength parameters of a material within the same model. When concrete fails in compression it may split or spall showing a complex relationship between compressive and tensile failure. We therefore set ourselves the simple task of producing a single peridynamics model which can predict the stiffness and strength behavior of concrete in standard compression and tension tests for which we chose the American Society for Testing and Materials standards for the cylinder compression test, the split cylinder test, and the modulus of rupture test. A parameter sensitivity study was performed based on the cylinder compression test to tune the key peridynamics parameters that determine the global material behavior. The compressive and tensile strengths were then determined from the combined simulation data. While the fracture modes, crack branching pattern and also the stress–strain curve show promising results, the maximum tensile strength was found to be significantly larger than physical experiments suggest. This is probably due to imperfections within real concrete at the interface between aggregate particles and cement paste and it shows that the detailed numerical modeling of the failure of concrete is highly complex with a large number of unknown material parameters.