An investigation on the coupling of data-driven computing and model-driven computing

The aim of this work is to investigate the coupling of data-driven (DD) computing and model-driven (MD) computing for the analyses of engineering structures. The data-driven computing was initially introduced by Kirchdoerfer and Ortiz (2016), the main idea of which is to directly embedding the experimental material data into mechanical simulations, thus bypassing the empirical material constitutive modeling. The model-driven computing in this work refers to the standard constitutive model-based simulations, which allows to benefit from its computational efficiency. The idea of this proposed DD-MD solver is to employ the data-driven computing for the local region of the solution domain where the material constitutive models are difficult to be determined, whilst the model-driven computing is applied to the remaining regions where the material models can be easily accessed. Several numerical examples are presented to demonstrate the robustness and reliability of the proposed method.(c) 2022 Elsevier B.V. All rights reserved.