Stress–Strain Behaviour of Lattice Structures Using a Surrogate Modelling Approach

The lattice structures are efficient for designing tailored lightweight structures due to their excellent properties like high strength and energy absorption during impact, which are suitable for engineering applications. Among these different lattice structures, Voronoi based open-cell honeycomb lattices have received considerable attention in the last few decades for developing lightweight yet high-strength structures at multiple length scales. Numerous studies in the literature report that the microstructure of the honeycomb lattices can be tailored to enhance mechanical behaviour. It is observed that cell regularity is one of the critical parameters that will affect the ultimate stress and modulus of elasticity of the honeycomb lattices. The literature is limited to a few regularities due to the higher computational cost and the experimental time. Considering these drawbacks, researchers have been exploring surrogate models for solving complex problems in different fields of engineering. The present study aimed to develop a surrogate model to predict the stress–strain response honeycomb lattices using a decision tree algorithm. As the target stress–strain response consists of both linear and non-linear behaviour, the decision tree outperforms to capture these types of targets. The Sobol sampling technique is adopted to obtain the input parameters as random combinations of cell regularity (0, 0.3, 0.5, 0.7, 1), relative density (in the range of 3–9%), and elastoplastic material property (in the field of ±20%). The FE simulations are carried out using these random combinations to obtain the stress–strain response data for decision tree model. The developed surrogate model predicted the stress–strain response of honeycomb lattices with high accuracy.