Tuning structural dynamics through architected inner material designs: Numerical, experimental, and machine learning analysis
I. O. Hassan, Agyapal Singh, Dimitrios C. Rodopoulos, Alexander Morawietz, Andrea Bergamini, Nikolaos Karathanasopoulos
Abstract
Architected materials and structures have garnered significant interest due to their potential to furnish mechanical performances beyond the bounds of customary designs. The present work investigates the flexural and modal response of Triply Periodic Minimal Surface (TPMS)-based architected beam structures, engineered with different metamaterials, relative densities, and structural configurations. The work combines experimental and numerical analysis, performing 3-point bending tests on additively manufactured architected beams. Designs with exceptional flexural rigidity approaching 3 N . m 2 are identified, while insights into their inner stress profiles are provided. Moreover, their dynamics are assessed, revealing a topology-dependent modal response. The architected beam mechanics are compared with analytical and Finite Element Analysis (FEA) results for solid, non-architected elements, identifying alterations in the modal response, infeasible for solid beams. Evidence is provided that the appropriate microstructure selection allows for a controlled appearance of torsional and bending modes below 2000 Hz. Furthermore, machine learning models are developed to predict and explain the recorded performance, classifying the importance of underlying influential parameters. It is shown that the effective bending stiffness can be used as a lever to control the effective dynamic response. The analysis provides benchmark results for the engineering of advanced, lightweight structural members, with extraordinary dynamic attributes.