Artificial neural network-based optimization studies for efficient energy harvesting from auxetic laminated composite smart beam

Abstract

In the present study, a smart laminated composite beam with auxetic configurations is considered to improve its energy harvesting capabilities. For this purpose, three different cases of auxetic geometries are considered on the cantilever laminated composite beam near the fixed end, and the piezoelectric patches are attached in bimorph configuration over this auxetic region. The coupled structural and piezoelectric equations are developed based on beam element theory. The principle of minimum potential energy is implemented to solve the equations of motion. Effective Young’s modulus and Poisson’s ratio of the auxetic geometric region are utilized in the beam model, and the results are validated with three-dimensional finite element solution. An experimental study is carried out to validate the finite element studies. The energy harvesting capability of laminated composite beam has significantly improved by using reentrant auxetic geometry. In addition to auxetic geometry, auxetic cubical inclusions are also considered to study their effect on energy harvesting potential. On validating the results obtained from auxetic geometry and auxetic cubical inclusion cases, some parametric studies are conducted to know the influence of auxetic inclusion dimensions, ply orientations, and thickness ratio on the voltage energy output. It is found that with an increase in the length and thickness of the auxetic geometry, the cross-ply oriented auxetic energy harvester’s output potential raised considerably. To obtain the optimum parameters required to enhance average output power and minimize the energy loss factor, grey-relational analysis tool, and artificial neural network-based optimization schemes are implemented. The energy harvesting system with the optimized parameters has shown about 60% improvement in the average power output and conversion efficiency.