Towards the design of campaniform sensilla inspired strain sensors: an FEM-based structural analysis and geometrical optimization

摘要

In this thesis, a structural analysis is performed to gain insights on the strain amplification characteristics of campaniform sensilla. In order to simplify the analysis, a single hole and four holes in arrays of two in orthotropic lamina are considered. Firstly, a finite element method (FEM) analysis is performed to discretely assess the influence that different geometrical parameters and material properties have on the mechanical amplification of the hole(s). Secondly, an artificial neural network is used to obtain an approximated multi-dimensional continuous function, which models the relationship between the geometrical parameters and the amplification of the hole(s). Thirdly, an optimization is performed to identify the geometrical parameters yielding the maximum mechanical amplification. Finally, results are validated with an additional FEM simulation performed by varying geometrical parameters in the neighborhood of the identified optimal parameters.