Topology optimization of self-sensing nanocomposite structures with designed boundary conditions

Controlling volume fractions of nanoparticles in a matrix can have a substantial influence on composite performance. This paper presents a topology optimization algorithm that designs nanocomposite structures for objectives pertaining to stiffness and strain sensing. Local effective properties are obtained by controlling local volume fractions of carbon nanotubes (CNTs) in an epoxy matrix, which are assumed to be well dispersed and randomly oriented. The method is applied to the optimization of a plate with a hole structure. Several different allowable CNT volume fraction constraints are examined, and the results show a tradeoff in preferred CNT distributions for the two objectives. It is hypothesized that the electrode location plays an important role in the strain sensing performance, and a surrogate model is developed to incorporate the electrode boundary as a set of additional design variables. It is shown that optimizing the topology and boundary electrode location together leads to further improvements in resistance change.