Simulated annealing algorithms for the optimization of particulate composite structures analyzed by X-FEM
This research objective was to implement a simulated annealing, SA, optimization algorithm to an extended finite element, X-FEM program called XMicro. The purpose was to combine the advantages of the optimization method with the auto re-meshing advantage of the X-FEM and thus minimize the burdens of combinatorial selection of particulate composite material reinforcement architecture. The optimized parameter was the strain energy and the varying parameters were the aspect ratio which varied from 0.1 to 10, the area of the inclusions which varied from 1.8 to 90 μm2 and reinforcement volume fraction which varied from 10 to 50%. For a SiO2/ZrO 2 particulate composite material subjected to a temperature load of 500 K, the simulated annealing arrived to a sub-optimal strain energy state of 1.37e-2 GPa/m3. The corresponding zirconia volume fraction was 11% formed of 10 sided polygon inclusions with an aspect ratio of 8.56 and 6.19 μm2.
Renova, Elvia Paola, "Simulated annealing algorithms for the optimization of particulate composite structures analyzed by X-FEM" (2008). ETD Collection for University of Texas, El Paso. AAI1456741.