Sensitivity clustering-based multi-scale topology optimization method for metamaterial thermal actuators
Zheng Qiu,Quhao Li,Shutian Liu
DOI: https://doi.org/10.1007/s00158-023-03729-z
IF: 4.279
2024-01-21
Structural and Multidisciplinary Optimization
Abstract:Thermal elastic metamaterials with extreme positive or negative thermal expansion properties have great potential applications in designing and manufacturing thermal actuators. In this paper, a concurrent topology optimization method is proposed for metamaterial thermal actuator design. The actuator is composed of multiple metamaterials with different microstructures. The microstructures are composed of two common materials, aluminum and invar. The concurrent optimization method is capable of simultaneously designing the topology, the arrangement of the multiple metamaterials, and the microstructures of the metamaterials. The macro-topology and the multiple metamaterial microstructures are optimized simultaneously through density-based topology optimization method and multi-phase material optimization method. A sensitivity clustering-based method is proposed and used to partition the macrostructure into several sub-domains, each of which is occupied with appropriate individual metamaterial. The sensitivity of the optimization objective with respect to the thermal expansion coefficients is chosen as the criterion to generate data points for clustering, and the fuzzy c-means (FCM) method is adopted for data clustering. The grouped microstructures will greatly reduce the computational burden and at the same time improve the manufacturability and structural performance. Several numerical examples are presented, including the design with single output point and multiple output points. Optimization results show that both the macro-topology and the metamaterial microstructures are significantly affected by the actuation impedance adopted in the optimization. The effect of number of clusters on the optimization results is studied.
mechanics,computer science, interdisciplinary applications,engineering, multidisciplinary