Sensitivity of the thermomechanical response of elastic structures to microstructural changes
This paper is focused on the analysis of the sensitivity of the thermomechanical response of a macroscopic elastic body to changes that occur at the microstructure. This problem is a key issue in material design. The sensitivity analysis relies on an accurate determination of the effective propertie...
| Autores: | , , , |
|---|---|
| Tipo de recurso: | artículo |
| Fecha de publicación: | 2015 |
| País: | España |
| Institución: | Universitat Politècnica de Catalunya (UPC) |
| Repositorio: | UPCommons. Portal del coneixement obert de la UPC |
| Idioma: | inglés |
| OAI Identifier: | oai:upcommons.upc.edu:2117/76868 |
| Acceso en línea: | https://hdl.handle.net/2117/76868 https://dx.doi.org/10.1016/j.ijsolstr.2015.06.009 |
| Access Level: | acceso abierto |
| Palabra clave: | Thermodynamics Continuum mechanics Microstructural material design Structural optimization Sensitivity to microstructural changes Computational homogenization of materials Response surface methodology COMP-DES-MAT Project COMPDESMAT Project Termodinàmica -- Assaigs de materials Àrees temàtiques de la UPC::Enginyeria civil::Materials i estructures Àrees temàtiques de la UPC::Física::Termodinàmica |
| Sumario: | This paper is focused on the analysis of the sensitivity of the thermomechanical response of a macroscopic elastic body to changes that occur at the microstructure. This problem is a key issue in material design. The sensitivity analysis relies on an accurate determination of the effective properties of the heterogeneous material. These effective properties are determined by computational homogenization. And their sensitivities, with respect to the parameters defining the microstructure, are then computed. For an efficient evaluation of the thermomechanical response, we propose to build response surfaces for the effective material properties. The surfaces are generated in an offline stage, by solving a series of homogenization problems at the microscale. In such a way, the fully online multiscale response analysis reduces to a standard problem at the macroscale. Thus, an important reduction in computational time is achieved, which is a crucial advantage for material design. The capability of the proposed methodology is shown in light of its application to the design of a thermally-loaded structure with variable microstructure. Considerable improvements in the structural response are achieved. |
|---|