CCES Unicamp

Topology Optimization of fluid-actuated cellular mechanisms.

Date: Jan 30, 2019, 14:00

Candidate: Daniel Candeloro Cunha

Advisor: Renato Pavanello

Faculdade de Engenharia Mecânica – UNICAMP


It was proposed the development of a topology optimization program to aid the project of bioinspired mechanisms, more specifically, cellular fluid actuators. The physical problem, its formulation for numerical analysis and ultimately the optimization procedure were described. Large displacements and strains were considered in the model, as well as hydrostatic loading, which is displacement dependent and also topology dependent. Therefore, the problems are geometrically nonlinear. The actuator is described by a macrotopology, which defines the cells’ distribution over the macrodomain; and by microtopologies, which define each cell’s shape. Two performance functions were considered for the actuators: output work, a measure of actuation amplitude; and mechanical transmittance, a measure of energetic efficiency. The optimization was performed in order to maximize one of those functions. Three modalities of optimization were defined: microscale, where a single isolated cell is considered; macroscale, where a multimaterial optimization establishes the actuator’s macrotopology, for unchanging microtopologies; and multiscale, where the cellular actuator’s microtopologies may be optimized together with its macrotopology. When working with the cellular structures, a representative model is used, where each cell corresponds to a single macroelement. A series of tests was performed to validate the formulations and implementations, and also to understand the limitations and potentialities of the proposed models and chosen methods. The validation was successful. Although the proposed representative model has been shown improper for working with mechanical transmittance, the developed program was effective for all other cases and it is suited to be used in a wide range of problems. Some categories of problems were approached: unicellular actuators, reactive structures, radial constrictors and articulated actuators. The obtained results were satisfactory for all cases and effective designs for the considered mechanisms were systematically provided by the developed program.

Related posts

Construction of reduced order models for fluid flows using deep neural networks

cces cces

Wrapping Up Viruses at Multiscale Resolution: Optimizing PACKMOL and SIRAH Execution for Simulating the Zika Virus

cces cces

CCES student wins CEPETRO Prize for Research, Development and Innovation 2018

cces cces
WP Twitter Auto Publish Powered By :