Research Projects

Summary:

In the last few years the optimum structural design problem have gained an increasing importance due to the growing trend on the use of numerical applications in order to automate the design process. In particular, topology optimization results a powerful tool to support the choice of the designer in new and efficient conceptual designs. These methods deal with the design problem from a deterministic approach, without to take into account several sources of uncertainties presented in real design (materials, loads, boundary conditions, etc.). Consequently, deterministic optimal design can present different real behaviours than the expected ones by the numerical results.

The global aim of the project consists of the formulation, development, implementation and validation of a methodology for the robust topology optimization of structures, for continuum and discrete ones. The objective of this methodology is to provide efficiently optimal topologies of structures that are non-sensitive to uncertainty, which allows its application to real problems. The project addresses the following research lines in order to reach the objectives:

1. The problem formulation of robust topology optimization of continuum structures.
2. The problem formulation of robust topology optimization of truss structures.
3. The research and evaluation of the more suitable techniques for the efficient uncertainty management of geometry, material properties, boundary conditions and loads.
4. The formulation and implementation of the strategies (surrogate models) needed to solve the high computational cost suffered from robust topology optimization problem.
5. The implementation and validation of the more suitable optimization algorithms to solve the robust topology optimization problem.

This project has a double interest: scientific and technical. Under the scientific point of view, the interest is the generation of new knowledge about robust topology optimization of structures using stochastic parameters and/or stochastic variables. Therefore, the formulation of robust topology optimization problem, requirements from the stochastic features of design variables and high computational cost of topology optimization will require to evaluate, and probably to take, novel uncertainty management techniques of surrogate models and optimization. Under the technical point of view, two software applications will be developed at the end of the project, namely: SIDORTE-C (Integrated System for the Optimal Robust Design of Topology of Continuum Structures) and SIDORTE-A (Integrated System for the Optimal Robust Design of Topology of Truss Structures), which will be very useful to the conceptual design of structures.
The research group of Structural Optimization has a wide experience, scientific and technical both, in the research lines of the project (optimization techniques of topology continuum and truss structures, stochastic optimization and surrogate models). Such experience is backed by publications in journals and proceedings about structural optimization, and by projects, supported by public and private funding, that have been previously achieved.