EnginSoft - CAE Conference Abstracts

EnginSoft International Conference 2011
CAE Technologies for Industry

Multi-Objective Optimization of Electromagnetical Devices Integrating modeFRONTIER with ANSYS Maxwell

Nicolich Matteo - ESTECO (Italy)
Clarich Alberto - ESTECO (Italy)

Abstract

Multi-Objective Optimization tools, and in particular modeFRONTIER® product from ESTECO, are nowadays successfully adopted as standard design methodologies in many application fields (aerospace, automotive, appliances, biomedicals, etc.) but they have not yet been intensively employed in electromechanical design. The big advantage offered by this technology is the possibility to automatically optimize, through the execution of series of different CAE simulations of parametric models, several and contrasting objectives at the same time, achieving important targets such as efficiency maximization, consumption minimization, performances maximization and weights and dimension minimization, fulfilling all the user required constraints and targets. modeFRONTIER® not only offers a large variety of Optimization Algorithms, such as multi-objective Genetic Algorithms, Evolutionary Strategy, Game Theory, Particle Swarm, Gradient Based and SQP, etc., but allows the user to build a complete workflow in which all the modular components of the optimization chain can be defined for a full automatic execution (I/O variables, I/O files, models and scripts to be executed, etc..), and moreover offers a complete post-processing environment which includes Statistical Analysis tools, Six Sigma and Quality Analysis, Response Surfaces and meta-models, and much more. Several direct interfaces for many CAE software, including the complete ANSYS® suite, can be used in the modeFRONTIER® workflow, allowing to automatically choose the I/O parameters defined in the CAE model for the optimization. In addition, a Grid functionality (or queuing system) is available in modeFRONTIER® to launch the different design simulations in parallel on different machines available in the local network, fully exploiting the available CPU resources. The application illustrated in this paper simulates and optimizes an axisymmetric solenoid; the armature force will be maximized, while minimizing the amount of steel for the core and housing, in order to create the most cost effective design.
A solenoid is a coil that generates a magnetic field. Due to this magnetic field, a force acts on a ferromagnetic projectile, called the plunger, in this magnetic field. This force moves the plunger toward the centre of the coil. ANSYS Maxwell® application has been used to simulate the electromagnetical field; modeFRONTIER® is able to automatically drive the geometrical parameters using the direct interface with ANSY Workbench®. The brand new optimization algorithms of modeFRONTIER®, HYBRID (which combines the robustness of a GA and the accuracy of SQP) and FAST (which combines any efficient Multi-Objective Optimization algorithm with automatically updated Response Surfaces able to extrapolate the response of the system reducing drastically the needed number of CAE simulations) have been applied and compared in terms of efficiency and results.