Presentation slides

Abstract

Over the past 12 years Tenneco has been able to significantly accelerate the development process of lightweight pistons for gasoline and diesel engines through the use of sensitivity-based shape optimization and is now able to calculate reliable piston designs for highest engine requirements within a very short time [1-4]. The in-house software FMShape was developed for this purpose, which uses Abaqus as solver to calculate both stresses, eigenfrequencies and their design sensitivities.

In contrast to the classic design methodology with manual design and analysis iterations, shape optimization iteratively changes the surface of the FE model of a starting design. The manufacturability of the component should be guaranteed as far as possible in order to facilitate the transfer of the optimized geometry back to a CAD model which is permissible with regard to production. In optimization with FMShape, this is done through the use of geometric constraints, so-called manufacturing constraints, which are maintained in each iteration.

The aim of shape optimization is often to improve the strength of a component by reducing stress concentrations. Compared to topology optimization, the geometry changes in shape optimization are typically local and small in relation to the overall dimensions of a component. The relative stress reduction achieved with FMShape is in the range of 10 to 40%. Improvements of this magnitude are significant, as a stress reduction of about 15% leads to an extension of the fatigue life of the aluminum materials used in piston development by a factor of 10.

Recently, FMShape has also been extended to optimize the bead layout of sheet metal structures to improve the NVH behavior of other Tenneco components like exhaust systems or heat shields. A new feature of bead optimization with FMShape is a bead ratio constraint that can be used to control the degree of bead usage of a shell structure. The application of this constraint favors the development of clear bead structures, thus improving the interpretation and transfer of the optimization results into the CAD geometry and thus increasing the effectiveness and acceptance of the methodology within the development process.

As part of the presentation, the strategy and application of shape and bead optimization with FMShape will be presented for several academic and industrial examples.

[1] R. Meske, M. Scherer. Sensitivitätsbasierte parameterfreie Formoptimierung. Deutsche SIMULIA-Konferenz, Dresden, 25.-26. September 2014.
[2] R. Meske, M. Scherer. Berechnest Du noch oder optimierst Du schon? – Revolution der Produktauslegung durch sensitivitätsbasierte Formoptimierung. NAFEMS Seminar Simulation Driven Engineering, Neuendettelsau, 20. - 21. November 2017.
[3] R. Meske, M. Scherer, G. Stankiewicz. New approaches in bead optimization using semi-analytical sensitivities. 3DEXPERIENCE Conference 2019, Darmstadt, 19.-21. November 2019.
[4] M. Scherer, R. Meske, G. Stankiewicz. Sickenoptimierung auf Basis semi-analytischer Sensitivitäten. NAFEMS DACH Konferenz 2020, online, 13.-14. Oktober 2020.