CFD related optimization is a very demanding task in terms of computational resources. The key advantage of the adjoint solution based method is that it requires fewer design iterations regardless of the number of design parameters, and therefore drastically cuts the number of Design of Experiment (DOE) simulations in the optimization process.

A joint case study was made by Volkswagen and ESI Group on a HVAC (Heating Ventilation Air Conditioning) air duct using both the adjoint solver and a conventional optimization method. The adjoint solver demonstrated improvements in both HVAC performance and turn-a-round time. It took two design iterations to achieve 20% reduction of the head loss, with automatic shape update in six hours. A single alternative conventional optimization solution required 600 design iterations (600 CFD solver runs) in order to achieve the same level of head loss reduction.

“The adjoint solver shows great potential to cut the number of prototypes in the entire design process. It is a major breakthrough for car designers who use CFD solutions. At Volkswagen, in collaboration with ESI Group’s engineers, we are looking very deeply into this method to extend the applications from the air duct, to aerodynamics drag/lift or yaw moment, aero-acoustics and engine intake port design etc.”, says Dr. C. Othmer, CAE Method Division, Volkswagen, Wolfsburg, Germany, who has presented the EUROGEN 2005 paper for aerodynamic shape optimization with ESI Group’s PAM-FLOW™ adjoint solver.

“ESI Group considers this adjoint solver as one of its strategic technologies for Simulation Based Design, to supply end users a transparent process within their simulation environment.” adds Dr. Ming Zhu, CFD Product Manager, ESI Group. “Today we are working with Volkswagen and other industrial partners across the world through joint studies to adapt and enhance our technologies for all our CFD solutions.”

ESI Group plans to have the PAM-FLOW™/AS product released by spring 2006.