CFD-procedures for Air Induction Systems

Maria Dybeck (Luleå University of Technology)

CHALLENGE - The first is dedicated to predicting the flow through the air mass flow sensor present in the system. Experiments show that at low mass flows the air mass meter signal is affected by a noise. Current simulation procedures cannot predict this noise and a new procedure were needed. The procedure developed could not find the signal noise but can be used as a basis for a future transient analysis procedure at Volvo. The second is an optimization procedure designed to find the optimal design for any given geometries in the system with respect to pressure losses and flow uniformity index. The optimization procedure showed promise in finding optimal designs and will be of used in future development.The procedures are tested and developed on existing air filter systems. Experiments were performed with the system in a flow rig and verified the simulated results.







SOLUTION - The goal of the optimization was to be fast, reliable and easy to implement. Thanks to modeFRONTIER it was possible to achieve an optimal solution to the optimization problems of this thesis.  The workflow used was set up with ANSA and STARCCM and in the end modeFRONTIER registered the results and produced new designs according to the algorithm set for evolution.

BENEFITS - modeFRONTIER enabled the integration of different simulation models into a single synergetic design tool. In addition, the designers took advantage of various optimization procedures available in modeFRONTIER, which are an essential part of the MDO approach.