Flow Topology Optimization of a Turbochargers’ Inflow Duct

Printer-friendly versionPDF version

In order to guarantee an efficient operation of a turbocharger, a ducting with low pressure losses is needed. Furthermore, the inflow duct of the compressor has to provide an air flow with a homogeneous velocity field. Otherwise, the compressor blades may temporally be faced with a critical aerodynamic load, which could result in a decreased operating range of the compressor.

With the help of 3D Computational Fluid Dynamics (CFD), engineers are able to predict accurately the flow behavior inside of flow channels and are able to judge if an adaption of the duct shape is needed in order to meet predefined flow parameters – like the uniformity index. To prevent costly manual work, automatic optimization schemes could be used. Topology optimization and shape optimization techniques represent possible schemes to obtain designs with improved flow behavior in an automated way.

This paper deals with the flow optimization of a compressors’ inflow duct, conducted as a customer project. The objective of that project was to deliver a duct design with a reduced pressure loss and an increased uniformity index at the outlet plane.  The optimization was conducted with the topology optimization software SIMULIA Tosca Fluid from Dassault Systèmes. Contrary to traditional parametric approaches where a CFD run is needed after each design modification, a topology optimization performed with Tosca Fluid acts as a co-simulation and is finished after a single coupled run.

In order to provide a high flexibility for the optimization process, the initial geometry represents the available design space delivered by the customer. To consider manufacturing constraints and further geometric aspects, the design proposal is reconstructed with CATIA after the optimization run. A flow simulation based on the reconstructed design proposal revealed increased flow uniformity and an impressive reduction of the pressure loss. By running STAR-CCM+® Adjoint Solver followed by a sensitivity-based morphing process, a further decrease of the pressure loss was achieved.

Author Company: 
Dassault Systèmes
Author Name: 
Jens Iseler