A time-dependent approach for calculating steady inverse boundary-layer flows with separation

Abstract

In this paper we describe an unsteady inverse boundary-layer method that can be used to compute steady flows with separation. The method uses Keller’s box method with Cebeci’s Mechul function formulation. De­pending on the complexity of the flow, two versions of the box method are used. In regions of positive streamwise velocity component u , the regular box is used; in regions where u becomes negative (t > 0), the zigzag box is used. When t = 0, and u becomes negative in some region across the layer, the regular box with the FLARE approximation is used. Calculations per­formed with this approach indicate that the use of a time dependent inverse boundary-layer method in which time is used as an iteration parameter provides a good approach in improving the accuracy of the solutions obtained from the FLARE approximation.