For a so-called backward facing step problem the local velocities and fluctuating components are measured in detail by several authors. When solving this problem numerically and comparing it with measured data still a lot of models and parameters can be varied. These models and parameter settings are the choice of the user and all of them influence the calculation results. The used complexity level for the selected models is a balance between what is necessary and what can be done. Comparison of measurements and calculations with each other are based on comparison of axial velocities and axial turbulent intensities at one and eight times the stepheight behind the expansion. The calculated centerline velocity at the expansion and the reattachment length are compared with literature data. For a couple of models and parameters the influence on the results can be explained on physical or numerical grounds. For other parameter variations it was expected that they would hardly influence the accuracy of the solution. Surprisingly this is not always the case. Computer Fluid Dynamic simulations and especially the interpretation of results requires a good deal of experience and knowledge.
|Number of pages||11|
|Journal||Phoenics Journal of Computational Fluid Dynamics and Its Applications|
|Publication status||Published - 1993|