Separated Regions of Centrifugal Impeller
 

Hello people!

i`d like to compare different impeller-geometries by the size of their separated regions. I`ve already compared the massflow rate that goes backwards through the inlet and outlet of the impeller and also through the outlet of a vaneless diffuser downstream. I implemented this by creating velocity contours with only 3 levels and defining the range with the same size of velocity in +/- direction...so that the recirculating regions become visible. Now i would like to have another criteria for comparing separation. I tried to make a contour of the wall shear stress on the impeller blades. Wall shear stress should become zero at the point of separation and finally negative when there is backflow but as CFX calculates the stress by vectors there is no +/- change at all. So all I can do is plotting the streamwise velocity over the walls (with conservative data) and compare the regions of negative velocity. Does anybody has an idee how to involve the wall shear stress or how to make it change its algebraic sign? I use the SST-Model...

I would be very happy about a hint...