[Julian121]

I would like to find the location where flow would seperate from a rotor blade. I found that there are many velocity components in CFD-Post. Which of them should be used?

[ghorrocks]

In 2D flow separation has a clear definition - it is where the wall shear stress goes to zero.

But in 3D flow separations do not have a clear definition. As there is no universal definition you then have to think about what definition makes sense in your application. You might be able to use the component of wall shear stress in the axial direction going to zero, or you might be able to draw an isosurface of zero axial velocity and where that intersects the surface - or you might have to work out a definition more suited to your application.

[Julian121]

There is a flow seperation from the blade tip suction surface at 99% span, but the shear stress predicts positive wall shear stress. Should it not go to zero?

Originally Posted by ghorrocks

In 2D flow separation has a clear definition - it is where the wall shear stress goes to zero.

But in 3D flow separations do not have a clear definition. As there is no universal definition you then have to think about what definition makes sense in your application. You might be able to use the component of wall shear stress in the axial direction going to zero, or you might be able to draw an isosurface of zero axial velocity and where that intersects the surface - or you might have to work out a definition more suited to your application.

[ghorrocks]

That is a demonstration of exactly what I am talking about. You are thinking the flow is 2D, but in 3D separations like this are often fed by a small flow perpendicular to the flow direction along the wing surface.

So you have found an example of why, when you look at the details, there frequently is not actually a traditional separation in 3D flows at all.

[Julian121]

Does it mean that since the flow is not 2D around the blade in reality, the 2D flow vectors mislead us to believe that the flow is seperated from the blade although it is not?

If it was, the shear stress would go to zero?

Originally Posted by ghorrocks

That is a demonstration of exactly what I am talking about. You are thinking the flow is 2D, but in 3D separations like this are often fed by a small flow perpendicular to the flow direction along the wing surface.

So you have found an example of why, when you look at the details, there frequently is not actually a traditional separation in 3D flows at all.

[ghorrocks]

I would recommend you look at this area of your model in detail. Look at what the velocity vectors are doing at what appears to be the separation point. Zoom right in so you can see the details. Use vectors and/or streamlines so you can see the flow paths.

Hopefully you can answer your own question when you see the details of what is going on.