Is there any way of drawing streamlines on a body itself, analogous to oil flow patterns on a wind tunnel model????

Common sense would suggest to plot the surface shear stress on the body, but neither the streamline nor vector plot options allow that as plot variable, only velocity, u v and w velocity gradients, and gradients of density, pressure and total enthalpy. The options for velocity gradients are not useful because the specify only the component gradient, not the gradient of the total vector...

Why, if I can plot the magnitude of wall shear on a surface, can I not also get the direction?

Hi Louwrens,

You will be able to plot Wall Shear vectors in version 5.6.

Robin

Hi Robin

Is there no way that I can compute it as a new variable from my existing res files, from the wall shear and the velocity vector? Problem of course is that the velocity is zero at the wall...

Louwrens

Hi Louwrens,

If you use the "conservative" velocity, it will be non zero and point in the right direction. However, even if you calculate the components, you cannot create a new vector variable in Post.

Regards, Robin

Hey this works... plotting streamlines or vectors of the conservative velocity on a surface gives me what I'm looking for. Thanx!

Which brings me to another point: Whats the difference between the hybrid and conservative values? I know I can specify them in the function calculator as well, and they give slightly different answers.. When would you use which one (apart from making pretty pictures?)

Hi Louwrens,

Since CFX-5 builds control volumes around the nodes (see solver theory for details), you end up with a half control volume next to the boundary whose node is on the boundary. This means that the boundary node has two values, the "conservative" control volume value and the "hybrid" boundary value.

For example: at a wall, the hybrid value gives you the value at the wall. If you are looking at temperature, the hybrid value is the surface temperature, whereas the control volume value is the fluid temperature.

Regards, Robin

Yes, this would work, i.e. you can make the plot, but it may not be right. The wall shear acts tangential to the boundary, but unfortunately, conservative velocity is not necessarily always tangential. If it is in your case, then it might still be OK.

Unfortunately there is no way to get the tangential direction in CFX-Post and you will have to wait for the wall shear vector in 5.6.

Neale