CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Problem with rotational mesh deformation (http://www.cfd-online.com/Forums/cfx/20719-problem-rotational-mesh-deformation.html)

Problem with rotational mesh deformation

Hi all,

I busy playing with the mesh deformation functionality in CFX 5.7.

I've set up a simple test problem as follows: http://www.imagedump.com/index.cgi?pick=get&tp=137134

The origen of the coordinate system is directly above the flap on the upper border of the fluid domain. The idea is for the flap to rotate around the coordinate system origen to present a varying obstruction to the flow.

I have named the cut out "flap" and specified its motion explicitly using CEL. In the Flap boundary condition mesh motion dialog I have set "Specified displacement" and used two CEL functions "XDisplacement" and "YDisplacement" for the X and Y displacements respectively.

I then defined these as follows:

AngleAmp = (30/360)*2*pi [rad] Angle = AngleAmp*sin(((2*pi)/To)*t) [rad] To = 1[s] XDisplacement = ((x*cos(Angle))-(y*sin(Angle)))-x YDisplacement = ((x*sin(Angle))+(y*cos(Angle)))-y

The mesh outline at time 0: http://www.imagedump.com/index.cgi?pick=get&tp=137135

The mesh outline at time 0.25s: http://www.imagedump.com/index.cgi?pick=get&tp=137136

As you can see the rotation is pretty much correct i.e. 30* but the flap has "shunk" down in size.

I surmise this is because I am transforming the mesh using the previous time steps x,y positions rather than the initial x0,y0 positions.

I had thought of creating 2 user scalars and writing the initial x0,y0 coordinates into them but that seems like a very inelegant way of doing things.

Any other solutions to this inappropriate scaling down problem?

Thanks ...

 Glenn Horrocks October 17, 2004 18:43

Re: Problem with rotational mesh deformation

Hi,

A comment: You don't need to calculate 30 degrees into radians. Just define AngleAmp = 30 [degree] and CFX will do the conversion for you.

There is obviously a problem with your mesh deformation equations. It often takes a bit of sorting out, but there will be an error somewhere in your CEL.

In general it is easier to specify movement to a known location with displacement from original location, and easier to specify incremental movement (that is where the fluid forces contribute to the mesh motion) from the previous timestep positions.

Glenn

 All times are GMT -4. The time now is 00:56.