[jaxk]

Hello,

In CFX Workbench I use Parameter Set to vary the angle of attack of a Blade. I want to know the drag coefficient for each angle, but because of vortex shedding it is not a constant value. I therefor want the (time) average value.

I have created an expression to calculate the drag coefficient:
CD = force_x()@Blade/(0.5*RhoWater*BladeArea*Uinf^2)

I then made a Monitor Point to monitor the CD value during the run.
With Monitor Statistics I am taking the Arithmetic Average over a Moving Interval of 100 Iterations.

In the Solver Manager, I can see both the Monitor Point and the Arithmetic Average.

However, in PST I am unable to add the Arithmetic Average as a Workbench Output Parameter.

My only option now is to manually open the Solver Manager for each finished run, and then manually extract the Arithmetic Average for each angle from the graph, but this is very time consuming.

Does anyone have some advice to directly add the average CD as Output Parameter?

Thanks!

[ghorrocks]

If you want to automate extraction of data from the Solver Manager you can use the command line "cfx5mondata -res [res file] -out [output file]" and it will generate a csv formatted data file of all the solver manager plots.

But I cannot answer your question about making it an output parameter until I get to work tomorrow.

[jaxk]

Originally Posted by ghorrocks

If you want to automate extraction of data from the Solver Manager you can use the command line "cfx5mondata -res [res file] -out [output file]" and it will generate a csv formatted data file of all the solver manager plots.

But I cannot answer your question about making it an output parameter until I get to work tomorrow.

Hey Ghorrocks,

Can you come back to your previous post? I'm really interested in making a Monitor Point an Output Parameter.

Thanks!

[ghorrocks]

The only way I can see of doing this is to define a CEL expression which uses the .Trnavg variable operator. As far as I know the variable operator only works on simulation variables (u,v,w,p,T etc) and cannot be applied to CEL expressions (like your CD equation) or CEL functions (like your force_x() function). The .Trnavg variable operator also does not allow much control - it only applies to all points in the simulation so you cannot do a running mean.

So you will be very limited in what you can do if you want to output transient statistics.

Personally, I would not do this sort of stuff in CFX because of these limitations. I would export the raw data from solver manager and do all the processing in some other package where you have full control over what it is doing (python, excel, matlab etc).

[jaxk]

Hmm, thanks, I'll keep trying then, maybe I can think of some workaround.