Community
CFD Forum
Welcome to Autodesk’s CFD Forums. Share your knowledge, ask questions, and explore popular CFD topics.
cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Wall Calculator inconstancy. Maybe User error? Please help.

2 REPLIES 2
Reply
Message 1 of 3
kprokopa
388 Views, 2 Replies

Wall Calculator inconstancy. Maybe User error? Please help.

Wall Calculator inconstancy.  Maybe User error?  Please help.

 

 

I seam to be getting sum FX results in 2 different modes.   The report appears to randomly switch.   One mode is typically 2 to 3x higher forces then the other, however I do not always know which one I am looking at.  By randomly I mean that some tests always report the higher mode, some test runs report the lower mode, and a few test switch from the higher mode to the lower mode when the solver is restarted for additional iterations.  (I have never seam it go from the low force mode to the high mode).    FY is more consistent then FX. (air flow is along the X axis)

 

 

Specifically I just ran this test:

 

1)  I set up my test to run adaptive meshing x3, 250cycles per.  I left the wall calculator window open and watched it converge to 6.5Newtons FX.  This ran to completion and stopped.  It maybe as a little early for convergence, but close enough to be in the ball park.

 

(Last time I ran this model I got an FX of 2.2Newtons... )

 

2) So Scratching my head a bit I clicked 'solve' to let it run for another 250 iterations (on the last mesh. no more adaptation).   I am not aware of any setting changes or doing anything other then clicking 'solve' again.

 

on Iteration 251, it jumped to approximately 2.2Newtons.   cycle 500 was still approximately 2.2Newtons. 

 

 

 

 

Did I hit some mode switch?  why is there 2 'sets' of wall force numbers.

 

This is very baffling and causing me to waste a lot of time second guessing which of my tests are meaningful numbers.  Sometimes I get one set of numbers, sometimes the other.

 

 

----

Some background on what I am trying to do if it helps: I am trying to calculate drag on a set of simple land vehicle shapes.  30m/s. 12.35psi 26.6degC.  My shape is symmetrical so I am only simulating half of it.   The model is a block of air with the half vehicle shape removed.   I am only modeling the Air, no solids.     I have 4 slip surfaces, a zero pressure on the right face and a 30m/s inlet on the  left face.

 

The test is a single rounded surface that I am measuring the force on with the wall calculator.

 

2 REPLIES 2
Message 2 of 3
kprokopa
in reply to: kprokopa

 

I found more inconsitent behavior with the wall calculator.   Yesterday 2 different designs blew up in the wall calculator.

 

1-  it was fine through the first 2 adaptive meshing iterations,  I left the wall calculator open as I often do.  After it finished FX was reported as 4newtons or so.   I thought it might need a few more iterations to converge so I hit 'solve' and on the next iteration it dropped to 2newtons.    Except this time I found I could get the more reasonable 4newton number back if I closed the wall calculator, and then reselected the surfaces I was interested in.    The total surface area was always correct at 2800in2, however, it seams to have gotten inconcsitent notion of which sufaces it was summing up the forces for.

 

2-  a different design was 'ok' through most of the run... but at some point in the 3rd adaptive meshing it went completely bonkers in the wall calculator.   it was watching when this happened.  it went from ~5newtons to 56000newtons.  surface area calculatos went nuts also.     restarting the wall calculotor did not help.      the analysis was nearly done when this happened.  the other CFD features looked ok, just the wall calculator went crazy.

 

 

so the question is...   how do YOU get consistent meaningfull results from the wall calculator. ?

 

thank you for your  help,

 

 

 

Message 3 of 3
marco.mueller
in reply to: kprokopa

The forces update at the end of the run since only at this point the friction effects are calculated (during the run only pressure effects). This may be the reason.

Dipl.-Ing. (FH) Marco Müller
Application Engineer Digital Simulation
Mensch und Maschine Deutschland GmbH
www.mum.de/cfd

Can't find what you're looking for? Ask the community or share your knowledge.

Post to forums  

Autodesk Design & Make Report