Simple testCase Blasius Boundary Layer, just NaNs for rhou,rhov,rhow......

[Gabriel Axtmann]

Hi,

since I am new to pyFr I have a little issue. I set up a simple test case. 2D Blasius Boundary layer at Ma=0.4

The mesh was created by Pointwise and exported in pyfrm format.

Also the ini file are adapted by using pyfr tutorials

The problem is I got always NaNs in the residual.csv, so where is my error?

Thanks in advance

[Vincent, Peter E]

Hi Gabriel,

What CFL number are you running at?

Peter

Dr Peter Vincent MSci ARCS DIC PhD
Reader in Aeronautics and EPSRC Fellow
Department of Aeronautics
Imperial College London
South Kensington
London
SW7 2AZ
UK

web: www.imperial.ac.uk/aeronautics/research/vincentlab
twitter: @Vincent_Lab

--
You received this message because you are subscribed to the Google Groups "PyFR Mailing List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pyfrmailingli...@googlegroups.com.
To post to this group, send email to pyfrmai...@googlegroups.com.
Visit this group at https://groups.google.com/group/pyfrmailinglist.
For more options, visit https://groups.google.com/d/optout.
<thumm_pw.ini><thumm_pw.pyfrm><residual.csv>

[Gabriel Axtmann]

Hi Peter,

I use a time step of 5e-05, which equals in CFL<0.3. Sry this is not updated in the files I uploaded

[Gabriel Axtmann]

Hi Peter,

you were right, time step was calculated by using delta x, should have used delta y since this is my smallest cell direciton. With a time step of 1e-08 no NaNs occur.

But next question arises. What do the residuals exactly show? I see that rhou rhov and rhow are printed, just wondering about the absolute value, since this is around 1e+06 for all values. Are they normalized in a strange manner?