Blasius Boundary Layer case NaNs for rhou,rhov,rhow

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

thumm_pw.ini (1.05 KB)

thumm_pw.pyfrm (2.85 MB)

residual.csv (309 Bytes)

Hi Gabriel,

What CFL number are you running at?

Peter

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

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?