[gbettega]

Hello forum,
I made a SU2 run in a simple domain in which air (ideal gas) is flowing.
It enters the domain @ 400K and exchanges heat with environment through convection realizing at the fluid domain walls (film coeff 100, Tfar 300; all MKS). Mesh is tetrahedral with inflation layers (SST model with wall functions).

In the results file (ASCII Tecplot) I found, for an => intentionally <= not well converged run, two lines, in which for a point Y+ = 0.0, and in a point Y+ = not a number.

See below.

TITLE = "Visualization of the solution"
VARIABLES = [...] ZONE NODES= 23484, ELEMENTS= 58624, DATAPACKING=POINT, ZONETYPE=FEBRICK
8.604952e-01 6.812059e-01 5.743822e-02 [...] 0.000000e+00 2.085533e+03
1.209390e+00 7.550436e-01 4.031287e-02 [...] nan 1.300315e+03

How should I interpret these findings?
I would like to stress that this is a partially converged run, intentionally.
Thanks
Giovanni

[bigfootedrockmidget]

Y+=0 can occur in corner nodes where the wall shear stress is zero. The nan value is more alarming, do you know where this is occurring?

[gbettega]

Thank you,
I found a certain number of times this issue, because my results reader used (now not anymore) std::stringstream >> value (a double), so finding "nan" It generated an error. I had to introduce a parsing, for jumping over this issue,. So I can exactly recover the node number and its location. I will re-generate the case, and post mesh and .cfg. Thank you
G