Temp & Pressure jump at&near wall during iteration
 

Dear all advisers and fellows

Now I'm implementing high temperature nozzle flow surrounded by a structure to deflect plume jets.
This is 3D case contains symmetry plane and the size is about 10x2.5x14(m). And time step is 1e-4sec, Courant number & Under Relaxation Factors are 0.4.
All descritization is 1st order for a quick solution.
I use UDF for developing temperature and pressure of nozzle flow.

My problem is..
During iteration, high temperature and(or) pressure points show up in domain abruptly.
those always occure on a cell neare wall.
To solve this problem, I've remeshed, and patched temp and pressure relative zone when those show up.
But they couldn't remove that point after all.

Now I'm solving my case with fixed wall boundary temperature(300K).
Until now, it works and I will change this wall temperature condition to isothermal or non zero heat flux at a proper time.

I wonder it can be a proper approach to solve my problem.

Somebody give me an advise please.

Thanks in advance.

additional action for this..
 

While waiting comments, I got abnormal point on symmetry plane and cells around here had low orthorgonality(?). So I used Adapt>sweep/smooth function several times and worst orthorgonality went from 0.29 to 3.7.
After initialization, abnormal point didn't appear yet. I guess my problem may be caused by mesh quality.

After several hours...
Nozzle flow seems to be developing well and normal shock is going out from nozzle exit now...
Converged solution is not obtained yet, but at this moment, I maybe able to summary my problem and solution.

there were two problems and their remedies.

1. abnormal pressure&||temp point occurance in fluid zone (those showed up frequently near wall, symmetry)
-> fixed temp at wall (now 300k), patch with marks

2. convergence fail during flow development in nozzle
-> patch with marks until certain time

for the 2nd problem I had similar problem in 2D axisymmetry nozzle. At that time I solve manipulating URF, Courant #, and time step...

Thanks any way