Coupled Calculation--Roughness Wall Boundary Conditon

[luyitang]

Hello,

I did coupled calculation with Syrthes, in my case, superheated steam flows in a tube, where the tube wall is heated. The heat conduction of the tube wall is calculated with Syrthes, and the inner face of the wall is set as interface between C_S and Syrthes. When in C_S, the boundary condition for the surface is set to be smooth, the calculation runs properly, with a reasonable pressure drop. However, when I changed the Smooth wall to Rougn wall with a roughness coef, the case can't run. So I suppose the coupling calculation doesn’t accept rough wall as a boundary condition.


So I want to add an additional head loss to model the friction pressure drop.

In the module of head loss, the head loss coefficient K should be defined by the user. The head loss term in the momentum conservation equation is that:
ρ*∂U/∂t=-grad P+⋯KρU(1)
K=1/2 αlpha |U| (2)
KρU^2=1/2 αlpha*ρ*U^2 (3)
I compare equation ∆P=0.5*f_D*L/D_h *ρU^2 and (3), it shows that when the head loss only dues to friction, αlpha is equivalent to f_D*L/D_h . While in the code, friction factor of smooth wall is computed, what we should define is the roughness part in addition to the smooth part, so I set αlpha=(f_D_rough-f_D_smooth)∙L/D_h.

the αlpha=4.2 in my case, I define the whole fluid volumical domaine as a head loss with αlphaXX=4.2, αlphaYY=4.2, αlphaZZ=4.2.

Finally, however the pressure drop I get is higher than I expected.

Could you help me find out a way to solve my case? Is my calculation with alpha right?

Best regards
Luyi

[Yvan Fournier]

Hello,

How much bigger is the pressure drop ? If it is a factor 2, there might be a mix between K and K/2.

Otherwise, did you check the friction part ? Depending on the combination of mesh refinement and turbulent model used, you might get a bad estimation of the wall friction, which could explain your incorrect results, so this is the first thing to check.

Otherwise, what error message do you get when using a rough wall with Syrthes ? This is not tested, but it might be a minor bug, and relatively easy to fix.

Regards,

Yvan

[luyitang]

Hello Yvan,

Thank your for replying.But I didn't get what you mean by mix between K and 1/2K. In GUI, the head loss module, it is demanded to enter alpha XX, alpha YY,alpha ZZ, αlpha is equivalent to f_D*L/D_h, am I right?

And I didn't really understand you said about "check the friction part". As the Y+ number is reasonable (around 30 ) and in the smooth calculation, the pressure drop is right compared to f_D=0.184*Re^(-0.2) (or find the corresponding f_D from Moody Chart). Is f_D the parameter should I verify?


For the error file that in C_S is :

Code: Select all

Call stack:
   1: 0x7ff1ab2ebd73 <vericl_+0x8703>                 (libsaturne.so.0)
   2: 0x7ff1ab1d1d97 <condli_+0x57d>                  (libsaturne.so.0)
   3: 0x7ff1ab2d4454 <tridim_+0x2fb4>                 (libsaturne.so.0)
   4: 0x7ff1ab1bdd95 <caltri_+0x2805>                 (libsaturne.so.0)
   5: 0x7ff1ab18bde5 <cs_run+0xa65>                   (libsaturne.so.0)
   6: 0x7ff1ab18bf45 <main+0x155>                     (libsaturne.so.0)
   7: 0x7ff1a89eec8d <__libc_start_main+0xfd>         (libc.so.6)
   8: 0x4007c9     <>                               (cs_solver)
End of stack

And there is no error file in Syrthes. The error appears in the Terminal is :

Code: Select all

Post-processing.. 
  .syrthes (.res) --> ensight... 
     --> /home/tangly/syrthes/syrthes4.0.1/arch/Linux_x86_64/bin/syrthes4ensight -m /home/tangly/SH2/solid/Solid_3.syr -r /home/tangly/SH2/RESU_COUPLING/20130528-1757/solid/resu1.res -o resu1
Traceback (most recent call last):
  File "./runcase_coupling", line 151, in <module>
    save_results=True)
  File "/home/tangly/code_saturne/arch/calibre_V3.0beta/lib/python2.6/site-packages/code_saturne/cs_case.py", line 1476, in run
    self.save_results()
  File "/home/tangly/code_saturne/arch/calibre_V3.0beta/lib/python2.6/site-packages/code_saturne/cs_case.py", line 1384, in save_results
    d.copy_results()
  File "/home/tangly/code_saturne/arch/calibre_V3.0beta/lib/python2.6/site-packages/code_saturne/cs_case_domain.py", line 1225, in copy_results
    raise RunCaseError(err_str)
cs_case_domain.RunCaseError: 
   Error during SYRTHES postprocessing

No error is produced when the wall is smooth.

Best Regards
Luyi

[Jacques Fontaine]

Hello,

The error in Syrthes is a consequence to the error in CS (Syrthes is trying to post-pro the inexistent results).

Could you post the last lines of the listing (and not only the error)? And could you run your case with CS complied in debug mode?

Regards,
JF