Fixed a bug on pi_inf and enabled subgrid bubble model for viscous flow

[hyeoksu-lee]

1. Fixed a bug that the capital 'pi_inf' was used where the little 'pi_inf' should be used. Also, computing speed of sound is also corrected.

2. Enabled subgrid bubble model for viscous flow

   • Added codes that computes the Reynolds number properly when 'bubbles=T'

3. Modified codes to make 'vel_profile' and 'instability_wave' options compatible with 'bubbles=T' and 'num_fluids=1'

4. Changed grid numbers in example/3D_turb_mixing to prevent memory issue (3d_turb_mixing not working #149)

[anandrdbz]

These changes seem fine, you might wanna change the definition of B_tait in m_qbmm as well. Other than that everything looks fine. The tests won't pass as of now due to this change as the golden files need to be regenerated for this definition of pi-inf

[sbryngelson]

Thanks for looking @anandrdbz . Agreed @lee-hyeoksu you'll have to generate new golden files for the test cases involving sub-grid bubbles. I believe this is documented, but you can also look through the toolchain/ code to see how this is done.

[hyeoksu-lee]

Thanks @anandrdbz and @sbryngelson. The definition of B_tait in m_qbmm is now changed, and the golden files for the cases with sub-grid bubbles are updated.

[sbryngelson]

@lee-hyeoksu your PR is failing GPU tests on the bubble cases I think.

[hyeoksu-lee]

@sbryngelson The failure on GPU was because I didn't add Re_L and Re_R to the private clause for OpenACC when I added codes for computing Re_L and Re_R in m_riemann_solvers. Now it is fixed.

[sbryngelson]

Great, thanks @lee-hyeoksu ! Can you also add new tests for the case of viscous effects + subgrid bubbles? So we can track if code associated with those effects changes or breaks. Again, there should be some documentation on this (adding new tests).

[hyeoksu-lee]

@sbryngelson I added new tests, and they appear in the list by ./mfc.sh test -l. But is there anywhere else I should add documentation?