Pressure P on the embed boundary

[Morhaf Aljber]

Hello,

I simulated a flow around an object using the centred NS and embed boundaries. my simulation is closer to Karman.c, http://basilisk.fr/src/examples/karman.c.

I have a few questions regarding the pressure.

1: If I defined, p[embed] = dirichlet (0.)

                        pf[embed]   = dirichlet(0.)

I have the pattern in Fig "P_x_dirichlet" and the vortex develops directly behind the object like Fig "level-60_pressure_dirichlet"

However, without setting this condition the pressure is too high compare to with the condition Fig "P_X_free" and the vortex does not develop directly, Fig "level-60_pressure_free". 

I do not clearly understand what is the meaning of this difference and why it is happening.

2: As you see in Fig "level-60..." the domain is square (500x500). If I changed the domain to a rectangular using the mask function (1000x500), I cannot set this condition "p[embed] = dirichlet (0.)" when the resolution is 2^12. But with lower resolution 2^9, the simulation works fine. with a resolution 2^12, I receive this error:

"WARNING: convergence for p not reached after 100 iterations

  res: 38.4083 sum: 12102.4 nrelax: 100 tolerance: 0.07744

WARNING: convergence for p not reached after 100 iterations"

Can you help me to know the reason.

3: what is the difference between p, pf, and g. When I extract those variables pf is similar to p but with higher value. If pf is the face pressure and p is the centred pressure, why does the two variables are different in value. u and uf have similar values when I extract them. For g, what is the final value that g has.

Best regards,

Morhaf

[Haochen Huang]

Hi,

For question 3, the great doc written by Edoardo Cipriano may help.

Best,

Haochen

[Morhaf Aljber]

Dear Haochen,

Thank you very much.

The document and further study helped me a lot.

Best regards,

Morhaf

--
You received this message because you are subscribed to the Google Groups "basilisk-fr" group.
To unsubscribe from this group and stop receiving emails from it, send an email to basilisk-fr...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/basilisk-fr/b921de0d-2846-43bc-ad3c-50747f0258c2n%40googlegroups.com.