CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   How to model this case in fluent? (https://www.cfd-online.com/Forums/fluent/45395-how-model-case-fluent.html)

hjzy July 13, 2007 18:34

How to model this case in fluent?
 
Hi, fluent fans:

I have a problem in solve this case:

the tube has solid walls all around - same boundary conditions on all four sides - left side, right side, top side, and bottom side. There is a high pressure region on the left and a low pressure region on the right separated by a diaphram. At t=0, the diaphram is removed and the pressure will create a flow from left to right.

Which kind of model can I use to simulate this condition. I can try it as a 1d or 2d problem.

Thanks!

hjzy July 16, 2007 18:53

Re: How to model this case in fluent?
 
this is a shock tube problem. I have tried many method, but still can not the result, did any one can help?

Thanks in advance!

sujith July 17, 2007 15:20

Re: How to model this case in fluent?
 
This problem should possible using unsteady density based solver. you have to patch the initial values of pressure. I havn't tried this myself. If you have already tried this and had any difficulties, please share it with us


yzheng July 17, 2007 20:11

Re: How to model this case in fluent?
 
I have tried this problem, but still can not get accurate result.

I just make a rectangular, 8.2m Ã- 2m, four sides all defined as walls, and make a line just between the center(x=0) as an interior, then make two zones, left has high pressure at about 3.5atm and density=0.445, the right side has the lower pressure at about 0.5atm and density=0.5. I just use patch after the initialize.

In fluent, read the case, then check the grid, it's ok. Define-> models->solver->coupled->unsteady Define->energy Define->viscous ->k-epsilon(2-eqn) Define->material ->ideal gases Define->operation condition ->operation pressure=0 Define->boundary condition(nothing need to define)

Solve->controls->solution controls->pressure velocity coupling->piso

Solve->monitors->convergence criterion(change all values to 0.000001)

Solve->initialization->gauge pressure=3.5atm, temperature=400k

Solve-> initialization->patch initialization->set the pressure for the left zone3.5atm, temperature=2789k, set the pressure for the right zone0.5atm, temperature=400k, set all the velocity=0

Solve->iterate, set the time step size=0.0001s, then iterate. But it can not convergence, after about 500 steps, it is said, divided by zero.

The problem is the pressure of the left side changed a lot, especially at the x=-4.1m, but the center is changed not so much, and the pressure in the right zone changed a little near the center. This is not as expected, I don't know why?

Any advices? Thanks in advance.


yzheng July 17, 2007 20:13

Re: How to model this case in fluent?
 
I only have the version of 6.2, so it don't have density based solver.


All times are GMT -4. The time now is 07:59.