Convergence issues in Natural convection heat transfer problem

[smukthiHT]

Hi All,

I am simulating a steady state Natural convection heat transfer problem in Acusolve. The solutions seems to be not converging.
I have no idea what migght be causing the convergence issues.
I am experienced in Structural analysis and trying to learn CFD analysis.
I would really appreciate, if you can let me know what could cause convergence issues in Natural convection problem and how do I debug them and fix them.

Thanks
Sridhar

 

[IRstuff]

Why wouldn't you post this question in the Acusolve forums?

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529


Of course I can. I can do anything. I can do absolutely anything. I'm an expert!
There is a homework forum hosted by engineering.com:

http://www.engineering.com/AskForum/aff/32.aspx

 

[smukthiHT]

Sorry..did not know there was an Acusolve forum.
I was looking for more general answer for factors that could lead to convergence issues in a CFD model for a Natrural convection problem.
Answer specific to Acusolve would be great..I am trying to work with Acusolve tech support too...still waiting for a response from them.
I the mean time I am trying to playu with the model too..

Thanks!

 

[btrueblood]

AFAIK there is no "Acusolve Forum" on this site...nor a standalone forum for many other CFD solvers.

First, do you know if your solver formulation can handle natural convection (bouyancy effects primarily, and heat flux boundary conditions)? Why do you believe the solution is not converging - is there an oscillation occurring in the mass flow rates, residuals, etc.? Can you describe the problem geometry, and what boundary conditions you are using? I would suggest you look at the outlet flow boundary, and see if you can compute the outgoing heat flux, and compare to input heat flux, averaged over time (hopefully many oscillation cycles).

 

[IRstuff]

Not here, but on the Acusolve website:

http://forum.altairhyperworks.com/index.php?/forum/45-acusolve-and-virtual-wind-tunnel/

A "general" question about convergence is a bit moot; all CFD solvers, while superficially similar, will have different coding and implementation of the finite-element equations and solvers, which will manifest as different idiosyncratic behavior with respect to convergence.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529


Of course I can. I can do anything. I can do absolutely anything. I'm an expert!
There is a homework forum hosted by engineering.com:

http://www.engineering.com/AskForum/aff/32.aspx

 

[tkordyban]

Convergence with natural convection is a common issue, because the driving force for flow is so weak. I would check things first:
1. Make sure gravity is turned on. By default many tools leave this option turned off. Check the Help for how to specify the value and direction of gravity. Without gravity there is no buoyant force to drive natural convection.
2. Turn on double-precision math if you have this option. The buoyant force is so weak it can be "rounded out" in single precision calculations and so again there is not driving force.
3. Is there a temperature difference from one place in the fluid to another? You can't just specify power dissipation someplace. Somewhere else there must be a temperature boundary condition. Temperature differences in the fluid drive the flow.

There are hundreds of other reasons for divergence, but these are the first things to check.

 

[SomptingGuy]

There may not even be a converged steady-state solution.

Steve