# Divergence

 Register Blogs Members List Search Today's Posts Mark Forums Read

 April 13, 2005, 19:49 Divergence #1 Docker Guest   Posts: n/a Can anyone tell me the fundamental reasons why divergence occurs if one tries to use a mass flow at one boundary and pressure at the other? Regards, Docker

 April 13, 2005, 23:12 Re: Divergence #2 pUl| Guest   Posts: n/a I do not think divergence errors are directly related to boundary conditions chosen. Instead they are linked to your problem definition. Personally, I follow some of the guidelines in the Fluent manual and things shape up very well. For instance, once you find continuity diverging after a particular amount fo time, you might want to: a. Lower some URF's, say momentum and body forces (judiciously) at the problem start and then increase them by small amounts. b. Try to solve for lesser equations till the flow develops. Like turn off "volume fraction" solution till the initial flow field has developed. For my own multiphase case, the 6.2 manual now says that one can now restrict the application of lift forces only to the specific areas in the domain where the shear rates are significant, which is quite sensible and IMO will avoid divergence issues. But they have not mentioned how this can be done. Perhaps through a UDF?

 April 14, 2005, 07:16 Re: Divergence #3 Docker Guest   Posts: n/a Thanks very much for taking the time to respond. I am new to all this. On a seperate issue of mine, which is causing the most trouble: I've set up an unsteady run with an inlet manifold as my domain. Pressure inlet and UDF for pressure outlet. I now want to be able to turn the outlet from a 'pressure outlet' to a 'wall', when then valve is closed. So I need to able to change the boundary condition at certain time steps. Any ideas? Heard of a method using a 'hand-written' journal file? Thanks again, Docker

 April 15, 2005, 06:04 Re: Divergence #4 Abu-Khawlah Guest   Posts: n/a Hi, there I think you should only run your case with pressure-outlet for define time, then re-define the outlet as a wall and start the problem again, Fluent will use your last case data as an intial data for the new case, Howeve, if you need to open and close the out let valve very frequently, then I thing you need to use a UDF macro. All, the best Abu-Khawlah

 April 15, 2005, 07:47 Re: Divergence #5 Docker Guest   Posts: n/a A very good idea. Unfortunately you pre-empt the problem. The valve does open and close frequently. And there are 4 inlet valves (outlets to the domain)! How, in a macro can you define a change in boundary condition. Now I don't mean a change in a value - that is easy, I mean a change form 'pressure-outlet' to a 'wall' (at stated time steps)? This is the thing that I do not understand? Am I missing something, or is this tricky? Thanks for your help, Docker

 April 15, 2005, 09:20 Re: Divergence #6 Abu-Khawlah Guest   Posts: n/a I dont know, if you can change the boundary condition status using the UDF, However, I mean you should use a proper UDF macro to represent you physical bounday conditions, like maybe you can define your outlet as an WALL and link it with a UDF source macro allowed the fluid to leave the system once it hit that wall and also pre-define the surface with pressure profile that change with the physical status of your case (open valve=>run the sorce code+applie a pressure) or (close valve=>disactivate the source code+solid wall pressure) I hope it is cleare for you, it is just a suggestion, I never didt before, so if you will go for it, make sure it will give you the same pressure distribution like real case, I suggest you run different cases with and without avalves to be a reference cases for you All the best AbuKhawlah

 April 17, 2005, 05:46 bbbb #7 bbbb Guest   Posts: n/a bbbbb

 April 17, 2005, 19:02 Re: Divergence #8 Docker Guest   Posts: n/a Thats clear, yes. I know what you mean re the pressure in the UDF. I am using that technique. And, yes, it is a good suggestion to try both methods, to compare. Thanks for the suggestions. Docker

 Thread Tools Display Modes Linear Mode

 Posting Rules You may not post new threads You may not post replies You may not post attachments You may not edit your posts BB code is On Smilies are On [IMG] code is On HTML code is OffTrackbacks are On Pingbacks are On Refbacks are On Forum Rules

 Similar Threads Thread Thread Starter Forum Replies Last Post cris FLUENT 3 September 4, 2014 18:06 MY FLUENT 3 January 11, 2014 05:46 SamCanuck FLUENT 2 August 31, 2011 11:34 shekharc Main CFD Forum 7 July 5, 2005 12:08 Maciej Matyka Main CFD Forum 2 December 19, 2000 11:43

All times are GMT -4. The time now is 09:44.