CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   How can I make FLUENT solve 1D problem (http://www.cfd-online.com/Forums/fluent/44311-how-can-i-make-fluent-solve-1d-problem.html)

 Saleem March 31, 2007 03:39

How can I make FLUENT solve 1D problem

Hi I have unsteady stratified two phase (air-water) flow in one dimension , I want to predict the change of phase velocities and volume fraction wiich both change with time and axial distance. I have a problem to how can I make FLUENT solve my case with one dimension only (i.e. with the axial distance a long the pipe) ? Is there anyone have idea about this?

Thanks all K. Saleem

 Ramesh.K March 31, 2007 21:08

Re: How can I make FLUENT solve 1D problem

Consider a plane flow in which only parameters in axial direction changes.In Engineering Practically speaking there are no 1D and 2D problems. they are just our assumptions and simplifications of actual 3D problems

 Saleem April 1, 2007 01:39

Re: How can I make FLUENT solve 1D problem

Thanks to the replay. Can you tell me how can i consider my problem as a plane flow with more details?

Regards Saleem

 bashu April 2, 2007 08:28

Re: How can I make FLUENT solve 1D problem

One way to do this is to use 1 cell for your other dimesion when you create the model in Gambit.

 Saleem April 2, 2007 08:58

Re: How can I make FLUENT solve 1D problem

Thank you bashu can you tell me how can i did this in more details? I think drawing 1 dimensional geometry not possible in Gambit.

Saleem

 Tondyn April 2, 2007 10:30

Re: How can I make FLUENT solve 1D problem

I think, for 1-D treament, just image the model consists of tank-in-series. Therefore, in Gambit, for X-Y-Z coordinate system, the grid interval for X-,Y-direction is set to 1. For Z-direction, you can assign the grid number you expected (say,120). Then, you will get the 1-D geometry in Gambit.However, in fluent, you should choose 2ddp/3ddp when starting this software. Evidently, for 1-D treatment, it is better to take 2-D meshing in Gambit (interval for X-direction equals unit, and for Z-direction is 120).

However, there are some things should be paid attention to.

(1)the momentum equation in X-direction should be deactivated;

(2)Whether the 1 interval assignment for X-direction will voliate some convergence (CFL, from grid discretization viewpoint). It is not clear to me. Since from 2d standpoint, the CELL aspect ratio under the predefined X-dir grid (=1) might too large or too small, depending on the system you simulated.

(3)The third problem is how the effect of boundary condition is imposed in the interior Z-dir fluid flow,especially for multiphase flow, say gas-solid two phase flow?

 bashu April 2, 2007 11:55

Re: How can I make FLUENT solve 1D problem

I ran a 1D case which I used for kinetics only. In other words I did not solve for the momentum equations and imposed the velocity/voidage profile on the domain. When I ran a single phase on this 1D domain, the results showed that the velcoity is not changing with length so I am not sure if you can solve the momentum equation.

I used a 2D case and the x direction had 1 cell (I don't think the dimension is important in x-direction), the y direction meshed it as usual.

 Saleem April 3, 2007 03:13

Re: How can I make FLUENT solve 1D problem

I want to command about the replay to both of you:

Mr. Tondyn: 1)Can you tell me where can I find the grid interval in Gambit? 2)In my case I need to solve with the axial distance so I cannot ignore the momentum equation in the axial direction because usually as I did this no change of the velocity will happen? 3) When I asked about the ability to solve like this problem in 1 D my purpose from this is to solve at the interface region only along the pipe length. Such I have two phase air-water stratifed flow with unsteady state and I use the VOF model for this.

Mr. bashu: I asked you many times to how can I did one cell in x-direction?

Thanks a lot Saleem

 bashu April 3, 2007 09:30

Re: How can I make FLUENT solve 1D problem

I am sorry I assumed you are familiar with Gambit and you know where to find the mesh menu in Gambit.

As an example run this journal file in Gambit. Open a Gambit session, and then File --> read journal. It assigns a single cell in x-direction.

face create width 0.1 height 1 xyplane rectangle undo begingroup edge picklink "edge.3" "edge.1" edge mesh "edge.1" "edge.3" successive ratio1 1 intervals 1 undo endgroup undo begingroup edge picklink "edge.4" "edge.2" edge mesh "edge.2" "edge.4" successive ratio1 1 size 0.1 undo endgroup face mesh "face.1" map size 1

 bashu April 3, 2007 09:37

Re: How can I make FLUENT solve 1D problem

face create width 0.1 height 1 xyplane rectangle

undo begingroup

edge mesh "edge.1" "edge.3" successive ratio1 1 intervals 1

undo endgroup

undo begingroup

edge mesh "edge.2" "edge.4" successive ratio1 1 size 0.1

undo endgroup

face mesh "face.1" map size 1

 Saleem April 3, 2007 10:10

Re: How can I make FLUENT solve 1D problem

Sorry i failed to open this can you send me it as journal file by the address khalidb77@gmail.com

 bashu April 9, 2007 07:52

Re: How can I make FLUENT solve 1D problem

Did you run the journal I sent before? You do not make a 1D geometry, but mesh one direction with one cell only.

 Saleem April 9, 2007 11:35

Re: How can I make FLUENT solve 1D problem

No I failed to run the journal you sent. I asked you to send me it by the address below: khalidb77@gmail.com

 bashu April 10, 2007 08:26

Re: How can I make FLUENT solve 1D problem

Just copy these lines in a text file and then save it as a *.jou file, then run Gambit and read the journal file.

face create width 0.1 height 1 xyplane rectangle undo begingroup edge picklink "edge.3" "edge.1" edge mesh "edge.1" "edge.3" successive ratio1 1 intervals 1 undo endgroup undo begingroup edge picklink "edge.4" "edge.2" edge mesh "edge.2" "edge.4" successive ratio1 1 size 0.1 undo endgroup face mesh "face.1" map size 1

 Saleem April 10, 2007 10:30

Re: How can I make FLUENT solve 1D problem

Unfortunately I still unable to make JOU file from text file because this extension not available under text file. I appreciate you to send me the jou file through my email to the last time. My email address is : Khalidb77@gmail.com

 nime April 10, 2007 12:17

Re: How can I make FLUENT solve 1D problem

I should say that I cnnot tolerate the way of asking problem by Saleem.

If you tell me you have some technique problem after making some efforts, then ask bashu to get some solution. It will be understandable. Now, you are trying to rest youself on a untrying level;and ask Bashu several times for the simplest problem even after 10 days. Why not try by yourself and then ask some questions beyond that?

I should say that here I have no any intention to offend or attack anyone. Just give a unbiased comment.

 Saleem April 11, 2007 02:07

Re: How can I make FLUENT solve 1D problem

Mr. Nime thank you for your advice. I want just to remind you this site designed to the all and we must not reduce the importance to every question even if it simple. So what the problem in my question if I not understood the answer from a person he was offered his help to me, second thing who tell you I did not make effort about this? I told Mr. Bashu I failed to make jou file from the text he sent me. So if you want to participate with us in a scientific discussion that's welcome and let the personal views away.

 nime April 11, 2007 10:01

Re: How can I make FLUENT solve 1D problem

Hi,Saleem,

I highly respect the importance of your post. So,I suugest that you can make some efforts to read the Gambit help file which will definitely be useful for understanding the Bashu's meaning. And I am very glad to participate in the interesting discussion. My question is:

For case of gas-solid two phase flow, if adopting kinetic theory of granular flow model as well as using 1D model in Fluent, then how to evaluate the radial segregation effect imposed by the KTGF?

Because we know in this case, all the radial information is highly lumped. But from viewpoint of the mathematical model of KTGF, the radial distribution functions are still there. Therefore, from physical standpoint, I cannot unify the understanding of it.

 Saleem April 11, 2007 10:26

Re: How can I make FLUENT solve 1D problem

Dear Nime: I have two phase (air-oil) flow into a pipe and I was used the VOF model to simulate my results. My need to use 1 D problem is due to predict the value of phase velocities and phase fractions along the interface region. So I think there was a difference between my case and yours because you have gas-solid flow and I am not familiar with the granular flow model so I cannot offer more help in this sorry.I hope you found the person who solve your problem. About my question to Mr. Bashu I understood what he mean and I open journal files before this time and I have no need to read the manual for this but the problem I cannot save the text he sent me as a journal file to use it by gambit so that I send this text below I am so thankful if you kindly save me it as jou file and mail me it by the address khalidb77@gmail.com

face create width 0.1 height 1 xyplane rectangle undo begingroup edge picklink "edge.3" "edge.1" edge mesh "edge.1" "edge.3" successive ratio1 1 intervals 1 undo endgroup undo begingroup edge picklink "edge.4" "edge.2" edge mesh "edge.2" "edge.4" successive ratio1 1 size 0.1 undo endgroup face mesh "face.1" map size 1

 bashu April 11, 2007 10:40

Re: How can I make FLUENT solve 1D problem

I try to reply the best I can. I think since we don't know each other personally there can not be any personal issues involved in our discussion. I also assume if anybody is posting any qustion in this forum, they can not solve it on their own. So here we go:

Dear Saleem, This is one way to create the journal file: Frist you copy the text in a text editor and save it with whatever name and extension that the editor allows you, then you can just click on the the file name and rename it as you wish with any extension as you want. I don't think that the extension of the file is important for gambit anyway!!

Each line is a command in gambit. I am resending the commands.

Just so you know, I don't think this is related to Gambit or Fluent or CFD in general. Sorry I do not want to send any separate emails.

face create width 0.1 height 1 xyplane rectangle

undo begingroup

edge mesh "edge.1" "edge.3" successive ratio1 1 intervals 1

undo endgroup

undo begingroup