CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Generate solution (fluent) (http://www.cfd-online.com/Forums/fluent/29405-generate-solution-fluent.html)

 Nic February 25, 2002 15:43

Generate solution (fluent)

I have been struggle to generate result for my model. However, when i try to iterate my model first with k-e model then using laminar model, it suprrisingly solved... my question -- are these steps were allowed? will it generate a fault result? i know this question sound silly, but i really need your help. i am new in using the Fluent :) thank you

 Jin-Wook LEE February 26, 2002 00:39

Re: Generate solution (fluent)

In general, for turbulnt flow, it is impossible to obtain cponverged solution by laminar solver. However, if you obtain two different solution by laminar solver and by turbulent solver, two results should be different from each other. Especially, for laminar flow, if you use turbulent option, you can in general easily obtain the converged solution. But it is very unreallistic result.

Actually speaking, as a concept, all turbulent flow can be solved by time-dependent laminar solver. This is the concept of the DNS(Direct Numerical Simulation). And there might be two different solutions in the real situation, one belong to turbulent regime and one belong to laminar regime. However, this is fairly difficult topic for the beginner. It is related to the hydrodynamic stability, the degree of pertubation and so on. So, just refer this comment in the future. And at this time, for the beginner, I can say that two flow regime are separated, so that, first of all, you should check that your flow is turbulent or lamninar. And use appropriate option(laminar or turbulent).

Sincerely, Jinwook

 Nic February 26, 2002 06:54

Help agian.. Generate solution (fluent)

Thank you to Jin-wook Lee.

My situation is like this. I am working on Jet impingement, and require to determine the pressure distribution along a flat surface which is normal to the air jet. I am using Gambit and Fluent. My model based on low Re. no. less than 2000. I have no problem to get the solution solved (using laminar model) when the distance between the jet exit to a non-moving flat surface is r(radius) and 5r. However, when i increase the distance between the jet exit and the non-moving flat surface to 10r and 25r, it can't be converge(using laminar model). Jet velocity inlet = 5m/s Air density = 1.225kg/m2

To all the Gurus, senior....i need hint and guide line when dealing with this problem.

@@I feel sad as i can't get it right. However, now i am more 'addicted' to CFD, is this kind of strange?? :)@@

 Jonas Larsson February 26, 2002 07:36

Re: Help agian.. Generate solution (fluent)

Impinging jets tend to become unsteady very easily, you might even have transition occuring when you increase the jet-length. This would of course make it difficult to converge - try running it unsteady. You should also compute the Reynolds number of the jet to see if you can expect transition - wihtout having done any calculations on it a jet-length of 25r sounds long to be laminar, are you sure it should be laminar? On which dimensions did you base the Re=2000 figure?

 nic February 27, 2002 06:27

Re: Help agian.. Generate solution (fluent)

Thank You Jonas Larsson. After reading your reply, i had make a double check on my calculation on the re. no. For my model(1 of the unsolved), distance between jet and flat non-moving surface = 10r distance spacing between jets = 10r Air density = 1.225 kg/m^2 velocity = 5m/s viscosity = 1.7894 x 10^-5 Re.no = (density x velocity x length)/viscousity Assumption r=2mm

 Jonas Larsson February 27, 2002 09:59

Re: Help agian.. Generate solution (fluent)

My guess is that you are passed the critical Reynolds number - for pairs of jets with no impingment you will start getting Hopf-bifurcations and unsteady/chaotic flow at Reynolds numbers around something like 50 (based on jet diameter) I think. In your case your Reynolds number based on jet diameter is 400 if I interpret your formulas right (I assumed that "length" in your Re number formula is 10r).

Hence, you are most likely entering an unsteady transitional flow regime which is very difficult to simulate using a laminar steady approach.

 Nic February 27, 2002 12:00

Re: Help agian.. Generate solution (fluent)

thank you, Jonas Larsson. I will try to solved it by reducing the velocity inlet, and make sure the flow will be in the laminar region. i hope to obtain solution soon... :)

 All times are GMT -4. The time now is 14:40.