|
[Sponsors] |
August 24, 2010, 08:50 |
transsonic nozzle with rhoSimpleFoam
|
#1 | |
New Member
Basti E
Join Date: Jun 2010
Location: Germany
Posts: 7
Rep Power: 16 |
Hi, I am trying to simulate a transsonic nozzle (like a laval nozzle with a convergent and a divergent part) using the rhoSimpleFoam solver. The case is set up to use the SST turbulence modell.
Boundary conditions are as follows: At the inlet I specify T_0 and p_0, and at the outlet I set a static pressure and temperature to zeroGradient. Now the pressure ratio p_static/p_0 determines how the flow through the nozzle will look like: completely subsonic, transsonic or completely supersonic in the divergent part. This setup works fine for CFX and sonicFoam, but for comparitive reasons I want to simulate this with rhoSimpleFoam too. Actually I even think it would make more sense using rhoSimpleFoam instead of sonicFoam since I want to find a steady solution. But I can't get it to work. Here is the output from rhoSimpleFoam: Quote:
1.) bounding omega, min: -4.70953e+07 max: 5.36335e+08 average: 2.42753e+06 omega is not supposed to become negative at any time! 2.) bounding k, min: -0.0701715 max: 4987.9 average: 162.727 same for k, it is not allowed to become negative! 3.) huge continuity errors 4.) when I look at the produced results during these few timesteps the pressure gets massive negative values (-1e1 0Pa region). So obviuosly I am doing something wrong. I just can't figure out what it is ... I haven't changed anything from the setup except adding the SIMPLE dictionary to the fvSolution and changing e to h when going from sonicFoam to rhoSimpleFoam. Any help would be greatly appreciated, thank you |
||
August 24, 2010, 12:06 |
|
#2 |
Member
|
hi Basti
there is a problem in setting the initial guess values or may be with boundary conditions or even it may be with grading part of the mesh at the inlet. so explain these things in detail or share your files. so that you may get some help for solving your problem. you said this setup had worked with sonicFoam solver (did you compare the results). will you tell me which version of OpenFOAM you are using. |
|
August 24, 2010, 15:29 |
|
#3 | |||||||
New Member
Basti E
Join Date: Jun 2010
Location: Germany
Posts: 7
Rep Power: 16 |
Hello,
sry I forgot some info, you are right I should post some more So, I am using OF 1.7.x updated today. Here are boundary and initial settings: Its a 2D mesh. It represents half of the problem since it is a symmetric problem. p: Quote:
Quote:
Quote:
Quote:
Quote:
Quote:
Quote:
|
||||||||
August 24, 2010, 19:53 |
|
#4 |
Member
|
hi
its better to attach your files. i found mistakes you are giving pressure inlet and pressure outlet and zerogradient for inlet and outlet velocities. why k and omega are zero at inlet. please upload your files. |
|
August 25, 2010, 04:34 |
|
#5 |
New Member
Basti E
Join Date: Jun 2010
Location: Germany
Posts: 7
Rep Power: 16 |
Hm, why is it a mistake to specify the total pressure at the inlet and the static pressure at the outlet? ... Based on the pressure difference between inlet and outlet there has to develop a flow with initially unknown velocities at inlet and outlet. Is there any other way to set this up WITHOUT specifying a velocity at inlet and outlet, since I don't know these yet?
|
|
August 26, 2010, 10:30 |
|
#6 |
Member
|
we all know that any fluid flows from high pressure region to low pressure region.
but when we specifying the boundary condition for a subsonic problem we should specify it as velocity inlet and a presure outlet (since there is a characteristic line(in fact there are two but one travels inside) travelling inward to the nozzle so it should be given as pressure oulet). refer CFD book by John.D.Anderson. Jr so define velocity at inlet and pressure(static) at oulet. And give some initial values at inlet for K , Omega hope this solves your problem are you validating this problem. you said you have tried with sonicFoam. would you share that results. |
|
August 27, 2010, 07:23 |
Same issue
|
#7 | |
Senior Member
Nilesh Rane
Join Date: Apr 2010
Posts: 122
Rep Power: 16 |
Hello all,
I am also facing same issue. i am using rhoPisoFoam and k-epsilon model. After some iterations the values of P, k, eps blow up to so hign numbers and the error is shown below. Quote:
I am also giving same kind of BCs as Kiran has alredy suggested, still the error is not getting solved.. I have tried evry possible BC specification methods. But no luck.. PFA my case files, mesh is not there as its 'engrid' mesh and too large to post here.
__________________
Imagination is more important than knowledge..
|
||
April 16, 2014, 03:38 |
|
#8 |
Member
|
Hi all,
As you already know that rhoSimpleFoam is a steady-state solver, while, somebody is getting results from rhoSonicFoam and rhoPisoFoam. I think you must give a try with rhoPimpleFoam with reduced time steps. This thing worked fine for me in a particular case. Further you can validate the results with your theoretical calculations or some empirical considerations. Thanks and Regards, Adarsh Tiwari |
|
July 1, 2022, 06:54 |
|
#9 | |
New Member
Join Date: May 2018
Posts: 18
Rep Power: 8 |
what about a supersonic problem? Which kind of inlet boudary woudl be better?
Quote:
|
||
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Convergent Divergent Nozzle | Hrishikesh | Main CFD Forum | 12 | June 25, 2016 02:06 |
Nozzle Flow - Divergence Problem | Ijaz | FLUENT | 9 | January 11, 2014 04:36 |
C-D nozzle supersonic jet boundary | Gland | FLUENT | 4 | May 24, 2012 00:25 |
calculate nozzle length | oe | Main CFD Forum | 0 | July 30, 2008 14:06 |
compressible flow in a counterflow nozzle | d.vamsidhar | FLUENT | 0 | November 24, 2005 01:45 |