|
[Sponsors] |
January 22, 2012, 15:41 |
Problem with FloatingObject
|
#1 | |
Member
Pierre
Join Date: Sep 2010
Posts: 57
Rep Power: 16 |
Hi,
i used the floatingObject tutorial as basic case and changed some stuff. (Other box geometrie, no water pillar crashing, but waves by groovy BC, etc.) blockMesh, topoSet, subsetMesh and setFields run without any errors and they create the geometrie i wanted. But when i run interDyMFoam this is what happens: Quote:
It crashes and writes floating point exception. So i tried the refine the mesh (up to 2 million cells) and to reduce the time step (down to 0.0001 s). Both didnt helped. What i can also see in the log is that the velocity of the floatingObject goes very high! So i guess its just falling through the water.. (not so much of a floatingObject ) Any idea why the object isnt bound to the surface? What further information you need? Shall i post some of the case files here? Thanks a lot! Have a nice evening Leech |
||
January 23, 2012, 11:17 |
|
#2 |
Member
Pierre
Join Date: Sep 2010
Posts: 57
Rep Power: 16 |
Hi,
i found this here: http://www.cfd-online.com/Forums/ope...-than-1-a.html There they found that there is a impulse in prgh crashing the case. I tested my floating obejct case (the original one) and it crashed. Then i tested the floating object case on my new setup machine (Foam 2.1.0) and there it runs. Then i used the 2.1.0 files and again changed the mesh. But still it keeps crashing after the velocity goes extremly high and then alpha1 goes high.. Any suggestions? Thanks! Second question: In the linked Thread they solved the problem by fixing the body. I want to fix the body to move in x and y direction. How can i do this? Last edited by Leech; January 23, 2012 at 12:16. |
|
January 25, 2012, 11:48 |
|
#3 | |
Member
Pierre
Join Date: Sep 2010
Posts: 57
Rep Power: 16 |
I tried now the following:
Quote:
But it still crashes. The body still begins to move along the z-axis? I got 4 constraints? How can it move? |
||
January 26, 2012, 12:42 |
|
#4 |
New Member
Mark Beal
Join Date: Feb 2011
Posts: 24
Rep Power: 15 |
Leech,
I suggest running the interFoam solver first until it reached a steady state, and then start using the interDymFoam solver. In the past I've had experiences of large forces in the z direction at the beginning of the simulation which lead to large velocities and displacement of the object. The object might not be able to recover from these initial kicks. Regards, Mark Beal |
|
January 26, 2012, 15:46 |
hello openfoamers
|
#5 |
Member
|
I tried to solve the problem of 3-D pipe flow and i want to use the K-Epsilon turbulence model. it is two phase flow problem. water and air is fed at inlet with different velocity and inside pipe it mix with each other and coming out to the another hand. so i want to plot different regimes like slug, plug, stratified, wavy etc. so plz tel me what is the inlet and outlet boundary condition for K, epsilon, alpha, pressure and velocity. and also tel me is there any roll of nut and nutilda? and tel me which solver i used rasinterFoam, interFoam or interDyMFoam?
plz help me i am so confused. i am waiting for reply jignesh thaker |
|
January 26, 2012, 17:05 |
|
#6 | |
Member
Pierre
Join Date: Sep 2010
Posts: 57
Rep Power: 16 |
Quote:
Anyway i go some strange problems: Just to be shure i copied the floatingObject tutorial fresh and had it running on 4 cores. And after 1,8s (case time) it crashes. But when i run it single core it goes through without crashing. I am getting really confused about this floating object thing. Sometimes it works, sometimes not, sometimes on multicore, sometimes not... |
||
January 30, 2012, 10:15 |
|
#7 |
New Member
Mark Beal
Join Date: Feb 2011
Posts: 24
Rep Power: 15 |
Leech,
Yeah.. that's an annoying problem and I'm not sure of the solution. To me, there appears to be a problem with the interDymFoam solver when used on multiple cores. If you find a solution (or the cause) then I'd be very interested to know it. I'm not working on any dynamic mesh problems right now so I can't offer a solution. Regards, Mark |
|
January 30, 2012, 10:18 |
|
#8 | |
New Member
Mark Beal
Join Date: Feb 2011
Posts: 24
Rep Power: 15 |
Quote:
Jignesh, To start I'd work with some pre-existing tutorials and try to adapt them to your case. You might need to combine a number of different cases to get what you need. I suggest you start a new thread and ask again. Regards, Mark |
||
March 26, 2012, 18:01 |
disable time step adjustment fixes the floatingObject tutorial
|
#9 | |
Senior Member
Nicolas Edh
Join Date: Mar 2010
Location: Uppsala, Sweden
Posts: 123
Rep Power: 18 |
Quote:
I've had similar problem twice now. The first time was with a separate case with interFoam. I experienced unexplainable crashes. The second time was when I tried the floatingObject tutorial. The simulation crashes after 0.73s. Both parallel and serial. In both cases I tried different fvSchemes but it didn't help. What solved the issue for me was to disable time step adjustment. In the floatingObject tutorial I chose the standard time step of 0.01s and disabled time adjustment. After that the simulations completed without issue. I should perhaps say I'm running 2.1.x but there hasn't been any relevant change to the tutorial files since the first release of 2.0.x. Best Nicolas |
||
March 29, 2012, 05:44 |
|
#10 | |
Member
Pierre
Join Date: Sep 2010
Posts: 57
Rep Power: 16 |
Quote:
Hi! Thanks for that tip! It seems to help at least to get it running a bit longer. If you havnt seen it yet, i did a bug report on the crashes concerning interDyMFoam. Link: http://www.openfoam.org/mantisbt/view.php?id=417 Ill continue testing. |
||
March 29, 2012, 16:24 |
|
#11 |
Senior Member
Nicolas Edh
Join Date: Mar 2010
Location: Uppsala, Sweden
Posts: 123
Rep Power: 18 |
Hi,
You could also try to select a time step that insures that the interface courant number is below 0.3. This is suggested inte this report [URL="http://publications.lib.chalmers.se/cpl/record/index.xsql?pubid=100901"]http://publications.lib.chalmers.se/cpl/record/index.xsql?pubid=100901[/URL If you google the author you can find the report as a pdf. Regarding the floatingObject tutorial I could run it in parallel on ubuntu 10.04, with adjustable run time. This was on a three year old amd phenom II. The floatingObject crashed on ubuntu 11.10 on a intel i5 sandy bridge. I've just recompiled on the sandy bridge machine and am going to try to run in in serial once more. (the other case was to big to run in serial and was crashing on SLES10, on 2.0.x) If I still experience crashes I'll post back in your bug report. Good luck with your simulation! |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Problem with interFoam; Wave/wiggle alpha1 behavior | JonW | OpenFOAM | 10 | February 4, 2023 08:27 |
UDF compiling problem | Wouter | Fluent UDF and Scheme Programming | 6 | June 6, 2012 05:43 |
natural convection problem for a CHT problem | Se-Hee | CFX | 2 | June 10, 2007 07:29 |
Adiabatic and Rotating wall (Convection problem) | ParodDav | CFX | 5 | April 29, 2007 20:13 |
Is this problem well posed? | Thomas P. Abraham | Main CFD Forum | 5 | September 8, 1999 15:52 |