Foam::error::printStack(Foam::Ostream&) with interFoam -parallel
Hello everyone !
I am trying to simulate water coming into a chamber. So I have one inlet, two outlets, walls and an atmospherical top. According to the amount of the flow, water will go to one exit rather than the other. I got this error message but I don't really understand it. I guess it comes from my BC but I don't know if my outlets BC are good or not... Can anyone help ? Thanks ! Code:
nProcs : 4 Here are my BCs Code:
object alpha.water; Code:
object U; Code:
object p_rgh; Code:
object nut; Code:
object k; Code:
object omega; |
Hi,
Let's decompose your error message. To do this let's collect printStack output with the same processor number: Code:
[0] #0 Foam::error::printStack(Foam::Ostream&) 1. The error is FPE (Floating Point Exception) 2. Reason for the error is located in variableHeightFlowRateInletVelocityFvPatchVectorFi eld::updateCoeffs() Since you are running Opt build, there are no line numbers (just ??:?). If you open variableHeightFlowRateInletVelocityFvPatchVectorFi eld.C in editor, go to updateCoeffs method and look for the line that can cause FPE, you find this one: Code:
scalar avgU = -flowRate/gSum(patch().magSf()*alphap); To solve the problem, change the way you decompose your domain. |
Thanks a lot Alexey !
I tried to initialize all the domain with water and the problem has been solved. Now calculations have started ! Let's wait and see if results are good ! Many many thanks ! I have been trying all different kind of BC for my outlets yesterday, thinking that this was where the problem was coming from ! |
Lethu:
I recently encountered the same problem. Can you elaborate on how to solve this problem? |
@RayJohn
I am not lethu but will try to answer you. For variableHeightFlowRate boundary condition to work, you need fluid in cells (at least in one) adjacent to the boundary using this BC. Why you do not have fluid in adjacent cells? It depends on your setup. One possibility, which I assumed, when answered initial question, is parallel decomposition, i.e. fluid-containing region is cut-off by decomposition. Yet, there are LOTS of other variants. |
Quote:
Thank you very much. I successfully solved this problem. |
Good morning,
I'm completely new is the area of cfd. I'm having a subject in the university where one of the works is to analyse a venturi. The projects consists in doing the geometry and mesh in ansys and then export it and do all the rest in openFoam. Given this fact, the part of export is going ok, but i need to use the comand os "makeAxiaMesh -overwrite". The problem apears here, because it gives ame a lot of errors because of segmentation fault. I already saw and the files U,etc and boudary are in agreement. I also saw tons of posts here and in other places and everyting that I did, incluing making the mesh in a different way in ansys don't result. I'm in this for three days, desesperatly need help. the error is this one Revolving nodes Radius to axis: min = 0 max = 2.12867 #0 Foam::error::printStack(Foam::Ostream&) at ??:? #1 Foam::sigSegv::sigHandler(int) at ??:? #2 ? in "/lib/x86_64-linux-gnu/libc.so.6" #3 ? at ??:? #4 ? at ??:? #5 __libc_start_main in "/lib/x86_64-linux-gnu/libc.so.6" #6 ? at ??:? Segmentation fault ps: When I do the command checkMesh it always says that is ok Thanks a lot :) |
Hi,
Your printStack lacks all useful information (since OpenFOAM was built without debugging information). Also your post lacks LOTS of relevant information (OpenFOAM version, makeAxialMesh version, etc, etc). So answer will be generic. If you take a look at header comment in makeAxialMesh.C: Code:
Assumes the following about the grid: |
Hi,
I'm sorry I'm still new in this "World". My version of openFoam is 3.0.1 and of ansys is 18.1. I'm working with the solver simpleFoam and just making half of the geometry. How do I know if the mesh is according with the requirements? Thanks a lot |
Hi,
Let's start with checkMesh output. Import mesh from Ansys, run checkMesh in case folder. Post output as a CODE section in a post, of as a txt attachment. |
1 Attachment(s)
Quote:
This is what I got |
1 Attachment(s)
Attachment 64274
Sorry the other file was really confuse and disorganised. |
Well...
- Grid is 2D - Grid is parallel to XY plane Yet, the rest is difficult to deduce from checkMesh output. So could you also post rotationDict (or if you invoke utility with just command line switches, post command line)? I was not able to find version of makeAxialMesh for OpenFOAM 3.0.x. The last version is for 2.x. |
Quote:
Code:
/*---------------------------------------------------------------------------*\ We are using makeAxialMesh from openfoam-extend, more specifically the Breeder_2.0 version. You can check it out using Code:
svn checkout https://svn.code.sf.net/p/openfoam-extend/svn/trunk/Breeder_2.0/utilities/mesh/manipulation/MakeAxialMesh Code:
makeAxialMesh -overwrite Code:
/*---------------------------------------------------------------------------*\ Here's the snippet Code:
const polyPatch &axisPatch=patches[patches.findPatchID(axisName)]; axisName is axis (as specified in the rotationDict). That's how far I can get not knowing a single thing about CFD... Hope it helps, antipatico. |
Your friend explained the reason for the problem: you set axisName to non-existent patch. Set it to the name of patch you would like to convert into axis.
|
I did it and now it works!
One more question, for the post-processing, to plot the variables in function with time, is there another way, besides paraView? and while in paraView, which are te best options? Thanks a lot, Inês Sebastião |
I am not quite sure, I got your question right. If you would like to visualise your case, there are (to name a few):
- ParaView - VisIt - EnSight - Tecplot Also you can use sets/surfaces function objects to sample your flow fields as a raw/gnuplot/whatever files and then you can use anything you'd like to visualise your fields. |
To visualize my case I know how to do it .
My problem is when i want to plot for example the pressure or the velocity |
"Plot for example the pressure" is rather vague description of what you are trying to achieve. Do you need to plot evolution of a pressure value at a given point? Do you need to plot 2D iso-contours of a pressure field?
|
3 Attachment(s)
I'm a little bit lost with the post processing.
About the question that I did i already found information and got cleared about it! the problem now it's in the residuals parts. I don't understand that much about this topic, so I tried to do different simulations (change the end time, the relaxation factors (because my teacher advised it). The graphics that i got were the ones above, however according to what I have read I think they are not that good. What us your opinion? Thanks a lot for the help! Attachment 64305 Attachment 64306 Attachment 64307 |
All times are GMT -4. The time now is 01:39. |