Hi, again,
I've tried to use the user defined coordinates, but I get the same 'SIGFPE signal (floating point exception) intercepted!' error.
I need to have the the velocity vector perpendicular to my inlet, which has his normal inclined of 22.5° respect to the x axe, so I used as user defined coordinates cos and sin of 22.5°...but it does not work!
Are these the correct value to input or or my thoughts are wrong?
wind speed profile
Forum rules
Please read the forum usage recommendations before posting.
Please read the forum usage recommendations before posting.
-
Yvan Fournier
Re: wind speed profile
Hello,
Using cos and sin if the inlet is inclined is probably right, but the problem lies in the (z/10)^0.34 formula,, which is not defined if z < 0. Checking your mesh (as posted above), z ranges for the inlet from -0.6 to 23.6. Also, the inlet seems perpendicular with the x axis (whether imported by Code_Saturne or by SALOME), so I am not sure your mesh is the correct one.
Did you check and visualize your mesh (using the mesh quality check) before running the calculation ?
Best regards,
Yvan
Using cos and sin if the inlet is inclined is probably right, but the problem lies in the (z/10)^0.34 formula,, which is not defined if z < 0. Checking your mesh (as posted above), z ranges for the inlet from -0.6 to 23.6. Also, the inlet seems perpendicular with the x axis (whether imported by Code_Saturne or by SALOME), so I am not sure your mesh is the correct one.
Did you check and visualize your mesh (using the mesh quality check) before running the calculation ?
Best regards,
Yvan
-
Filippo Monari
Re: wind speed profile
excuse me Yvan...I didn't tell you something...
I'm running several simulation rotating the far domain with a step angle of 22.5, in order to simulate the main wind directions I have.
The first (the mesh you checked) worked fine, also didn't using the user specified coordinates.
Then I have tried to run the second simulation (rotating the inlet) setting the user coordinates as I said in the previous message, but I got the usual SIGFPE error.
The parameters I used to built the meshes are almost the same the only difference is the perpendicular direction to the inlet.
Anyway here is the mesh...
http://dl.dropbox.com/u/5081283/Mesh_1.med
Thank you for yours very useful helps and I apologize for my continuous questions but my knowledge about CFD and the tool is growing right now with this project.
I'm running several simulation rotating the far domain with a step angle of 22.5, in order to simulate the main wind directions I have.
The first (the mesh you checked) worked fine, also didn't using the user specified coordinates.
Then I have tried to run the second simulation (rotating the inlet) setting the user coordinates as I said in the previous message, but I got the usual SIGFPE error.
The parameters I used to built the meshes are almost the same the only difference is the perpendicular direction to the inlet.
Anyway here is the mesh...
http://dl.dropbox.com/u/5081283/Mesh_1.med
Thank you for yours very useful helps and I apologize for my continuous questions but my knowledge about CFD and the tool is growing right now with this project.
-
Yvan Fournier
Re: wind speed profile
Hello,
If you rotate the mesh with user defined coordinates, I would need your coordinates subroutine to check your case.
In any case :
for each rotated mesh, run the code with quality criteria enabled, so as to visualize the mesh and checl both that your rotation is correct and the z >= 0 everywhere on the inlet. If you have user subroutines modifying the mesh, checking quality from the GUI will not be enough, but this can be done by setting the following option in the runcase file :
Regards,
Yvan
If you rotate the mesh with user defined coordinates, I would need your coordinates subroutine to check your case.
In any case :
for each rotated mesh, run the code with quality criteria enabled, so as to visualize the mesh and checl both that your rotation is correct and the z >= 0 everywhere on the inlet. If you have user subroutines modifying the mesh, checking quality from the GUI will not be enough, but this can be done by setting the following option in the runcase file :
If you have a stack trace (as before), check if the call stack is the similar to the one you had previously (i.e. you are crashing in the same place, or you made progress and are running into a different issue).ARG_CS_VERIF=-q
Regards,
Yvan