segmentation Error

[jith178085]

I'm working on wood pyrolysis. The domain has 12 million cells. My HPC has 500 GB ram and 60 cores. I divided the domain into 32 mesh parts for parallel, among 4 domains have 2 million cells each. Can anyone help me what im missing here. When I run it's throwing segmentation error

[drjfloyd]

We would need your input file to determine the cause.

[jith178085]

Grouped.doc

sorry for the delay,
This code runs perfectly, when I remove the obstructions (Both wood and brick wall)
(did not include the obstructions due to new design)

[mcgratta]

The cases Grouped is running for me using the latest source code. Does the case you posted work for you? I do not understand what you mean by "did not include the obstructions due to new design".

[jith178085]

Grouped.docx
Could you please try this. Previously I did not include the wood due to large no of lines of code. I thought some error in mesh. This is updated code. I also attached the screenshot of error

[drjfloyd]

I ran the new file with T_END = 0 so I could see the overall geometry. You seem to have a 5x4 bundle of wood rods. The STL file conversion as converted these all to just faces. All your OBST inputs are 2D objects. For example here is one OBST input:

&OBST ID='Body_Top_mesh_Dia_3.STL', XB=0.015,0.0225,0.136,0.138,8.0E-3,8.0E-3, SURF_ID='WOOD'/

I suspect that you want these wood rods to be solid and not a collection of 1000s of little hollow cubes which is what your inputs currently specify. Having this large number of tiny enclosed volumes surrounded by thin single grid cell obstructions may be the cause of your segfault.

Check out 7.6.2 in the User's Guide. You could just define the cylinders using OBST with SHAPE and then use HOLE to truncate them on the lower end as you have now in the input.

[jith178085]

Now I created all the obst
Grouped.docx
in FDS itself, now the segmentation error rectified, the simulation ran for 12 time steps (0.001 to 0.00127) then numerical instability came

[drjfloyd]

Do you have the correct units for your A and E for your wood reaction? E=185 is orders of magnitude lower than expected for an activation energy.

Did you run a virtual TGA and cone test for your wood reaction as recommended in the User's Guide 9.3 Testing Your Pyrolysis Model?

[jith178085]

I didn't perform the experiment, I took the values from literature. Is there any database for these values, that I can directly use for FDS

[drjfloyd]

We give the units for E used by FDS in mulitple places in the User's Guide. There are even multiple example inputs provided for you. As the User you are responsible for reading the Guide and making sure you understand the units being used and how those units are being used. You should not just blindly apply values from a paper without first making sure they have the correct units and that the reaction mechanism correspond to the pyrolysis model in FDS for the specific version you are using. This is why we suggest in the Guide that you test your reaction mechanisms before moving on to a full scale simulation.

[jith178085]

My code is running fine. Please close this. Sorry for the delay