Finding the reason for numerical instability in FDS

[vijay00901]

Dear friends,

I am running a simiulation in FDS and it is causing numerical instability. I am trying to figure out the reason for that. Would someone be able to point me in the right direction. I cannot attach certain files so I am sharing it here. Thank you so much.

https://drive.google.com/drive/folders/1eA7058j7EIW48ynioIjA2vIaxa2t80P9?usp=drive_link

[vijay00901]

Hi friends, I found out that the initial time step was 0.0583 and the time step at which instability occurred was 6.13 E-6.
There is a condition i.e. if the time step goes below 0.0001 that of initial time step, calculation stops. The value I got is very close to this limit. should I change the LIMITING_DT_RATIO? Please advise.

[rmcdermo]

I cannot look at the Google Drive you posted. Please add the file extension ".txt" to your input file and post it here.

[vijay00901]

Thank you Dr. McDermott. I have attached the files here.
error_file.txt
out_file.txt
model3_0_1g_steps.csv

[rmcdermo]

You have not included an input file. From the .out file, I cannot really tell what is wrong. Also, please try with the latest code. Quite possible we have sorted out a problem that will help.

[vijay00901]

Sorry I forgot to include that. I have attached it here
input_file.txt

[rmcdermo]

OK, so we are doing microgravity combustion with a premixed flame?

Please let's start simple. You are way into research territory here and the very first thing you send me is super complicated. Can you just do a simple flame speed calculation with the chemistry you are giving FDS? Try that with the latest code and report back.

[vijay00901]

Yes Dr. McDermott. I am actually doing research. What you said will be a digression from my end goal. I got the simulation run successfully for various gravity values like 0.5g, 0.3g, 0.2g but 0.1g is having this issue. Is there any way I can pinpoint the issue or solve it? Thank you so much.

[mcgratta]

If you are doing research and you do not do the simple verification exercise that @rmcdermo suggests, then your research is doomed to failure. You are using FDS in a way that most people do not normally use it. It is very important that you check to ensure that the model assumptions are appropriate for your case.