Hello everyone, I met some problems in the process of running the case, could you please help me check it out? I want to do a simulation about the pyrolysis of plastics (the product contains some liquid). I ran it without any errors, but after about 1.5 hours, it reached 0.04s and stopped going further. In addition, if the product involves liquid, do I need to add any other files to the folder?
suliaoronghua.zip (30.5 MB)
This is displayed on the Anaconda page , Could you please have any suggestions to improve it?
Thank you very much!
Please attach all relevant inputs - go to the main menu, select “Submit bug report”, and attach the ZIP file. Thanks.
I’m sorry, it didn’t stop with an error, but Mfix suddenly quit. I need to re-enter mfix in Anaconda to open it, but I can only continue to run for about half an hour, and the time in the lower right corner stops moving forward, so I have to stop this process.This is my ZIP file. I want to know why it stops and turns itself off every time. Is there something wrong with my Settings or is it my laptop? I look forward to your suggestions. Thanks for your attention.
suliao_2022-04-19T153628.163007.zip (459.7 KB)
Hi, we’re looking at this now.
The simulation time is proceeding very slowly - the dashboard shows an estimate of over 18 days to complete the simulation, and it seems to be getting worse.
I got an out-of-memory error and traced this to the number of messages in the console window - your write statements in usr_rates_des.f are creating a lot of output, and while this is good for debugging, the GUI keeps all this text in RAM and ultimately crashes due to memory pressure. We need to address this in the GUI (so that it doesn’t attempt to keep all the console output in memory), but in the meanwhile try reducing the verbosity and see if this helps.
– Charles
Hello Charles, I have deleted all the outputs in the usr_rates_des.f according to your suggestion. Dashboard now shows an estimate of 22 days to complete the simulation. I am puzzled why 0.5s takes so long. I think that the time step may be too small, almost 1E-6 or 1E-7. Do you have any idea to make the time step bigger?
Please try to turn off the reactions. If it runs faster, this means the reaction rates are the limiting factor. Then you would need to review the rates to see if these are too large.
Yes，when I turn off the reactions, it runs fast. I will try to reduce the rate of the reaction. Thank you very much. That’s useful.
Hello, I’m sorry to bother you again. I slowed reaction rate down a lot, and the simulation was a lot faster, down from 24 days to eight hours. However, as time goes by, the time step also decreases from 1e-4 at the beginning to 1e-7 at about 0.04s, and finally dt < dt-min at 0.05s. I would like to ask whether this is normal and whether I should continue to reduce the reaction rate or minimum time step?
I suggest you plot the residuals and see which one is taking longest to converge. In the graphics pane, create a new tab if needed, then click “Residuals”
You can also plot Dt. This may help understand what’s going on.
You can also examine the .LOG file for status messages - I’ve found that this command is useful:
grep ':-' *.LOG | less
which will print out the happy/sad face messages that MFiX uses to report on its status
– Charles
My suggestion when debugging is:
if(pM==1) Return
While you can arbitrarily decrease the reaction rate while troubleshooting, you should do a simple hand calculation to make sure the rate you coded is actually what you want.
Okay,I will try it, Thank you very much for your reply.
Okay,I will try it, Thank you very much for your suggestion.