Floating-point exception - erroneous arithmetic operation

56 views
Skip to first unread message

Richard Petrey

unread,
Aug 3, 2023, 2:02:27 PM8/3/23
to Gpyro Working Group
Hi all - 

I am using Gpyro's capabilities as a thermal property estimator. I have 8 specimens to run through Gpyro. Each specimen was tested in the cone calorimeter at 25kW/m2, 35kW/m2, 50kW/m2, and a second test at 50kW/m2. I have managed to get 7 out of 8 specimens to run successfully. I am having trouble with one of my specimen's however. When launching the run on a parallel processor I am getting the attached error. 

I have tried a number of things to diagnose the issue (whether it be input related or exp. data related). My initial thought is that there is an issue with the ignition times being so soon after the start of the simulation; however, I think this would moreso affect the fitness rather than initialization of the program. Thus far I have tried the following to diagnose the issue:
  • Artificially increasing the ignition time
  • Changing some of the early outliers in the experimental data to align with expected outcomes
  • Running the simulation with a different set of experimental data
  • Removing experimental data for fluxes with many outliers
Has anyone encountered this type of error before? I have also attached my input worksheet for reference. Any feedback would be greatly appreciated!
Specimen6.xls
Capture.PNG
Reply all
Reply to author
Forward
0 new messages