Floating point error in sq_contact_newton dpmethod_mod_MOD_func_dp_a

MFiX How to
#1

Hi, I am trying to seed some SQP particle in some given geometry to generate particle output dat file (to be used in another simulation). However I am continuously getting floating point error after a certain time period.

It says:

Program received signal SIGFPE: Floating-point exception - erroneous arithmetic operation.

Backtrace for this error:

#0 sq_contact_newton_dpmethod_mod_MOD_func_dp_a

at des/sq_contact_detection_newton.f:964

#1 sq_contact_newton_dpmethod_mod_MOD_sq_contact_newton_dp_a

at des/sq_contact_detection_newton.f:267

#2 sq_contact_evolution_mod_MOD_sq_evolution_newton_method_a

at des/sq_contact_detection_evolution.f:115

#3 sq_calc_force_mod_MOD_calc_force_superdem

at des/sq_calc_force_superdem.f:354

#4 des_time_march_MOD_des_time_step

at des/des_time_march.f:193

#5 run_dem

at mfix.f:214

#6 run_mfix

at mfix.f:149

#7 main_MOD_run_mfix0

at main.f:85

Floating point exception (core dumped).

I have also attached the mfx file for reference. How can i possibly resolve this issue

MI_SQP.mfx (12.2 KB)

#2

Can anyone help me with this, if I can get some idea of what might be causing this, that will help a lot.
Thank You

#3

Hi @AnshT18

The Superquadric code can produce extremely large numbers when calculating particle overlaps ( overlap between two superquadric particles is too large!) - unfortunately at this time we do not have an easy fix for this, but you can try turning off overflow checks:
shot-2024-03-15_07-09-37

If you do this, make sure to check your results carefully.

The proper fix may be to just disable overflow checking in the section of code which is known to produce overflows, but at this time there is no easy way to do that.

Sorry for the difficulty and thanks for your report. Please keep us posted on any results.

– Charles