Skip to content

Solver crashes during initialization when particle # above certain threshold #97

Description

@stephendkang

The solver crashes when too many particles are put into the system. To replicate this problem, one can construct the simplest possible simulation (e.g. Nvol_c = 1, Nvol_s = 0, Nvol_a = 0, type = homog) and increase the particle number to "Npart_c = 3000." The program crashes while displaying "0% ..ETA."

  • Changing DAESOLVER settings (MaxNumSteps, IterIC, StepsIC, MemoryIncrements, etc) does not seem to affect this threshold.

  • Physical memory resources of the machine seems unrelated to the problem. Memory usage can be made minimal but the initialization will still fail. (about 600 MB for 2500 particles)

  • The threshold for crashing is platform dependent. On windows (daetools 1.9.0), the threshold is about 2453 (plus minus a few randomness each run) particles. On a linux machine the threshold seems to be above 4100, but eventually the same problem happens. This is possibly related to the C compiler used for these platforms rather than the platform itself.

  • Will probably have to debug daetools to find the exact reason.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Labels

bugSomething isn't working

Type

No type

Projects

No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions