Errors from OpenSeesMP, OpenSeesSP using Mumps

This forum is for issues related to parallel processing
and OpenSees using the new interpreters OpenSeesSP and OpenSeesMP

Moderator: selimgunay

Post Reply
YuZhang
Posts: 5
Joined: Fri Dec 19, 2014 11:27 pm
Location: UCLA

Errors from OpenSeesMP, OpenSeesSP using Mumps

Post by YuZhang » Sat May 28, 2016 4:13 pm

Hi everyone,

Here are three errors from my model using OpenSeesMP, OpenSeesSP with system of Mumps.
For problem 1 and 2, it seems that the Mumps solver is not able to switch to the backup algorithms when the present algorithm encounters singularity matrix.
For problem 3, the model did not run for even 1 step.
Hoping someone could help me out. Thanks in advance.

PS: My model worked for UmfPack, and I only changed the system to Mumps.

##########################################################################################################################
# No.1 # Error From OpenSeesMP Using 1 processors (Job aborted without any warnings when the 1st algorithm failed to converge rather than keep tring
other backup algorithms):

CTestEnergyIncr::test() - iteration: 44 last EnergyIncr: 0.00706986 (max: 0.01)
WARNING MumpsParallelSolver::solve(void)- Error -10 returned in substitution dmumps()
cause: Matrix is Singular Numerically
WARNING NewtonLineSearch::solveCurrentStep() -the LinearSysOfEqn failed in solve()
DirectIntegrationAnalysis::analyze() - the Algorithm failed at time 2.52
OpenSees > analyze failed, returned: -3 error flag
Dynamic analysis failed to converge at time 2.480 ...
Algorithm = NewtonInitial; Time Step = 0.040; Remaining Time = 50.520.
##########################################################################################################################


##########################################################################################################################
# No.2 # Error From OpenSeesSP Using 1 processors (Job aborted without any warnings when the 1st algorithm failed to converge rather than keep tring
other backup algorithms):

CTestEnergyIncr::test() - iteration: 34 last EnergyIncr: 0.00350111 (max: 0.01)
WARNING MumpsParallelSolver::solve(void)- Error -10 returned in substitution dmumps()
cause: Matrix is Singular Numerically
WARNING NewtonLineSearch::solveCurrentStep() -the LinearSysOfEqn failed in solve()
DirectIntegrationAnalysis::analyze() - the Algorithm failed at time 2.52
OpenSees > analyze failed, returned: -3 error flag
Dynamic analysis failed to converge at time 2.480 ...
Algorithm = NewtonInitial; Time Step = 0.040; Remaining Time = 50.520.

job aborted:
rank: node: exit code[: error message]
0: Yu-Desktop: -1073741819: process 0 exited without calling finalize
##########################################################################################################################


##########################################################################################################################
# No.3 # Error From OpenSeesSP Using 6 processors (Job aborted without even solving the 1st step of the gravity analysis):

Slave Process Running 5
FEM_ObjectBrokerAllClasses::getNewUniaxialMaterial - - no UniaxialMaterial type exists for class tag 3002

job aborted:
rank: node: exit code[: error message]
0: Yu-Desktop: 123
1: Yu-Desktop: 123
2: Yu-Desktop: 123
3: Yu-Desktop: 123
4: Yu-Desktop: 123
5: Yu-Desktop: -1073741819: process 5 exited without calling finalize
##########################################################################################################################

Post Reply