05-14-2017 08:14 PM - edited 05-14-2017 08:39 PM
I've been having some convergence issues with a surface-to-surface contact inside NX NASTRAN/Simcenter 11 and I was wondering if I can get some help/ideas on how I can figure out what is causing the non-convergence and how I can fix it.
I have defined two surface-to-surface contacts (with a coefficient of friction = 0.2) and four regions (a source and target region for each of the surface-to-surface contacts).
I don't have a nonlinear parameter modelling object/card being used right now (mostly because I don't really know how to use the card even after reading Blas' response to another question on the engineering tips forum).
*edit*
Looks like that I can't use the NLPARM card anyways for SOL601,129.
Here is a snippet of the input deck:
$* UNITS: mm (milli-newton) $* ... LENGTH : mm $* ... TIME : sec $* ... MASS : kilogram (kg) $* ... TEMPERATURE : deg Celsius $* ... FORCE : milli-newton $* ... THERMAL ENERGY : mN-mm (micro-joule) $* $* IMPORTANT NOTE: $* This banner was generated by Simcenter and altering this $* information may compromise the pre and post processing of results $* $*$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ $* $*$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ $* $* FILE MANAGEMENT $* $*$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ $* ASSIGN OUTPUT2='advnlin.op2',UNIT=21 $* $*$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ $* $* EXECUTIVE CONTROL $* $*$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ $* ID,NASTRAN,parts_x_t_sim1-solution_1 SOL 601,129 CEND $* $*$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ $* $* CASE CONTROL $* $*$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ $* ECHO = NONE BCSET = 100 BGSET = 100 TSTEP = 100 OUTPUT BCRESULTS(TRACTION,FORCE,PLOT) = ALL BGRESULTS(TRACTION,FORCE,PLOT) = ALL DISPLACEMENT(SORT1,PLOT,REAL) = ALL SPCFORCES(PLOT,REAL) = ALL STRAIN(PLOT,REAL,VONMISES,FIBER,CENTER) = ALL STRESS(SORT1,PLOT,REAL,VONMISES,CENTER) = ALL $* Step: Subcase - Nonlinear Implicit SUBCASE 1 LABEL = Subcase - Nonlinear Implicit LOAD = 5 DLOAD = 301 SPC = 101 $* $*$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ $* $* BULK DATA $* $*$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ $* BEGIN BULK $* $* SOLUTION CARDS $* BCTSET 100 3 40.200000 + + 5 60.200000 TSTEP 100 3001.0000-5 3 $* Simulation Object: Face Gluing(1) BGSET 100 1 21.000000 0.010000 $* $* PARAM CARDS $* PARAM ALPHA1 0.0000 0.0000 PARAM ALPHA2 0.0000 0.0000 PARAM F56 YES PARAM LGSTRN 1 PARAM OIBULK YES PARAM OMACHPR YES PARAM POST -2 PARAM POSTEXT YES PARAM UNITSYS MN-MM $* $* GRID CARDS
I'm not sure what kind of outputs you would be interested in seeing (and from where), so please let me know. I have .f06, .f04, .f56, .log, and .diag available.
Below is the output from .f06:
STEP NUMBER = 94 ( TIME STEP = 0.10000E-04 SOLUTION TIME = 0.94000E-03 ) DIAG ELEMENT (WITH MAX ABS VALUE) OF THE FACTORIZED MATRIX = 0.10420E+09 NODE = 50554 DOF = X-TRANSLATION DIAG ELEMENT (WITH MIN ABS VALUE) OF THE FACTORIZED MATRIX = 0.13450E+07 NODE = 22984 DOF = X-TRANSLATION I N T E R M E D I A T E P R I N T O U T D U R I N G E Q U I L I B R I U M I T E R A T I O N S OUT-OF- NORM OF CONVERGENCE RATIOS CONVERGENCE RATIOS OUT-OF-BALANCE LOAD BALANCE OUT-OF-BALANCE NORM OF INCREMENTAL FOR OUT-OF-BALANCE FOR INCREMENTAL VECTOR CALCULATION ENERGY FORCE MOMENT DISP. ROTN. CFORCE ENERGY FORCE DISP. CFORCE BETA RATIO NODE-DOF NODE-DOF NODE-DOF NODE-DOF CFNORM MOMENT ROTN. (ITERNS) MAX VALUE MAX VALUE MAX VALUE MAX VALUE COMPARE WITH COMPARE WITH ETOL RTOL DTOL RCTOL 1.00E-03 (NOT USED)(NOT USED) 5.00E-02 ITE= 0 1.32E+05 1.68E+05 0.00E+00 1.30E+00 0.00E+00 3.89E+04 1.00E+00 1.00E+00 1.14E+00 1.04E-01 56088 0 14723 0 3.75E+05 0.00E+00 0.00E+00 1.86E+04 0.00E+00 -3.77E-02 0.00E+00 ITE= 1 -1.06E+01 5.01E+03 0.00E+00 1.69E-02 0.00E+00 3.90E+04 -8.00E-05 2.97E-02 1.48E-02 1.07E-01 1.00E+00 4.74E-05 56103-X 0-F 20983-Y 0-F 3.63E+05 0.00E+00 0.00E+00 ( 1) -2.29E+03 0.00E+00 -2.63E-03 0.00E+00 ITE= 2 9.96E+01 3.66E+04 0.00E+00 1.62E-02 0.00E+00 3.09E+04 7.52E-04 2.18E-01 1.42E-02 8.44E-02 1.00E+00 1.10E+01 56103-Z 0-F 20983-Y 0-F 3.66E+05 0.00E+00 0.00E+00 ( 1) -1.85E+04 0.00E+00 2.48E-03 0.00E+00 ITE= 3 -4.59E+01 1.05E+04 0.00E+00 1.53E-02 0.00E+00 2.98E+04 -3.47E-04 6.23E-02 1.34E-02 8.17E-02 1.00E+00 4.61E-01 56103-Z 0-F 20983-Y 0-F 3.65E+05 0.00E+00 0.00E+00 ( 1) -6.04E+03 0.00E+00 -2.52E-03 0.00E+00 ITE= 4 9.66E+01 3.79E+04 0.00E+00 1.52E-02 0.00E+00 3.04E+04 7.29E-04 2.25E-01 1.33E-02 8.29E-02 1.00E+00 2.23E+00 56103-Z 0-F 20983-Y 0-F 3.67E+05 0.00E+00 0.00E+00 ( 1) -1.91E+04 0.00E+00 2.47E-03 0.00E+00 ITE= 5 -5.01E+01 1.19E+04 0.00E+00 1.52E-02 0.00E+00 3.04E+04 -3.79E-04 7.04E-02 1.33E-02 8.32E-02 1.00E+00 5.19E-01 56103-Z 0-F 20983-Y 0-F 3.65E+05 0.00E+00 0.00E+00 ( 1) -6.83E+03 0.00E+00 -2.46E-03 0.00E+00 ITE= 6 9.22E+01 3.80E+04 0.00E+00 1.51E-02 0.00E+00 3.08E+04 6.97E-04 2.26E-01 1.32E-02 8.40E-02 1.00E+00 1.92E+00 56103-Z 0-F 20983-Y 0-F 3.67E+05 0.00E+00 0.00E+00 ( 1) -1.91E+04 0.00E+00 2.41E-03 0.00E+00 ITE= 7 -5.29E+01 1.28E+04 0.00E+00 1.51E-02 0.00E+00 3.08E+04 -4.00E-04 7.62E-02 1.32E-02 8.44E-02 1.00E+00 5.74E-01 56103-Z 0-F 20983-Y 0-F 3.65E+05 0.00E+00 0.00E+00 ( 1) -7.39E+03 0.00E+00 -2.40E-03 0.00E+00 ITE= 8 8.84E+01 3.80E+04 0.00E+00 1.49E-02 0.00E+00 3.12E+04 6.68E-04 2.26E-01 1.31E-02 8.50E-02 1.00E+00 1.74E+00 56103-Z 0-F 20983-Y 0-F 3.67E+05 0.00E+00 0.00E+00 ( 1) -1.91E+04 0.00E+00 2.37E-03 0.00E+00 ITE= 9 -5.53E+01 1.37E+04 0.00E+00 1.49E-02 0.00E+00 3.11E+04 -4.17E-04 8.12E-02 1.31E-02 8.53E-02 1.00E+00 6.25E-01 56103-Z 0-F 20983-Y 0-F 3.65E+05 0.00E+00 0.00E+00 ( 1) -7.88E+03 0.00E+00 -2.36E-03 0.00E+00 ITE= 10 8.45E+01 3.80E+04 0.00E+00 1.48E-02 0.00E+00 3.15E+04 6.38E-04 2.26E-01 1.30E-02 8.60E-02 1.00E+00 1.60E+00 56103-Z 0-F 18097-Y 0-F 3.67E+05 0.00E+00 0.00E+00 ( 1) -1.91E+04 0.00E+00 2.32E-03 0.00E+00 ITE= 11 -5.75E+01 1.45E+04 0.00E+00 1.48E-02 0.00E+00 3.15E+04 -4.34E-04 8.63E-02 1.30E-02 8.63E-02 1.00E+00 6.80E-01 56103-Z 0-F 18097-Y 0-F 3.65E+05 0.00E+00 0.00E+00 ( 1) -8.37E+03 0.00E+00 -2.31E-03 0.00E+00 ITE= 12 7.98E+01 3.81E+04 0.00E+00 1.46E-02 0.00E+00 3.19E+04 6.03E-04 2.26E-01 1.28E-02 8.71E-02 1.00E+00 1.47E+00 56103-Z 0-F 18097-Y 0-F 3.67E+05 0.00E+00 0.00E+00 ( 1) -1.91E+04 0.00E+00 2.26E-03 0.00E+00 ITE= 13 -6.00E+01 1.55E+04 0.00E+00 1.46E-02 0.00E+00 3.19E+04 -4.53E-04 9.23E-02 1.28E-02 8.74E-02 1.00E+00 7.52E-01 56103-Z 0-F 18097-Y 0-F 3.65E+05 0.00E+00 0.00E+00 ( 1) -8.95E+03 0.00E+00 -2.25E-03 0.00E+00 ITE= 14 7.31E+01 3.81E+04 0.00E+00 1.43E-02 0.00E+00 3.25E+04 5.52E-04 2.26E-01 1.25E-02 8.86E-02 1.00E+00 1.33E+00 56103-Z 0-F 18097-Y 0-F 3.67E+05 0.00E+00 0.00E+00 ( 1) -1.92E+04 0.00E+00 2.19E-03 0.00E+00 ITE= 15 -6.32E+01 1.69E+04 0.00E+00 1.43E-02 0.00E+00 3.24E+04 -4.77E-04 1.01E-01 1.25E-02 8.89E-02 1.00E+00 8.64E-01 56103-Z 0-F 18097-Y 0-F 3.65E+05 0.00E+00 0.00E+00 ( 1) -9.75E+03 0.00E+00 -2.17E-03 0.00E+00 EQUILIBRIUM ITERATION IN TIME STEP = 94 NUMBER OF ITERATIONS = 15 ITERATION LIMIT REACHED WITH NO CONVERGENCE *** S T O P ***
Thanks.
05-15-2017 05:29 AM
1. Has the problem converged/solved without the friction?
2. What is you increase the max allowed iteration?
05-15-2017 08:07 AM
Yes, and yes.
The timestep that I started out with was 3e-5 s for 100 timesteps for a total time of 3e-3 s.
And it will get to around 2e-3 s with that timestep size before it will fail to converge.
But I've even tried lowering the timestep down to 1e-7 s from 3e-5 s and it fails around 3e-5 s which is "odd" given that with a higher time step, it can do it.
I've also enabled ATS now with its default parameters and it still fails to converge, so I am starting to run out of ideas.
Thanks.
05-15-2017 08:46 AM
ATS will only be effective if the loads are time varying, so that the loads can be reduced when the time is reduced. There is no point in subdividing tim if loads are not time dependent.
The case control snippet shown above contains both LOAD and DLOAD, so there are some static (non time varying) loads in the model. Consider making those time varying as well.
05-15-2017 09:48 AM - edited 05-15-2017 09:52 AM
Two of the loads are transient, one isn't. (And that is due to the nature of the underlying physics.)
There's a TABLE(D? I think is the name of the card) towards the bottom of the deck (below the GRIDs, CTETRAs, and contact regions, and I forget what else), but there are two transient loads and one that is a non-time dependent load (two PLOAD4s and one FORCE).
The "steady-state" load is what it is (based on the physics of what else is going out external to the model, that isn't modelled here, e.g. electromagenetics). The emag load has been decomposed into a mechanical load that I can apply using NX NASTRAN (SimCenter).
It is my understanding that NASTRAN (pretty much regardless of flavour or implementation) can't solve a multiphysics problem like that where it couples a steady-state (or transient) emag load with a transient, mechanical load.
But that provides additional background in regards to the loads themselves, but my question concerns more with the contact condition moreso than the load.
(The transient load is modelled with a sine function, so you actually have a progressive ramp-up in load (peak load is at t=1.5e-3 s), and then ramp-down. And the latest run, the non-convergence occurs when the load is being UNloaded. So...I'm not really sure why and I am looking for ideas/suggestions/things to try or things that I can run diagnostically to try and find out what's going on.)
In looking at whatever results are available before the time step where it fails to converge, there is little indication that the elements are getting severely distorted/skewed, but I don't know enough about NX NASTRAN (SimCenter) to be able to check/prove/verify that for certain. I can only "tell" just by "eye balling" it.
Thanks.
05-15-2017 03:02 PM
Huh. Weird - there was a reply that I had written, and then edited, but then it doesn't show up anymore.
Anywhoo - in short - two of the loads are transient in nature (modelled using sine function) and one is a constant (read: "steady-state"/non-time dependent load) due to the nature of the physics.
(The constant force load comes from electromagnetics, which it is my understanding that NASTRAN, regardless of flavour, can't do in conjunction with a nonlinear transient mechanical load, so that has been "decomposed" into a single, non-time dependent mechanical force load.)
I think that there is a TABLE(D?) (I forget the name of the card) that shows up below the GRID, CTETRA, contact regions, etc. that the snippet doesn't show.
Hope that helps.
Thanks.