Error message: Solver Problem, Zero pivot 1

[Low12]

Dear Abaqus specialists,

i am currently working on the FE Analysis of an indentor Test in order to examine the failure mechanisms (cracking and Delamination) of coated Tools. Unfortunately my Simulation Always Aborts giving the following error message:

Solver problem. Zero pivot when processing D.O.F. 1 of 49 nodes. The nodes have been identified in node set WarnNodeSolvProbZeroPiv_1_1_50_1_1.



I have a Picture of the model and the nodes specified in the warning set. These node are exclusively located within the Coating-Substrate interface. I would be glad About any ideas and Inputs. Thank you in Advance.

 

[FEA way]

This message indicates either underconstraint or overconstraint. The main question here is what interactions and BCs (I guess that in this case only interactions) are applied to the nodes identified in the warning set ?

Did you get any other warnings ? Could you share your model’s files here ?

 

[Low12]

Thanks for your answer.

For the coating-substrate contact i used a Node to Surface contact with finite sliding. Do you mean the Data file? The Data file for the interaction section looks like the following:


***NOTE: DUE TO EXCESSIVE REPORTING, THE ECHO OF THE *SURFACE OPTION IS BEING
SUPPRESSED.
*contactpair, interaction=_INTERFACE_CONTACT-PROP, adjust=ASSEMBLY_VERBINDUNGSKNOTEN
*contactpair, interaction=RW_CVD, type=SURFACETOSURFACE
*surfaceinteraction, name=INTERFACE
*surfacebehavior
*surfaceinteraction, name=RW_CVD
*surfacebehavior, pressure-overclosure=HARD
*surfaceinteraction, name=TEST_INTERFACE
*surfaceinteraction, name=XFEM
*surfacebehavior
*surfaceinteraction, name=_INTERFACE_CONTACT-PROP
*surfacebehavior
*contactpair, interaction=_INTERFACE_CONTACT-PROP, adjust=ASSEMBLY_VERBINDUNGSKNOTEN
*contactpair, interaction=RW_CVD, type=SURFACETOSURFACE
*enrichment, name=XFEM, type=PROPAGATIONCRACK, elset=ASSEMBLY__PICKEDSET14, interaction=XFEM
*surfaceinteraction, name=INTERFACE
*surfaceinteraction, name=RW_CVD
*friction, sliptolerance=0.005
*surfacebehavior, pressure-overclosure=HARD
*surfaceinteraction, name=TEST_INTERFACE
*surfaceinteraction, name=XFEM
*surfacebehavior
*fracturecriterion, type=VCCT, normaldirection=MTS, mixedmodebehavior=BK, unstablegrowthtolerance
*surfaceinteraction, name=_INTERFACE_CONTACT-PROP
*cohesivebehavior
*damageinitiation, criterion=MAXS

***WARNING: THE SECOND MAXIMUM SHEAR TRACTION AT FAILURE MUST BE GREATER THAN
ZERO FOR COHESIVE ELEMENTS, SURFACE_BASED COHESIVE BEHAVIOR, OR
XFEM ELEMENTS IN 3D; OTHERWISE, IT WILL BE ASSIGNED THE SAME VALUE
OF THE FIRST MAXIMUM SHEAR TRACTION.
*damageevolution, type=ENERGY, power=1, mixedmodebehavior=BK
*surfacebehavior
*fracturecriterion, type=VCCT, normaldirection=MTS, mixedmodebehavior=BK, unstablegrowthtolerance
*surfaceinteraction, name=INTERFACE
*surfaceinteraction, name=RW_CVD
*surfaceinteraction, name=TEST_INTERFACE
*surfaceinteraction, name=XFEM
*surfaceinteraction, name=_INTERFACE_CONTACT-PROP
*boundary
*boundary
*boundary
*boundary
*boundary
*contactpair, interaction=_INTERFACE_CONTACT-PROP, adjust=ASSEMBLY_VERBINDUNGSKNOTEN
*contactpair, interaction=RW_CVD, type=SURFACETOSURFACE
*enrichment, name=XFEM, type=PROPAGATIONCRACK, elset=ASSEMBLY__PICKEDSET14, interaction=XFEM
*surfaceinteraction, name=INTERFACE

***WARNING: PLEASE CHECK TO MAKE SURE THAT THE INTERACTION PARAMETER IS USED
ON THE *ENRICHMENT OPTION FOR LEFM CRITERION OR LOW CYCLE FATIGUE
ANALYSIS.
*surfaceinteraction, name=RW_CVD

***WARNING: PLEASE CHECK TO MAKE SURE THAT THE INTERACTION PARAMETER IS USED
ON THE *ENRICHMENT OPTION FOR LEFM CRITERION OR LOW CYCLE FATIGUE
ANALYSIS.
*surfaceinteraction, name=TEST_INTERFACE

***WARNING: PLEASE CHECK TO MAKE SURE THAT THE INTERACTION PARAMETER IS USED
ON THE *ENRICHMENT OPTION FOR LEFM CRITERION OR LOW CYCLE FATIGUE
ANALYSIS.
*surfaceinteraction, name=XFEM
*surfaceinteraction, name=_INTERFACE_CONTACT-PROP

***WARNING: PLEASE CHECK TO MAKE SURE THAT THE INTERACTION PARAMETER IS USED
ON THE *ENRICHMENT OPTION FOR LEFM CRITERION OR LOW CYCLE FATIGUE
ANALYSIS.
*contactpair, interaction=_INTERFACE_CONTACT-PROP, adjust=ASSEMBLY_VERBINDUNGSKNOTEN
*contactpair, interaction=RW_CVD, type=SURFACETOSURFACE
*enrichment, name=XFEM, type=PROPAGATIONCRACK, elset=ASSEMBLY__PICKEDSET14, interaction=XFEM
*output, field, frequency=2
*output, history, variable=PRESELECT
*output, field, frequency=2
*output, history, variable=PRESELECT, frequency=10
*elementoutput, directions=YES
*elementoutput, directions=YES
*output, field, frequency=2
*output, history, variable=PRESELECT
*output, field, frequency=2
*output, history, variable=PRESELECT, frequency=10
*elementoutput, directions=YES
*elementoutput, directions=YES


SLAVE : ASSEMBLY_S_SURF-4
MASTER: ASSEMBLY_M_SURF-4
DISTANCE
NODE ADJUSTED REMARKS/WARNINGS

14463 4.00000E-10 ADJUSTED TO ( 0.50000 , 4.00000E-02).
14464 4.00000E-10 ADJUSTED TO ( 0.49800 , 4.00000E-02).
14465 4.00000E-10 ADJUSTED TO ( 0.49600 , 4.00000E-02).
14466 4.00000E-10 ADJUSTED TO ( 0.49400 , 4.00000E-02).
14467 4.00000E-10 ADJUSTED TO ( 0.49200 , 4.00000E-02).
14468 4.00000E-10 ADJUSTED TO ( 0.49000 , 4.00000E-02).
14469 4.00000E-10 ADJUSTED TO ( 0.48800 , 4.00000E-02).
14470 4.00000E-10 ADJUSTED TO ( 0.48600 , 4.00000E-02).
14471 4.00000E-10 ADJUSTED TO ( 0.48400 , 4.00000E-02).
14472 4.00000E-10 ADJUSTED TO ( 0.48200 , 4.00000E-02).
14473 4.00000E-10 ADJUSTED TO ( 0.48000 , 4.00000E-02).
14474 4.00000E-10 ADJUSTED TO ( 0.47800 , 4.00000E-02).
14475 4.00000E-10 ADJUSTED TO ( 0.47600 , 4.00000E-02).
14476 4.00000E-10 ADJUSTED TO ( 0.47400 , 4.00000E-02).
14477 4.00000E-10 ADJUSTED TO ( 0.47200 , 4.00000E-02).
14478 4.00000E-10 ADJUSTED TO ( 0.47000 , 4.00000E-02).
14479 4.00000E-10 ADJUSTED TO ( 0.46800 , 4.00000E-02).
14480 4.00000E-10 ADJUSTED TO ( 0.46600 , 4.00000E-02).
14481 4.00000E-10 ADJUSTED TO ( 0.46400 , 4.00000E-02).
14482 4.00000E-10 ADJUSTED TO ( 0.46200 , 4.00000E-02).

***WARNING: FOR CONTACT PAIR (ASSEMBLY_S_SURF-4-ASSEMBLY_M_SURF-4), NOT ALL
THE NODES THAT HAVE BEEN ADJUSTED WERE PRINTED. SPECIFY
*PREPRINT,CONTACT=YES FOR COMPLETE PRINTOUT.


I hope this will help you. As you can see, other interaction related warnings occur during the simulation.

 

[FEA way]

Since you want to model delamination with through-thickness crack propagation using surface-based cohesive behavior and XFEM, try creating additional small partition of a substrate under the coating layer and assign enrichment to this region as well.

To clarify the damage definition, in this case it should combine approaches used in surface-based cohesive zone modeling and XFEM analyses (with cohesive or LEFM damage):

1) surface-based cohesive behavior and XFEM with cohesive damage:

- material: damage initiation and evolution
- interaction: cohesive behavior, damage

or:

2) surface-based cohesive behavior and XFEM with LEFM damage:

- material: damage initiation
- interaction: cohesive behavior, damage, fracture criterion (type=VCCT)

Make sure that interaction property with fracture criterion is assigned to XFEM crack definition in the second case.

Also try enabling slave adjustment in contact pair definition.

You can perform some quick tests using files provided for the documentation example "Debonding behavior of a double cantilever beam" in Example Problems Guide. Just add XFEM-related features to this model.

 

[Low12]

First of all, thank you very much for the elaborate answer.

Assigning the XFEM to a small portion of the substrate sounds like a good idea, i will try it as soon as i have access again to my work PC. It would'nt help regarding the zero pivot problem though right? Since the simulation aborts before the crack reaches the contact region of coating and substrate.

The 1st approach is quite clear for me and it is the one i was following so far. Here the delamination is solely controlled by the damage and damage evolution in the interaction property module right?

I have some difficulties understanding the 2nd approach though. The crack initiation within the material is clear to me, but i don't understand how the delamination is being carried out here. The VCCT needs a predefined imperfection since it can only model propagation not initiation. So i'm defining the initiation via 'Damage' and after initiation occurs the VCCT takes care of the evolution?

I'm gonna try this out when i'll be at work next time, i am still a bit unsure though what the reason for the Zero Pivot error could be.

Happy holidays!

 

[FEA way]

In the first case (surface-based cohesive behavior and XFEM with cohesive damage) damage intiation and evolution for XFEM crack is specified in material definition. Interaction property with cohesive behavior and damage (initiation + evolution) is defined for delamination as typically done when modeling surface-based cohesive zone.

In the second case (surface-based cohesive behavior and XFEM with LEFM damage) damage initiation for XFEM crack is specified in material but evolution is handled by fracture criterion (type=VCCT) given in the interaction property assigned to XFEM crack. Another interaction property with cohesive behavior and damage (initiation + evolution) handles delamination.

I think that the confusion comes from the fact that you are thinking about modeling delamination with VCCT approach while I'm talking about modeling it using cohesive behavior. Fracture criterion type=VCCT in the second case is used for XFEM damage, not for delamination.