Discontinuous CSLIP from step to step

[Bayram2014]

Hello,
I am making contact friction analysis with abaqus.
In a first step, I load my structure like this:

*Step, name=Step-2, nlgeom=yes, inc=100000,UNSYMM=YES
*Dynamic,ALPHA=-0.0,GAMMA=0.5,BETA=0.25
0.1,89.41,2e-10,0.50
***************
*Controls, ANALYSIS=DISCONTINUOUS
*CONTACT CONTROLS
STABILIZE
*CONTROLS, PARAMETERS=TIME INCREMENTATION
,,,,,,,20
*CONTROLS, PARAMETERS=FIELD, FIELD=DISPLACEMENT
,1.
****
** BOUNDARY CONDITIONS
*Boundary
B_Push, 1, 1, -0.6544812e-3


In the second step, the BC is unchanged:

*Step, name=Step-3, nlgeom=yes, inc=100000,UNSYMM=YES
*Dynamic,ALPHA=-0.0,GAMMA=0.5,BETA=0.25
1.0e-7,1.0e-4,2e-10,1.0e-6
***************
*Controls, ANALYSIS=DISCONTINUOUS
*CONTACT CONTROLS
STABILIZE
*CONTROLS, PARAMETERS=TIME INCREMENTATION
,,,,,,,20
*CONTROLS, PARAMETERS=FIELD, FIELD=DISPLACEMENT
,1.
****
** BOUNDARY CONDITIONS
*Boundary
B_Push, 1, 1, -0.6544812e-3

However, when I look to the CSLIP1, I find it discontinuous:
the variable is different between the last increment of step-2 and the first of step-3

regards

 

[FEA way]

What do you want to achieve in the simulation using these two steps ? Does the deformed shape of the model look as expected throughout the whole analysis (you can animate it for easier interpretation) ?

In the documentation chapter "Prescribed Conditions —> Boundary Conditions" there’s an explanation regarding prescribed displacements and their behavior in multi-step analyses, including jumps in displacements which are handled differently depending on the solver.

 

[Bayram2014]

I want to see how fracture propagates like in earthquakes. When the first sign of slipping appears (end of step-2) i stop pushing and wait to ses how slipping propagates.

 

[Bayram2014]

the attached plot shows the CSLIP1 at 2 points of the contact surface.
We can easily see that this variable is put to zero at the beginning of each step.
How can I turn it off to have a continuous slipping as if it was computed in one single step ?
thanks,

 

[FEA way]

Can you show how those prescribed displacement BCs look like for all steps ? Do you use amplitudes or otherwise change displacement values ? If you want to achieve continuous motion in subsequent steps then you may have to redefine your BCs in accordance with what is described in the aforementioned chapter of the documentation. Otherwise, if the motion is supposed to be intermittent, you may get peaks like that in CSLIP plots but they should agree with the behavior of the deformed model in subsequent frames.

 

[Bayram2014]

The BC are as I wrote in my first message.
I apply a displacement on a face during Step-2 until some value : -0.6544812e-3
in Step-3 I stop pushing and hold the face at the last displacement : -0.6544812e-3
the problem is not the intermittence of the CSLIP1 variable, but in the fact that it is put back to zero
at each new step. Which is not normal if we want to see how "slipping" propagates.

regards

 

[Bayram2014]

Any answer because I am stuck.
I would like to add the steps:
**
*Step, name=Step-2, nlgeom=yes, inc=100000,UNSYMM=YES
*Dynamic,ALPHA=-0.0,GAMMA=0.5,BETA=0.25
0.1,89.41,2e-10,0.50
***************
*Controls, ANALYSIS=DISCONTINUOUS
*CONTACT CONTROLS
STABILIZE
*CONTROLS, PARAMETERS=TIME INCREMENTATION
,,,,,,,20
*CONTROLS, PARAMETERS=FIELD, FIELD=DISPLACEMENT
,1.
****
** BOUNDARY CONDITIONS
*Boundary
B_Push, 1, 1, -0.6544812e-3
**
*Restart, write, frequency=0
**
....

and

**
*Step, name=Step-3, nlgeom=yes, inc=100000,UNSYMM=YES
*Dynamic,ALPHA=-0.0,GAMMA=0.5,BETA=0.25
1.0e-6,0.01,2e-10,1.0e-6
***************
*Controls, ANALYSIS=DISCONTINUOUS
*CONTACT CONTROLS
STABILIZE
*CONTROLS, PARAMETERS=TIME INCREMENTATION
,,,,,,,20
*CONTROLS, PARAMETERS=FIELD, FIELD=DISPLACEMENT
,1.
****
** BOUNDARY CONDITIONS
*Boundary
B_Push, 1, 1, -0.6544812e-3
**
*Restart, write, frequency=0
**


regards,

 

[FEA way]

I think that we can assume your BCs are correct (since that’s the kind of motion you want to achieve). So we should focus on other settings of your analysis. Usually, such issues are caused by contact definition and/or mesh. However, to be able to say more I would have to see your model. Can you share the input or cae file ? If not, maybe multiple screenshots showing the model and results (contour plots) will be sufficient.

Why do you change the controls in this analysis ? Did you try running it without adjusting those parameters ?