Discontinuous stick slip open zones

[Bayram2014]

Hi Everyone,
I am trying to reproduce the experimental results of a friction rig.
I have two plates (modelled here as 3d solids) one above the other.
After applying a normal (vertical) pressure (step 1) I start pushing the plate one at the top until the first sliding occurs (step2).
Then in step3, I stop pushing and wait for the fracture to propagate.
I am making implicit dynamic analysis.
I am using the exponential decay friction model in abaqus.
I have been playing with the friction parameters to obtain results comparable to our experimental ones.
With
FRICTION, EXPONENTIAL DECAY
μs, μk, dc

I noticed that with bigger dc, the slipping speed is higher.
I start to decrease dc to have physical speeds. However, when I got the dc value to have
a reasonable fracture propagation speed for the first increment, the propagation mode changes: In fact after a continuous
propagation of the slipping zone, I start to see disconnected small surfaces slipping, sticking or open.
We have not yet observed this phenomena experimentally.

A picture attached here shows the discontinuous zones of stick/slip.

Thank you for any help.

Bayram

 

[FEA way]

What are the contact settings (apart from the friction model) ? Is the pressure from thr first step propagate to the remaining two steps ? Which application type is used for dynamic implicit steps ?

 

[Bayram2014]

Of course the pressure and gravity are both propagated until the end of the analysis.
For the application I have used all of the available possibilities: Quasi-static, moderate dissipation, transient fidelity. I have even tried
with entering alpha, beta and gamma myself to have different numerical dissipation in the HHT method. The result is always the (almost) same and
the phenomena is observed.

regards

 

[FEA way]

What about the contact settings ? Which type of contact is that (pairs or general, node to surface or surface to surface and so on) ?

 

[Bayram2014]

** Interaction: Int-1
*Contact Pair, interaction=IntProp-1, type=SURFACE TO SURFACE
Part-B-1.B_Contact,Part-H-1.H_contact,
** ----

 

[Bayram2014]

I have tried also Node to surface contact and obtained similar results.

 

[NRP99]

Check "Frictional behavior" in help manual. Could you use test data points directly instead of coefficients which might help?

 

[Bayram2014]

Exponential decay or test data are the same. In the second option, you do not specify dc, that's it.

 

[Bayram2014]

Hello,
Still stuck on this intermittent slipping.
Thank you for any help