Surface to surface contact between deformable and rigod parts

[manaliac]

Hello everyone,

I have a contact problem between a rigid and a deformable surface.
I am using a kinematic algorithm, and a surface to surface contact. I am putting a normale pressure on the rigid surface so it can squeeze the deformable one, and it works finely for some steps while the pressure amplitude is increasing, but once the pressure is reaching the max, the rigid part began to bounce back with the deformable part a lot of time, so the displacements of the rigid and deformable parts are constantly going from negative to positive values.

I dont know how to explain this weird behavior, and I would like to know how to change the stiffness of the contact first and how to accept some penetration between the two bodies.
Please Help!
M.

 

[corus]

Look at the displaced shape more closely. It may be that the rigid part slips between the nodes of an element, so some penetration is made, before contact is made again and positive displacements occur. You might need automatic tolerances to smooth this out.

 

[manaliac]

Thank you so much for your answer! I think you are right about the displacements of the impactor (rigid), I am really a newbie on Abaqus, so please can you tell me how I can add automatic tolerances to smooth these slips?

Here is a picture of my model, with the pressure applied to the impactor (the rigid part) on the top and on contact with the deformable part(the box).Lastly the box is on contact with some supports too. The model is meshed with shells elements.

 

[corus]

*contact controls, automatic tolerances

 

[manaliac]

Thank you again, but I am using abaqus explicit so I dont have automatic tolerances as contact control options!! Do you know any other options that permit overclosure or penetration?

You will find attached a 3D shot of my model.

Thanks!

 

[corus]

If you're using dynamic explicit then the structures may well bounce about as you apply your pressure loading. Looking at the structure you might expect it to behave as a simply supported beam, with one of the outer supports being redundant under the pressure load as the box section lifts up (unless self wt keeps it down). As such does the free end of the box section vibrate as you hit it, causing the structure to vibrate up and down under the upper rigid surface where you apply the pressure, causing both to move up and down?
As far as the contact control option goes, I have used it in one of my input decks, but I never checked to see if there were any warnings to say it did nothing.

 

[manaliac]

Well, I am using explicit dynamic. I read in one of your many replies, that sometimes it can ben convenient to create a step where a displacement is applied to the first surface, while the second is constrained, to activate the contact mechanisms.

If you can tell me, some details about such step (like the displacement value, the duration of the step) It would be great , I am trying on my own right now, but I am not confident enough that I will find the perfect set ups. The duration of the step where the pressure is applied on the rigid applicator, is 0.08s, with the pressure goes from 0 to 1MPa on 0.05s and stay monotonic till 0.08s.

M.

 

[corus]

Generally for static analyses it is better to first create a 'dummy' step where contact is inititated, but in a dynamic analysis this isn't necessary as the job will run even though contact may not be made and equilibrium isn't reached. If you do want to try it then just create the first step where you apply a fixed displacement to the rigid body sufficient enough to make contact. As you have several bodies then you may need to apply a displacement to the upper rigid surface, and a second to the deformable body, so that one is greater than the other and the deformable body is effectively 'squeezed' by all the rigid surfaces. Don't apply too large displacements though in case you have non-linear materials and they yield. I tend to use a small value of displacement in the order of 0.01mm. As I said though, I don;t think you need to use this method in a dynamic analysis, and where contact is being made in the step.

 

[manaliac]

Yeah I tried to apply a small displacement, but as you said I am working on a dynamic analysis, even though I am trying to have a load and results that tend to a quasi static analysis. Thank you for all of your advices, it was very interesting to learn so much about contact intiation techniques!
Have a good day.

M.

 

[manaliac]

Sorry to bother again, but I added some round fillets to the wedges (supports) to lessen the level stress, I created contacts (penalty) between said round fillets and the bottom surface of the box. The round fillets are added to the existing surfaces of the wedges so the stresses had to be less important, but in resultats it's higher.

I dont understand these results, can you please help find the fluke?

 

[manaliac]

I would like to add, that when I desactivate the contacts between the round fillets and the bottom surface of the bow,the stresses are lower than the standard case(without round fillets).