Learning Forum

[Charlie Cresswell]

Hello,

Ive been working on a simple Ansys mechanical simulation for some days now, and after sifting through some similair posts on this forum I've decided to ask for help. 

I'm trying to simulate the instalation and subseqeunt deformation by pressure of a seal (the seal identified by the orange arrow):

My simulation is in two steps:

1. The simulation is initially ran to allow the seal to 'seat':

2. pressure is then applied to the back side of the seal:

The orange arrow here shows that the contact does not connect. 

I have tried the following with no success:

• Multiple (15) substeps on the second step.
• Large deflection on (which makes a mess of the simulation so i turned it off again)
• Trim contact off, on and specified to 3mm
• pinball radius increased from 1mm (no overlap) up to 20mm (HUGE overlap)
• Time step to automatic bisection and predict for impact
• I even tried a third step at one point where i allowed the seal to inflate then turned on the contact status of the contact element (contact step control)
• a 3D version of the model

Note that i applied a contact tool wich showed the contact as near open. The elastomer is not modelled hyperelastically, rather i calculated the bulk modulus and poisson ratio and substitued that into a polyethylene material. Though this shouldnt make much of a difference?   

Im not so sure what im doing wrong here, is anybody able to assist? 

Best regards,

C

[peteroznewman]

Turn on Large Deflection and work through the issues it causes.

A hyperelastic material model will converge more easily than a linear elastic material with a high Poisson’s Ratio.  Learn how to turn on the mixed u-P element formulation using a Keyop when you switch to this model.

A small blend on the tips of the seal may be helpful to avoid element distortion when contact is working and may help the contact algorithm as well.

[Charlie Cresswell]

Thank you Peteroznewman, 

Once i managed to get the simulation to stabalise this worked perfectly. 

I appreciate your help very much. 

Charlie C 

 

 

[Erik Kostson]

Thanks for letting us know - closing this topic as it might be useful to others.

Erik