Learning Forum

[RFV1]

So an overview. I'm attempting to estimate the stresses that would occur in a set of toe clamps and the related bolts that are clamping the bottom flange of a tank to an aluminum baseplate were the assembly to suddenly decelerate at 0.8g initially followed by 1.6g (twice the required load). Loads are applied as forces at the top of a cylinder coincident with the tank's center of gravity. The plate is fixed constrained on a portion of its bottom surface. Further to this I am intending to run 3 versions of this analysis. coefficient of friction is 0.4

1. All contacts are frictionless
2. Contacts between the plate and the bottom flange are considered frictional but clamps are considered frictionless (Gravity being the main driver here)
3. Contacts between the clamps and the plate, the clamps and the flange, and the flange and the plate are Frictional

The model is depicted below

 

Load Step 1 applies gravity force, step 2 applies bolt preload, Step 3 lets the system settle, Step 4 applies the 0.8g force, and step 4 brings applied force up to 1.6g

Image of where loads are applied.

Image of the mesh

Symmetry is applied across the Y-Axis, Materials are 304 Stainless and Aluminum, Frictionless contact is also applied between the bolt's outer surface and the hole wall in the toe clamps to try and limit the bolt from deforming through the hole wall. Nut and bolt head have a rough contact to the surfaces they bear against. The Nut is bonded to the bolts

Analysis settings have Large Deflection turned on, and I have tested runs with both program controlled newton raphson and unsymmetric. 

The analysis manages to converge a few timesteps and then fails. Looking at the stress and deformation plots of the converged steps I notice this somewhat comedic-looking result

The moment I start the simulation it notifies me of an abrupt contact change as the nut on one of the clamps shoots to infinity. I've been looking over the contacts and forces to see if there's something I missed but have so far been unsuccessful. I get no warnings of an underconstrained model. Beginning to suspect maybe I have a geometry error.

Any insight into what might be going on and how to remedy this? Let me know if any additional information is required.

 

[Ashish Khemka]

Hi,

One suggestion is try to load and lock the bolt first and then apply gravity.

Regards,

Ashish Khemka

[RFV1]

I discovered my issue late yesterday actually. I had set the inital step for the bolt as "open" while gravity was being applied which my intuition thought meant something similar to "inactive". That is, I assumed "open" would be effectively leaving the bolt alone until the status was set to load at which point it would start applying the forces to the bolt body. I was sorely mistaken as "Open" is used to simulate loosening a bolt (thus likely why they are shooting to infinity as there is no resistance to the opening effect.)

That said convergence is still difficult. But I'm suspecting that may be more to do with how my geometry is set up or changes in contact.