Discussion Forum

The offset is a mathematical construct where a number is added to (or subtracted from) the geometrical gap. When the boundaries are in contact, the visual impression is that they are at the given offset distance from each other.

I don't believe this has always been the case though has it?

I came across this exact situation today, repeating an analysis type I have done a number of times before (prior to v6 of Comsol) where a rubber gaiter is fitted over a shaft, and the gaiter expands in the model to snap onto the outside diameters of the shaft. However, it seems with v6 this visual anomaly exists as Shyam describes, where the final position seems to show the gaiter not in the intended final position but actually (in my case) expanded too much and gap between the OD of the shaft and the neck of the gaiter.

Thanks for the reply Henrik and Mark.

So Mark Are you saying that this appearance is only occur in V6 , not in earlier version. Actulay I am thinking that in interface fit in visual expersion there must be 0 or negative gap between the bodies came in contact.

Mark,

As far as I know, this has always been the case. Can you show an example highlighting the difference that you experience? If it is confidential, ask support about why there is a difference.

Hi Henrik,

I don't want to de-rail the initial thread, but attached is a quick description of what I have observed with v6. I am struggling to locate a solved model using an earlier version, but I recall the displayed (and calculated) deformation would be as intended, whereas the visual (and calculated) displacement in v6 seems to be almost twice what was asked for.

Of course, I am fully expecting this to be user error (!), but I do not believe I have changed how I approach this type of analysis.

Mark

Hi Mark,

This seems like a potential user error. If there is a geometrical overlap, then no offset is needed (except possibly to get good starting conditions).

In this case, I would use a negative offset, multiplied by (1-para). The negative offset removes the overlap at the first parameter value, so that the contact can be established without problems.

For shrink fit simulations where the geometry is not overlapping in the original configuration, you need a positive offset, and then the visual anomaly appears.

Hi Henrik,

Thanks for that feedback, I will try that approach... not sure if any of this helps with the original subject of the post, but I appreciate you taking the time to look at this related question!

Mark