Crazy results!

[HalMatt]

Hi all,

This is first post here so hello again to you all!

I'm a student currently on placement, I been using online tutorials to teach myself the basic skills of Mechanica, with which the company I work for hopes I'll be able run FEA testing on designs they're making. I am struggling with one particular model and assembly at the moment in which a metal plate is bolted to a plastic housing, with an internal bursting pressure applied. When I simulate the test on the plate alone, it shows a masximum displacement of around 1mm, which is roughly what's expected. When I run it with the plastic housing as well in an assembly the whole thing distorts and offers a max displacement of 5 METERS! I get these error in the interactive diagnostic box:

** Warning: One or more contacts in the model will
have no effect because it is not possible for
any of the elements on one side of the contact
region to contact any of the elements on the
other side. See the log file for a list of
contacts for which this applies.
(I assume this one is because part of the plastic surface extends past the plate, leaving an area on that surface that is not in contact with the model?)

** Warning: Poor tangential constraint detected during
contact analysis. Contacts are
frictionless and constraint must be provided
tangential to the contact.

** Warning: One or more contacts in the model will
have no effect because it is not possible for
any of the elements on one side of the contact
region to contact any of the elements on the
other side. See the log file for a list of
contacts for which this applies.


As it currently stands my knowledge of technical terms and Mechanica is patchy, so please go lightly with any help and explainations!

Thanks for any help

 

[shaun8567]

Sounds like your model is insufficiently constrained. To illustrate my point, image your hands and pushed them together, palm to palm with your fingers facing away from you. Now try to slide one hand towards you, and the other away from you. You can't, unless you apply a large enough load, because the normal force between your hand creates a frictional force of some value (depending on what the coefficient of friction is). This frictional force is tangential force. Contact in Mechanica will transfer the normal forces in your model, but it will not transfer the tangential (or frictional) forces unless you specify a coefficient of friction. Even in that case, it will only tell you whether or not it slips.

To see if this is the problem, change all of your contact interfaces to bonded and re-run the model. If it runs successfully, then you need to figure out which interface have their surfaces sliding with respect to one another.

 

[HalMatt]

Shaun,

Thanks for the advice, I've tried running the analysis with the interface set to bonded and it does seem to show a much more 'expected' outcome. I'm back at the drawing board now considering my constraints!

Another question that has come up from a colleague is- is there a way to run some kind of sensitivity study that will consider the location of bolt holes in the material? For example he is proposing to move on of the holes in the cover and would like to know how much effect it would have? I know that I could alter the model and rerun it with the hole moved, but I'm looking for some kind of linear result that shows the change in stress as the hole is moved from A to B. Thanks again!

 

[HalMatt]

Also, to add to this... I've been running another analysis on a symmetrical part. It's simple enough that I don't need to chop it in half and use mirror constraints to run an analysis on it. But if I do the results for max VM and displacement change? I've just used a cut to remove half the model, then just loaded and constrained the rest of the model in the same way as before, except with an added symmetry constraint down what would be the centre of the part. Could anyone explain what's causing this and which is the correct method? Thanks