Cast - Machine - FEA Models - Best Practice

[stu-aspinall]

All

Im more of a cad modeller/draughting rather than an FES specialist, but was just wondering if anyone had any ideas of best practices for creating models which are linked for casting to machining to FEA in Creo Simulate.

Our current procedure is to model the casting model fully (with all its rads etc).  This is then used in the machining model as the first feature, using the merge/inheritance option (merge is our default so users cant change any features).  We then apply our main machining feature to create the machining model,  At the end of this process we will then save the model (Using Save As) to an FEA Model so we keep the link to the casting model.  In the machining model, we will then add further machining features such as weld preps, if needed.  I have since been in touch with some more experienced users who have suggested adding a copy geometry feature into the machining model, then publishing this out to the FEA model, and thus we can drag the copy geometry feature down the model tree and this affect the FEA Model.

But this will not take any fillet rads out of the casting model, if this model has rads which the FEA guys would like to remove for their FEA models

Just wondered if anyone had any ideas

Thanks for your time

Stu

[tbraxton]

Create layers for features that would be suppressed so they are not included in the FEA geometry. You can then use select by layer functionality to toggle the regen  status in the model . This makes it very quick to suppress the features and export the desired geometry for analysis.

With layers you can set up the degree of control desired. I typically have layers for rounds, chamfers, etc. that can be ignored for FEA.

This does presume that the models are built such that suppressing the "FEA" features does not cause other geometry to be suppressed unintentionally.

[stu-aspinall]

Thanks for the reply

We have thought of using the casting model where the "important" rads and chamfers are added into the model, we then use the copy geom and publish geomety to output at that required makeup of the model to create an FEA model.  In the casting model we can then progress and add all the "unimportant" rads, ie sharp edge removal rads etc.

This fully feature cast model could then be used as the basis for the machining model by using the casting as the first merged feature, then add the required machining.  Any machining required for the FEA, could then be copied, using copy surfaces, from the machining model and copied into the FEA, creating a link.

Do you think this works ok ?

Your process of using layers sound like a workable idea, I'm just wary of our users and their knowledge of layers etc

Stu

[tbraxton]

I am a proponent of using merge/inheritance to represent the as cast and machined versions of the part. I use inheritance features for castings but I understand why you may want to use merge instead as you mentioned.

Using the merge or a copy geom as the starting point for an FEA part will work and I see no issue with that.  If this base feature at the start of the FEA model and the geometry needs to be defeatured then I would consider flexible modeling functionality to remove rounds/chamfers etc. These flex features will then reside fully within the FEA geometry part and be controlled within the context of the FEA model.

There are a lot of options to tailor this to your workflow. It is good that you are thinking about it before implementing a standard and I would suggest you create some simple test case models and evaluate them within the team.