in I create part modeling as a multibody is not there, therefore it must be reasoned in a different way by the cads that foresee this type of modeling.
If you are really forced to have a single part that encloses the maths of the entire welded assembly, you can do it through the cospy geometry of the advanced assembly, the steps would be:
-I would create a support set
- assemble the axieme from which you want to extrapolate the entire geometry
- insert by default (or with the coordinate system) an empty part and the active ingredients- control "copy geometry" and selections as type of external reference "part" and "only published geometry"
-selections "references" and selections the first component of the set, is asked if the positioning method is the default one, confirm
-go back to the menu "references" and expand it, in the first row the model name appears, move on the collector under "set of surfaces", thus opens the under window with the only selected component previously
- selecting a surface and then with the right button, collect all the surfaces of the component
- if you need points or axes or planes, move to the "referral" picker
- at the end of giving ok
- equally made with the chosen components, open the part with all the geometry copy and with the "solid" command, turn into solid all the geometry copy.
In this way you have a unique external part referenced to the set, which will automatically update to every change of the set.
n.b.
- If you delete a component, you will have to delete the geometry copy associated with it (otherwise it will fail)
- some complex components may not be solidified at the first attempt, it should be analyzed because from this problem and every situation makes a little history to itself, but elaborating a little starting mathematics, it is resolved.
- if you need to process the part of "summary" of the set," I suggest you put all copy geometry in "manual update" (from the right-click menu), so data traffic is less heavy.
there is also the shrinkwrap command, which does the same work in automatic but the greater the axieme and more precision loses and you can not solidify mathematics, even with maximum precision level (it is not studied for the purpose we are analyzing)
personal thought: multibody modeling is often abused (for convenience/mancancy of alternatives etc.) finds its ground for the modeling of extremely complex parts such as plastic parts, where certain geometries are much more complex to handle, as the feature tree stretches, but in other areas it involves a long series of project management limitations and however does not solve the problem of external references when then it is necessary to "spezzare" multibody in single parts.