Mass assignments for dynamic mechanism analysis in complex/unfinished assemblies

[cwenner-2]

MDO requires "every part" has mass assigned... and then collects those into bodies. But I want to perform some dynamics predictions of large complex assemblies that are unfinished AND... contain many STEP objects that don't necessarily have masses assigned. Essentially, I want to ASSIGN mass properties to BODIES so that MDO ignores the incomplete part masses. Is my interpretation of documentation correct that this is NOT possible? 

As a second option, my bodies are almost always constructed as stand-alone assemblies! So, I'd like to think I can ASSIGN mass of these assemblies and expect MDO to then properly ignore all information from subparts. However, I'm pretty sure I tried this a few years ago and MDO did NOT use the assigned masses as expected: It continued to compute the masses of subcomponents that had masses assigned. Or maybe just did that for 3D skeletons Anyway, it did not work and I spent a lot of time trying to get it to work.

Does anyone have advice on how to handle this situation. Or, conversely, can some expert simply say definitively that what I want to do is not possible with the tool (ergo, I must construct an analysis assembly instead of using our actual design models).

Thanks.

[skunks]

"... Or, conversely, can some expert simply say definitively that what I want to do is not possible with the tool (ergo, I must construct an analysis assembly instead of using our actual design models). ..."

Yes

[cwenner-2]

I appreciate the response but don't want to assume it correct without having some idea of your back ground with the tool.  Are you a PTC developer? Or, are you someone that has lots of experience using the tool for dynamic analyses? Your answer would be appreciated. Thanks.

[SweetPeasHub · Accepted solution]

I am a user since 1996 with a good amount of experience with mechanisms.  I am glad you are not assuming things about the ptc community.  It is best to prove things for yourself, as Creo is an extensive tool and used in ways that even the developers had not expected nor intended.  I realize that takes time and can be frustrating when, like in this case we think it should work a certain logical way and it does not.  I have not tried to use Assign mass for a while and it would be great if a developer confirmed that the assignment does not ever cross into mechanism and what they view the proper work around to be.  Like skunks attests to, it is often better to have a separately put together assembly that is just for analysis.  Also, those simplified reps can be tricky since you have to remember to set them and save them as needed.  On retrieval of an assembly you can choose "last stored" rep and I find that removes ambiguity about where I was last, in case I did not save the rep changes.  BTW this "last-stored" works for FEA - simulate study directory files as well - to know what exactly the rep. state was when it was run.

[cwenner-2]

Thanks.

[SweetPeasHub]

I would probably create an "analysis assembly" in the same assembly by just using simplified reps, possibly having dummy parts with the correct mass in each of the subassemblies.  This is done by making a small cube at the center of mass and assigning material.  In mechanism, you can override the density until the mass is correct, or to do studies on mass.  This is not perfect because it omits inertia properties, but even that could be done with some multiple cubes or some simplified geometry to create the inertia terms.  Maybe? you can still have the graphics for the non-mass or incorrect mass step objects by setting the representation to Graphics, rather than Exclude?  Also, as long as the body has a mass, it is ok to include non-solid non-mass components as part of the body.  In short, using simplified reps is the general path I would take to accomplish this.