Simulate workflow for efficiently replacing assembly components with mass idealizations

[JF_6099993]

What is the correct workflow in simulate - assembly to replace a component with a point mass idealization at the part C.O.G. and constrain it to the assembly with links?

When using mass idealization - component at point - the following notification appears: "You have included a mass idealization of a component which contains the highlighted volumes that will be meshed during an analysis. This will result in the mass being accounted for twice in the run."
If I create a simplified representation of the part which removes the solid geometry, the simplified representation removes the component from view entirely (including the cog datum point defined in the part).

My current approach is to open the part/subassembly to idealize, do mass properties, enable X/Y/Z COG features, then create an offset point with relations using those features to define it. Then I return to the assembly simulation and create an assembly datum point at the assy level defined by the COG location point within the part. Then I use refine - create mass - mass by part on the assy level datum point (thats when the notification about double mass pops up). Finally I go back into the part level, create a simplified rep with all solid geom features excluded, then return to the assy simulation and use this simplified rep.

Issues are that the simplified rep disappears at the assy level (perhaps because it no longer has any solid geometry?) It is also a very slow workflow for something that seems like it should be achievable in a few clicks. I have several dozen subassemblies that I must replace with mass representations or simplified reps - and I would like to maintain parametric definitions of the mass properties of the idealized components so that if someone on my team changes a subassy detail I can update the analysis quickly.

Curious if anyone else has come up with a better workflow for creating paramteric idealized mass reps in simulate - assy.

Here are the errors that I faced
You have included a mass idealization of a component which contains the highlighted volumes that will be meshed during an analysis. This will result in the mass being accounted for twice in the run

[Chris3]

Personally, I just remove the components and replace them with a point and a mass node. You can however create inheritance parts or parts with copy geom features and replace your production part with the simplified part. This would keep the parametric nature in place.

There was a PTC/User presentation on this but it looks like it was archived. If I can dig it up I will upload it for you.

[Chris3]

Attached is the presentation. It is old, but the fundamentals still apply.

[JF_6099993]

The fundamentals in there are good - it is interesting to me that inheritance models are used instead of simplified reps. Im curious why one is chosen vs the other. The linked presentation unfortunately does not address idealizations.

[Chris3]

At the time that presentation was put together, part reps were not part of the CAD package. That aside, inheritance models allow you to change features without changing the production part. For instance if you want to change a wall thickness you can make a varied dimension in an inheritance model to see what that change would do.

[JF_6099993]

In a broad sense that is what I am doing, however since I am doing this for dozens of components I am looking for an efficient workflow if one exists since this strikes me as a fairly straightforward analysis task. Often when my workflow feels this unwieldy it is because there is a different and better workflow available to achieve the same thing.

For reasons I think are obvious, I dont want to manually calculate each CG within the part CYS then calculate the equivalent offset in the assy CYS for dozens of parts. Especially since I would have to redo the calculation any time a part changes.

[JF_6099993]

Alternately, is there a way to prevent a simplified rep containing only datum geometry from being excluded from a higher level simplified rep?

[SweetPeasHub]

Hi @JF_6099993 ,

component at point workflow.

1. The component to be used should be already in the excluded state when you pick it (show excluded in your model tree and pick it from there)

2. The point to put the component does not have to be its COG. It will use the parallel axis theorem to relocate and still have the correct mass and inertial terms. I would put this point in a visible part of the assembly either within a part or at the assembly level. I put it sort of close to where the COG is, but it does not need to be exact, nor update to the cog per my previous statement.

Note on the separate issue of automatic cog point. (for some other workflows.)

• To streamline use a Parametric license with behavioral modelling that allows automatic creation of point or csy at cog. This will automatically update to the value at where the measure feature is in the model tree. Moving to the footer is the most common situation but in a rare case you might want to take the mass before some additional features are added.

• if you really want the component at point to follow exactly at the cog, even though that is not necessary...
  • In another simplified rep or the master rep with the component included, make a point at the assembly level that is ON the cog point (from the analysis feature) in the part . This creates a relationship that will update when regenerating, but only if the referenced component is included at the time of regeneration. This means some switching between reps to do an update, but then this new point can be referenced even when the part is excluded because it is in a part or assembly that is not excluded.

*******Bonus hack of CREO. Maybe they don't want you to know this 🙂 ******

• IF: Something is put in the model tree that requires a higher license or feature like behavioral modelling. (Like analysis feature with automatic COG point)
• THEN: This item will still function if the part/assembly is opened with a lower license. (otherwise lots of things might break)
• Two ways to hack...
  • Add the tree items to your start parts so whenever a part is created, this is already in the tree. (need access to the higher license to make the start part)
  • If you have one part with the "advanced" feature in it, you are not prevented (yet) from copy and pasting said feature into another part. You only need one part file with the "advanced" feature to copy from, and then can place it into any other part.
• I have seen this work for this analysis feature that creates the cog point and it works for some mechanism model tree items that only can be created if you have MDO. MDO has items added to the model tree for dynamic MDO that basic mechanism does not allow. It is not very useful if you don't also have need for them to do studies with the higher license, but some of the model setup can be done before grabbing the higher license, which reduces time spent using the more expensive Creo Parametric license.

Welcome to the PTC simulate community!

[JF_6099993]

Am I correctly understanding point 2: The initial location of the point chosen for part at point doesn't matter, once the part is assigned to the point, the point will be relocated?

[SweetPeasHub]

In a sense, yes. The mass is relocated to the correct location. The point is not actually moved. It uses the parallel axis theorem.

We used to do this manually, it is now automatic with this newer component at point mass type.

But if that seems confusing you can just use the automatic point at COG analysis feature I discussed.

From PTC technical support

"...For a mass
element where the attachment point (to the rest of the model) is not
coincident with the center of mass for the element, how should the
inertia terms be entered?

The inertia terms are with respect to the attachment point. Thus, the
user can figure out the inertia terms with respect to the center of
mass, and then use the parallel-axis theorem to compute the inertia
terms with respect to the attachment point:

Mxx = Ixx = Icxx + m * (y^2 + z^2)
Myy = Iyy = Icyy + m * (z^2 + x^2)
Mzz = Izz = Iczz + m * (x^2 + Y^2)
Mxy = Ixy = Icxy - m*x*y
Myz = Iyz = Icyz - m*y*z
Mzx = Izx = Iczx - m*z*x

where:
c = represents mass center C
m = mass of the body
x,y,z = coordinates of the mass center C..."

[JF_6099993]

Hmm this does not match my testing - here is a point off the centroid of the part to be replaced (node is rigidly linked to center of plate):

And here is the same thing but with the mass / centroid at the center of the plate:

This to me says that the location of the point chosen does matter.

[SweetPeasHub]

This is a pretty extreme case of being off from the correct point. Also the mass is not attached in your examples. It is not connected to an element. Rigids do not count as elements. Also, why is the rigid connection showing in the second and not the first image?

[JF_6099993]

• This is a pretty extreme case of being off from the correct point.
  • Chose an extreme case to test
• Also the mass is not attached in your examples. It is not connected to an element. Rigids do not count as elements.
  • Do you have a suggested alternate attachment for the point mass?
• Also, why is the rigid connection showing in the second and not the first image?
  • Just had different things blanked / included for the screenshot to show the test part. The cube is the part which is being replaced but the assigned mass. It is connected in the same way to the same surface as in the second image, only the original location of the adpt is changed.