I am having some trouble with determining the mass of an assembly that I am working on in Creo 2.0. What I did was created a full assembly and then inserted a cross-section at a plane at the center of my assembly. I would like to know if there is a way to determine the mass of the assembly on just the left side of this plane or on the right side of the plane. I am able to use the cross-section to view half of my part, but only able to determine the mass of the entire assembly. Any help is appreciated. Thanks!
Cross section is a visualisation tool only, mass properties will not reflect it. If you want true MP update you need to perform assembly level cut on the top assembly. You can use same datum plane for cut purpoce.
Well, if you have the cg calculated automatically, and with a CS placed there, you can create left and right assembly cuts, and create family table instances using these and the CS should update for each side. If you want CS's in the non-cut assembly, you'd have to measure the CS's distance/rotation from, say, the default and then manually place them (they won't update).
Not entirely visualization** - cross-sections create hidden family tables for all the parts that are intersected, and create separate in-memory models for them. It's software slight-of-hand that keeps the user from seeing these. It's the same method that an assembly-level cut produces, but with the distinction that the assembly-level cut hidden-family table models are included in mass property calculations.
I keep hoping for a memory monitor that would display objects PTC manages so the effect of modeling technique would be more apparent in terms of memory usage. Instead of the jet plane cockpit where the critical systems are monitored, we get one idiot light.
**or so I've seen from occasional failures where the seams in the software show. At one time every assy level cut modified each of the intersected parts by adding a hidden family table. This was before the files were encrypted and one could read the various entries, so it was very noticable. I was then told by one PTC tech rep that that was the only possible way to do things and that the cut information could never be part of the assembly file.
Fortunately, hidden instances mechanism is not used in cross sections creation. If it was you'd see huge memory increase upon simple frontal sectioning - same that would happen if you make central plane cut in large assembly. This mechanism is used in few more places besides assembly cut, but not for XSEC. And the fact that you noticed - that assembly level cut is reflected in mass properties while XSEC not - is exactly since XSEC does not really create any new geometry (even not a curve - it should be created as a feature, separately). XSEC is just a visual clipping.
Still puzzling over this - to make a section the software needs to trim intersected surfaces and generate capping surfaces, while eliminating surfaces that are to be removed from view. How is this different than creating a cut to the solid except for not building an independent solids component structure? Since only the affected surfaces and a list are involved what would cause a huge memory increase? I don't even know what visual clipping is WRT computer graphics**. Math gets done somewhere - do you mean it's dynamically offloaded to the GPU for surface subdivision and visibility culling? If so, why does Pro/E have to regenerate for a flipped viewing direction; wouldn't it be instantaneous if the GPU was doing the work? One normal vector change and done? I've seen dynamic sectioning, but that isn't adding a section to the model.
**I'm not new to the subject. That's how I own a copy of the Renderman 1.0 Spec I purchased from Pixar and somewhere have a copy of the FORTRAN card image input reference for MAGI's Synthavision - the software used to render images for the first TRON movie. If only PTC was as forthcoming about the underpinnings so that users could reliably budget projects ... I do keep up, including the t-spline technology that AutoDesk ended up with.
Well, I do not know all the details how this differ internally, just know these are 2 completely independent mechanisms. View clipping (same as XSEC) does not create real solid geometry in terms full calculation of intersection of features, that depends on accuracy, short edges etc.
My guess is that graphical method is way faster and less accurate, tolerable to geom checks, and definitely does not create duplicates in session - though I agree that task of proper shading of the shape is somewhat visually close to real intersection.