Interesting question. Certainly there are multiple ways to accomplish the task in SE, as mentioned. I tend to determine the method of modeling less from what's the quickest / snazziest way to get it into the computer, but more with the mind set of how do I capture the actual device/component/widget’s purpose and intent and then how will these models and associated draft file(s) be maintained throughout their life. I am somewhat envious of those who model something up once, send it off and never see it again. We model things the way manufacturing puts them together, or how they are purchased and then continue to discover how we should have done it over the next 10 years. So much more to this than just “how do I get it into the digital world” for some. But, we’re all entitled to our opinions.
What makes this interesting for me is how it relates to a recent class/seminar I attended on “Additive Manufacturing” or 3D printing… Many of the processes are inherently capable of changing the material properties on the fly. SLA for example, I’m told can change the wavelength and or power setting to change how the resin is cured and that can be changed inside the part. Binder Jetting processes can use multiple “print” heads with different binders, so changing the binder throughout the part can produce different material properties and colors. Directed Energy Deposition (laser or electron beam) use filler wire or focused spray of powdered metal to build the part. It is not uncommon in the existing machines to have multiple heads, as I understood it was for speed, but they are putting different alloys in the different heads and can vary the finished metal alloy on the fly by varying the flow in the heads and power in the beam. All of this is in it’s infancy, and as you know nearly all of the systems use the antiquated STL file format. There is a new format created by a consortium, the AMF format, fulfills many of the needs of information conveyance from modeling environments to the 3D printers. Solid Edge is definitely interested in additive manufacturing, as they have added features for that process. Since they already have multiple bodies (non-manifold even, which is a niche ability of additive manufacturing to produce where machining or molding could not) it only makes sense to continue forward with some way to vary material within a part to allow use of this evolving manufacturing capability.
I'd assume at a certain point there has to be limiting factor of tools available in each enviroment? At what point does it make sense to be able to create two exact assemblies (one in multi body other assembly) and have to keep updating those enviroment? I like to have my options too, but at a certain point one enviroment has to be different than the other otherwise why have two?
I see two points of view regading that issue.
First, it sounds great to have multiple materials for multpile bodies within one and the same part file.
This - of course - can make the developement easier and more flexible.
But - and this is the second point of view to that - is this method of working conform to the designing and development process as we have defined it and used it for decades?
Is a part with more than one material, with different sheet thickness, and this thena is an absolute must, with different document/article/serial numbers, will this part be lonegr a "part"?
Or should we consider taht as an assembly what it really would be then?
And this might be the real question, shall we change logistic rules, numbering systems, etc, to make CAD modelling easier?
I - and this will be true for me only - will see the opportunity to define the geometry in a single master part file with multiple bodies, publishing them and then define document number, material etc. as more conform to my designing process as the "advantage" of several materials within one part.
So to say, the request for more materials in multi body will not get the number one in my enhancements list.
But if there are no other wishes and requests and everything else in SE is working absolutely correct, then ...