What can bodies do that components cannot?

[Noah_Katz]

I've done a lot of reading and watched a lot of videos on bodies/components, which is a confusing

Most of the content is about the pitfalls of not using components, and as far as I can tell, anything you can do anything with components that you can do with bodies.

So why do bodies exist?

One of the videos had me momentarily thinking there was a reason, i.e. you might want to shell just one of the bodies of what would later become a multi-body component, but I believe you can do the same with components.

As a test I created a cylinder and a box as new components and was able to combine them, so I'm puzzled why it says "Unfortunately the combine command is going to limit you to bodies."

here   https://forums.autodesk.com/t5/design-validate-document/combine-components-instead-of-bodies/m-p/674...

[etfrench]

If I'm not doing a join or a cut operation, then 99.999% of the time I want a new component.  It would be a lot easier on my carpal tunnel to have new component as the default.  It's also one of the operations that would be easy to expose in the preferences.

[Oceanconcepts]

It could be in preferences. It seems odd to me, though, that if you were doing an extrude that you would want the resultant body to be in a different component than the sketch that it was created from. I'm maybe not understanding your workflow. For me the normal pattern if I want a new component is Create Component > Create Sketch > Create Body. 

[etfrench]

In direct modeling mode there isn't any compelling reason to create a new component for each sketch. 

[michallach81]

Hello again,

Guys ( @TrippyLighting, @Oceanconcepts, @Beyondforce, @jeff_strater) I think Noah ( @Noah_Katz) is right, we didn't answer his question. I think he understands our points, and at least to some extend know how Fusion is built.

Noah, direct answer to your question is: there are no special advantages of having this "component+bodies+single environment" build that Fusion is utilizing, over other ways to define designs in cad. Yes, you are right all properties could be assigned directly to bodies (geometry), without creating special containers. We also can do motion studies just with bodies only. That's the way Onshape do assemblies, it allows you to mate just bodies. Rhino 3D, for example, organize data through an idea of layers. A nice example where there are no components but everything is assigned to bodies is Ironcad.

Recently I had a chance to play with Ironcad, and this is what I've made through weekend (some time ago);

What was very new to me was exactly what we talking here about, the way Ironcad stores and assign properties. Sketches were not only assigned directly to bodies but were part of a feature. If for example, I was trying to create some "circular" part, instead of creating sketch profile and then using revolve feature, I had to start by enabling revolve feature and within that feature, I was sketching profile especially for that feature. Do I find that concept as a better one than one from Fusion? No, it was very unintuitive and very laborious. Does it mean I would call it bad? No, I know that there were methods to make all experience smooth, it would require an understanding of Ironcad "philosophy". I would have to go through some tutorials and lectures.

That's the main point, there are no special advantages of having this particular build in Fusion, but the decision which way we should define an environment for design is arbitrary. There are no good or bad choices since with such complex task we can't predict the outcome. We can't evaluate workflows robustness. It's all matter of taste. Therefore, the way Fusion looks today is a result of long development. Cad, in general, has some heritage. Fusion for that matter started as a plugin for Inventor (which is build in a very "classical" way, at least classical for parametric solid modelers), and after years of development looks as it looks now.

The last thing that bugs me is your rhetorical form of expression. What is the purpose of it? Since my question is also rhetorical, I will answer it. You don't like how Fusion works, it doesn't suit your habits. You don't want to learn, and you think that since there are no reasons (advantages) for this particular solution it should be different. Your wish is to have " just be parts (to which belong associated coordinate systems, sketches, etc., and which can exist on their own and within assemblies)". On this forum when we have an idea for new functionality we finding good arguments, and try to find allies on IdeaStation. We don't try to prove that through whole this time developers were wrong in their decision making. But fear not you're not the only one who likes to win the biggest (_*_) award. If it's any different from what I said, please give us some good arguments for another solution.

[Noah_Katz]

Ron,

Thanks, that's very helpful.

Going through the Assembly lessons, this time around a lot more of the bodies/components section sunk in.

[Noah_Katz · Accepted solution]

Michal,

>  there are no special advantages of having this "component+bodies+single environment" build that Fusion is utilizing, over other ways to define designs in cad. Yes, you are right all properties could be assigned directly to bodies (geometry), without creating special containers. We also can do motion studies just with bodies only.

Thanks for the direct answer, finally.

So yes, in that respect I am being critical, as I don't like things being more complex and obtuse than they need to be without conferring commensurate benefits.

That said, it could be a lot worse; I expect that at some point I will have a comfortable grasp on the concept and it will cease being an impediment, which is a lot better than being stuck with, say, a common operation that is inefficient and tedious because it requires a lot more mouse clicks than it ought to.

Also let me add some perspective on my overall experience with F360 - I'm *LOVING* it.

Every few minutes as I go through the tutorials, I'm amazed and delighted at the slick, thoughtful, and labor-saving features compared to what I had to do in I-DEAS to accomplish something, as well as outright advancements, i.e. things F360 can do that I-DEAS couldn't with any amount of effort.

Re IronCAD,

> What was very new to me was exactly what we talking here about, the way Ironcad stores and assign properties. Sketches were not only assigned directly to bodies but were part of a feature. If for example, I was trying to create some "circular" part, instead of creating sketch profile and then using revolve feature, I had to start by enabling revolve feature and within that feature, I was sketching profile especially for that feature. Do I find that concept as a better one than one from Fusion? No, it was very unintuitive and very laborious.

Yes, you can create features with sketches, but I believe they intend for features to be built up by dragging and dropping in shapes from the Shapes panel, which I found very slick.

As one of their resellers (tecnetinc) says: "I want to introduce you to a proven alternative to the standard sketch, sketch, constrain, constrain procedure of design. IRONCAD uses drag and drop functionality."

I didn't get to spend as much time with the trial as I wanted before it expired, and their tutorials were minimal, so an unanswered question I have is how "smart" are features created that way, and what happens when you want to modify them.

[michallach81]

Fair enough. Maybe I'm too sensitive to trolling. I wish you the best. I think that after some time you will like that "matryoshka" concept, it's very clear, and makes Fusion very flexible.

Cheers

[Noah_Katz]

Oops, forgot to "accept as solution".

In case that locks the thread, I wanted to say before I do that I do like the fact that in the browser each component contains its associated sketch, origin, joints, etc.