Understanding Components

[SGoldthwaite]

I'm trying to get a better understanding of components.  I downloaded a Solidworks model of a desktop CNC router from GrabCAD (link to my Fusion model is below).  Every part in the model is it's own component, this makes the brower tree very long and cumbersome.  One problem I ran into was I tried to move a bolt, and every bolt in the model moved along with it.  I read a post in this forum that mentioned it's easy to move a body within a component when you actually want to move the whole component. Apparently this is what I was doing.  The solution is to change the select filter to not select bodies.  I tried this and was able to move just the one bolt.  This is somewhat confusing and certainly not intuitive. Another issue I ran into which I think is related but I really don't understand is: I grouped a bunch of components together using a Rigid joint, then I moved them.  That worked fine, but sometimes when I'm working on the model (I forget what tool I used), the whole group of parts moved back to the original position, then when I'm done with the edit, the group moves back to where I wanted.  Any idea what it's doing?  Should I have originally moved the parts using a specific method.

I would appreciate any tips on cleaning up my broswer tree with all the components? There a bunch of washers and bolts that make it hard to use.  If I was building a desktop CNC from scratch (which I plan to do), what's a good way to think about organizing components and bodies? 

It there a way to highlight or select all the components that are the same? For example, how could I select all the bolts that are same component?   Also, if I want to make just one bolt shorter, how do I "unlink" it so the other bolts of the same component are not changed?

I've seem some Fusion 360 YouTube tutorials on components, but they are not very in-depth.  Are there any good resources that really go into this?

Link to the CNC desktop router:

https://myhub.autodesk360.com/ue298ce311/shares/public/SHabee1QT1a327cf2b7ae29e00bdc08db885

[Anonymous · Accepted solution]

Hi Scott, here are my thoughts on your questions.

During a move operation, you can also choose between selecting components or bodies in the Move tool dialog window. 

When you edit an existing sketch by actually going into it (i.e. will have Finish Sketch button appear in upper right) the timeline will automatically scrub back to the earlier position on the timeline where that sketch was made.  Hence anything that was created after that sketch in the timeline will temporarily "not exist," which includes Snapshots in the timeline (which establish a new position/orientation for a component coordinate system, and body moves (indicated by 4 point arrow icon in timeline).  Hence while doing in sketch editing, goes to a time before these component or body moves occurred, so will be in prior positions. 

As to why timeline auto scrubs back when in-sketch editing (while can drag existing points around and change sketch dimensions in out-of-sketch editing without timeline autorolling back), is that while in in-sketch editing, projections from other geometry can be referenced to, and could potentially get a "time loop" if referenced to a geometry created after (the sketch currently being edited) yet also referenced to the current sketch.  So needs to be this concept of linear time flow in the timeline. 

For a workflow to prevent things jumping to prior positions upon editing a sketch in a timeline, here are some guidelines I try to follow (although if a model is imported from say Solidworks obviously can't follow this), that I recently posted in another thread:

1) Define a new sketch on existing geometry whenever parametric updating of position etc. is desired from changes of the existing geometry that defined to, and also to prevent stuff "jumping around" from edits in different timeline positions.  But get in habit of first always creating and activating a new component, so that the sketch and subsequent body(s) will go into its respective component.  This allows for among other things nice timeline condensing based on which component is activated, and also the sketches also moved by a component move.  It really helps to turn on Inspect>Color Cycling for ease of component identification. Alternatively, uniquely naming the components helps with identification, even just front, front-right, etc. corresponding to view cube.  

2) For duplicate parts, create copy(s) of the original component such that they are all linked (i.e. one changes they all change).  The pasted component(s)'s positions can be defined to existing geometry via rigid joints.  It is true that these subsequent pasted components will revert position or outright disappear when editing prior timeline features, but the original component will stay in place at least.  

3) For performing fillets, shelling, etc. of the multiple components as one (or just reducing components as described below), they can indeed be combined first.  If there are two or more paste linked components, it is important to choose these as tools for combining, with an unlinked component being the target, otherwise will get weird behavior (multiples of the combined component).  The resulting combined body will go into the target component.  The components now potentially left empty of bodies can be soft deleted if desired for browser simplification (although corresponding joints will disappear, they are still there/functional, and will reappear if corresponding component(s) brought back into browser via deleting the deletion event(s) in the timeline).  If need to modify a sketch of one of these soft deleted components, can simply scrub back timeline to before the deletion events.  Allowing sketch relocations between components would be nice alternative.  But a nice side benefit of this timeline scrubbing of deleted "body barren components" is that the scrubbing easily turns on or off visibility of all those sketches in different soft deleted components at once.

Besides merging components described above, you can drag a component into another component in the browser to make the former a subcomponent of the latter. 

Selecting a component in the browser highlights blue the corresponding components in the work space.  I also like using the component color differentiation mentioned above.  This is especially good for multiple copy-pasted components that are linked -- they will all show up the same color.  Also identified as being linked via having the same number after the the colon ":" in the browser, so to highlight all linked components could hold shift down and select in browser.  As to breaking the link for one of these copy-pasted components, as far as I know, need to copy the component instance of interest, click on the very top level model component, right click and select Copy New, click ok if satisfied with current placement, then delete the other instance.

Jesse

[jeff_strater]

Hi Scott, and thanks for the post.  A lot in here.  I probably won't answer all of your questions, but I'll take a shot at some of them.  Maybe others can chime in as well.

First, to understand bodies vs. components, start with this help article:  bodies and components.  IMO, it gives a pretty good summary about why the two concepts exist in Fusion.  We realize that this can be confusing at times.  A bit of history, for your amusment, on this topic:   During the development of Fusion, we did try to eliminate the multiple concepts at one point.  We decided that we could dispense with bodies, and each independent chunk of material would create a new component.  It sounded like a good idea at the time, but it ended up to be a bit of a disaster, so we tossed that idea out.  It caused too many components to be created (cut a component into two, and you get a new component!).

So, for better or worse, we will continue to have bodies and components.

Re:  Long browser.  Yes, that model does have a very long browser.  But, that just reflects the structure of the data that you imported from GrabCAD.  The best way to control clutter in the browser is by making sub assemblies.  In this particular case, because the data was imported, I would advocate doing that just after the import, while the model is a Direct model.  You can do it in Parametric models, but it generates a lot of "restructure" features in the timeline.  I tried this with your model:  I added all of these components:

Into a single subassembly:

I will put together a screencast to show how I did that, as soon as I am done with this post.  There are a few tricks that make it easier.  That said, we could always make it easier.

But, in general, for designs you create on your own, planning out the design hierarchy as you go will go a long way.  Creating sub assemblies is the key.

Regarding move:  Yes, others have hit that problem.  You think you are moving a component, and you are really moving a body.  The key is the "Object" setting in the Move command, and knowing what it is you want to do.  Sometimes you actually do want to move a body...  More history:  At one point, we had two separate commands: "Move Component" and "Move" (for everything else).  There was a desire to avoid having too many commands (we didn't want a toolbar that looks like Rhino...), so we combined them.  The jury is still out as to whether that was a good move or not (personally I liked "Move Components").

Hope this helps clear up some of your questions.

Jeff Strater (Fusion development)

[jeff_strater · Accepted solution]

OK, here is a screencast on how to create a sub assembly from existing components:

 Also, I realize that I missed a couple of your questions:

1. Selecting all instances of a component.  Sadly, no, there is no way to do that today.  Actually, though, that's a good idea.  You should put that on the Idea Station.

2. Movement during edit.  Jesse answered this (thanks, Jesse!), but I can provide some more context.  This one is sometimes tricky for people to get, because Fusion works differently than some CAD tools.  In Fusion, the assembly itself is history-based (not just parts, as is the case in Inventor and Solidworks).  We think that there is a lot of power in this paradigm, but it does take a bit of a mindset shift.  Operations like creating a component, adding a new instance, or move all are captured in history.  So, when you edit an earlier feature, sometimes things can shift, because Fusion rolls back in history to the point of that feature.  Here is a very simple example:

This design has 2 components, and I added a fillet to one of the components:

Then, I move Component2:

The Snapshot UI in the toolbar indicates that some components have moved.  If I click on that, I get a snapshot feature in the timeline:

Then, if I edit the fillet, Fusion rolls to that point in history, which temporarily undoes the move:

And, once you OK the edit, the components will go back to their positions.

Hope that clears up even more of your questions.

Jeff

[SGoldthwaite]

Thanks for all the great info.  Also, what are these semicircles and how do I turn them off?

[jeff_strater]

Those are Joint Origin symbols.  To turn them off, you just have to find them in the browser (again, the "squiggle selection highlight" can help) and turn them off.  A big percentage of them, in this design, come from the "Socket Head Cap Screw_AM (3)" component.

Unfortunately, it's a little laborious today to turn these all off.  We have an enhancement in the backlog to provide a way to turn them all off, but is hasn't risen to the top of the heap just yet.

Jeff