Fusion 360 & FRC in 2019

Our team has been using Fusion 360 for a few years. I’m personally still learning how to use it. I’m decent at 2D sketch geometry analysis now, which I find quite useful, but I’m still climbing up the learning curve toward being able to model mechanisms & a whole robot in 3D. Of course, I’m also climbing up the learning curve to become an effective mech designer generally, irrespective of CAD.

We definitely like the collaboration & version control features (& CAM support). Here’s my question, since F360 has continued to evolve since the last major thread on it in 2017: What are teams using F360 now finding is the most effective workflow for FRC robot design from strategy decided to full robot model? Thanks!

I can’t bring myself to modeling in F360. The modeling seems a little too clunky (or I haven’t been using enough shortcuts).

However, I love doing CAM in 360 after bringing models in from Solidworks and OnShape.

6 Likes

Once you’ve learned it I guarantee you’ll find the SW modelling clunky.

1 Like

We typically have 3.5 folders
1 for OTS parks that don’t get modified
1 for new parts or OTS parts that are copied in and modified (kit base)
1 for old designs, we keep stuff but don’t want to clutter the other spaces. Think prototypes
The top level is were we keep the field, game pieces and main robot and sometimes (but not often) main sub-assemblies for quick clean access

I use both 360 and SW on a regular basis and both are clunky/speedy in their own ways. I tend to push 360 because it’s free and has T-splines and CAM built in. Really, if you get good at one, you can get good at the other in short order.

My team has used Fusion 360 to design our robots for the last 4 years. We do our best to organize the workspace- creating separate folders for COTS, robot subsystems, etc. This makes it extremely easy for members to find the file they are looking for.

This year, my team emphasized 2D sketching to speed up the initial prototyping process, and we were pleased with the results. @chrisrin For this reason, I’m glad you’re practicing 2D sketch analysis since it’s something most of the powerhouses do, and for good reason.

Others have mentioned it, but the CAM support in Fusion is pretty awesome. I can design a part and then immediately switch to the manufacturing workspace to generate the toolpath to router it.

I have a related question. Those that chose to use onshape for CAD and fusion360 for CAM: Why is this the optimal solution in your eyes?

2471 has been discussing moving to onshape, but it seems somewhat silly to make that change when we could kill two birds with one stone and use fusion for the whole process. One of the bigger issues we see is that if you CAM and cut a part, and want to make a revision, you would need to export a new model and start the CAM from scratch. Have you found a way around this? If not, how bothersome is it to re CAM parts multiple times?

We used solidworks last year for CAD but I was personally responsible most of the CAM this year in Fusion. If changes were made to parts, fixing the CAM was trivial. Often you could just copy and paste all of you setups over and re-select the geometry in ~30 seconds. This makes stuff like gearbox plates which are roughly symmetrical pretty easy. We have a Haas TM1 with 10 tools in the ATC carousel so we started out with an empty model to use as the master tool library, but we discovered that it’s much easier to just use the last part you made as the master for the next part. We basically never had to make new tools all year.

2 Likes

4607 has used Fusion for 2 years. We are switching to OnShape for the foreseeable future.

Could you give your reasoning for this?

Our school teaches Inventor only so for that reason Team 1989 uses inventor. So that is what I learned too and I am pretty ok with it.

Can you elaborate a bit more on this a bit? I’ve had a hard time trying to setup what I would consider a highly functional environment in Fusion. It seems the concept of a “design library” is not one that the software utilizes like Inventor and SW do.

As one who used Inventor for CAD, and (recently) F360 for CAM, I found that creating toolpath templates really sped up the process of importing and re-programming an updated model from Inventor after manufacturing.
Of course, depending on whether a design change concerns a simple hole size or core function, I may or may not use F360 for it. Overall, dedicated CAD software like Inventor seems to be easier to navigate in, and the assembly system seems far more streamlined there.

Why don’t you use inventor’s integrated cam software? That’s what we currently do.

OP here. Thanks for the interesting conversation & helpful information. Following up…

@Fields I’m also interested in understanding your team’s folder structure and workflow a bit more. Some specific questions I have:

  1. You mentioned having an unmodified COTS part/component folder. Which of following is your preference: A) There is a non-year-specific COTS part folder hierarchy, and there is also a folder each season for the unmodified COTS parts that are used (or at least considered for use) on the robot, B) There is only a non-year-specific COTS part folder hierarchy drawn from each year, or C) There is no overall COTS part folder hierarchy; there is just a COTS folder for each season, and CADs are pulled in from manufacturers as they are used / considered for use.

  2. When you say “Field”, do you mean several CADs of the different field pieces, or do you mean a single CAD of the whole field including all field pieces?

  3. I sketched out a folder structure in a Google sheet based on your description. If you have a moment, would you please take a look and either add comments or reply here. (Comments welcome from anyone BTW). Link: https://docs.google.com/spreadsheets/d/115OliPtIzgE3IAtr-aYOq9T0ZYlLVRdTptK55rsF89s/edit?usp=sharing

  4. [All thoughts on this last question are welcome - working hard to figure this part out…] As far as workflow that goes with the folder structure, from strategy decided to full robot CAD, after reading more about F360 & general FRC workflow, here’s what I came up with…

  • Decide robot archetype / mechanism options to investigate
  • Do 2D geometry analysis (CAD or otherwise), narrow list of options
  • Design prototypes, test them, narrow list of options
  • Decide on overall robot concept, including mechanism modules
  • Assign design lead roles (overall/integration, each module)
  • Overall/integration design lead(s) assign initial space, location, weight, linkage constraints for each module (to be negotiated / updated over time as details collide)
  • Module designers proceed with detailed design. Another pass at geometry is done, materials/linkages & power transmission are considered (including math), and COTs parts are pulled in, custom parts are designed, components are integrated to the overall module, there is iterative problem solving & refinement, etc.
  • As module designs are in progress & take shape, overall/integration design lead pulls WIP modules into main robot CAD (or applies space holders). Linkages / integration points are pondered. Obvious interferences are identified & addressed (sometimes meaning rework or other bad news for module owners).
  • The above is iterated as many times as needed… I feel like I might be missing something here, but it’s probably just grinding out iterations. Time. If I am missing some uber-integration step, please advise.
  • And eventually there is a whole robot CAD composed of base chassis & modules, which are composed of COTS and custom components that live in the sub-folders.
    Anything glaringly missing or wrong above?
1 Like

I don’t have both programs open in front of me right now, but I believe Fusion’s CAM has been in development longer than Inventor’s, and has more options.

You are correct in that the newest Autodesk CAM features and updates are generally deployed to Fusion first. This sometimes means that they break certain things, and since you’re at the mercy of Autodesk’s updates, you have no say in the matter. We are still running Inventor HSM Pro 2017 for CAM, and it does everything you would need, with no real bugs or glitches to speak of. I can only imagine the newer versions would be even better. So, long story short, don’t feel like Fusion gives you anything that Inventor doesn’t. They are essentially the same perhaps aside from both ways adaptive and some 5-axis stuff. You would be fine with Inventor.

As for using Fusion for FRC, we don’t, but we use it in the CAM class I teach. We get by just fine for designing and programming single parts, but for me, when it comes to assemblies, joints in Fusion are still confusing. Also, we have had numerous issues in keeping various computers all in-sync with their Fusion versions between school and home, and the different versions often do not play nice with each other, to the point where you can’t open your work. Due to the somewhat unpredictable nature of the automatic updating and cloud integration, since we can’t be in control of what the software will be like when we boot up a computer on any given day, I doubt we will ever use Fusion for FRC.

Nothing wrong with Inventor… I’ve used it for over a decade, and wasn’t planning to switch to Fusion, but through one thing and another I’ve slowly been dragged into F360 and would recommend that you give it a shot. You may find yourself making a similar recommendation to your school, if you do.

For those swapping back and forth between CAD packages, a tip I got from the guys at SolidEdge was to get a 3D Connexion Space Mouse. I got one because I was using Solid Edge, Inventor and Fusion all at about the same time and couldn’t remember how to rotate, pan and scale the viewport… but now I’ve grown accustomed to the darn thing and couldn’t imagine CAD without it! (Okay… hyperbole there. I can imagine it, and I can do it… I just much prefer the 3D mouse!)

Jason

+1 for the 3d mouse. Its great! I woukd recommend one of the oned with lots of buttons though, I didn’t and I wish I had.

We use Onshape because of its ease of use and low entry point for new learners. Plus it makes collaboration simple and you can work with it anywhere you have connection to the internet. We use F360 for CAM because it’s free and relatively easy to learn. I imagine that if Onshape had a free, easy to use CAM solution we’d try it but that isn’t the case so far.

and @craigboez since you guys specifically asked

Top

At the top level (for that year) we keep the PLAYING field for reference. It’s dumbed down by removing a lot of non-essential stuff for faster loading.
That pic of a Round bot is just a squarebot outline (with bumpers) and a circle around it. We can quickly slide this around the field to get a better perspective on size compared to the bot and look for choke areas.

OTS

Everything right off from FIRST for THIS year. Space is not an issue and it’s easier to keep track of what’s new or phased out each year.
We could have added a folder for 3rd party parts, but they ended up in the Parts folder this time

Old

Most of these designs started out in the top level folder, then as we looked at resources and stuff, we started dropping unused designs in here. Including our main idea when we found out 1/2 through the season that a supplier fell through. Oh well, move on

Parts

Finally, new parts and modified parts went here. If it got modified at all we would copy them here so as to not mess up the original. Second part down is a modified piece of the kit frame. Had we not made a copy, our drive base would have exploded.


Overall (for in-season items), I would say try to keep it simple and make a new top folder for each year.
Space is not the issue it was even 10 years ago. This way you can iterate on your folder structure and keep from inheriting a mess from years past. Next year we may start a legacy folder for common items like sprockets and gears, but I do like our online folder for that (McMaster)

If you do want to clean up folders from past years, just be sure to break all the links in the top level assembly first.

1 Like