Fusion 360 is able to handle large assemblies, even complex marble machines

[nathan.skalsky]

I thought I'd share Martin Molin's latest Wintergatan Wednesday video since this one focuses on, what can be a delicate topic,  large assembly performance :).   This isn't intended to declare all things solved, large assembly wise, more of a journey, however, the MMX (Marble Machine X) team is back to making rapid progress and enjoying the experience of building it (both in Fusion and near-parallel in real life!) and I suspect others might benefit from Martin's story.

We had a very candid deep dive on this over the last couple weeks and believe we have preserved "the joy of Fusion" (TV show idea?), even as Martin's machine approaches the limits of detail in assembly complexity in the current (and constantly improving) Fusion platform.  Maybe I should be a diplomat? 🙂 

How does this video compare to your experience?

-Nathan

[TrippyLighting]

Hmmmm.... do I take the bait, or do I not ....

Ahh, what the heck, I'll take it! 😉

So, right, the title does not phrase this as a question but as a statement and provides this video as proof that it can be done. But I knew the already. I've analyzed hundreds of Fusion 360 assemblies. Everything from noob to pro. Here on the forum and for people that have approached me privately.

The key takeaway from that is that if you have enough time and experience in Fusion 360 under certain circumstances, this can made to work. The key word that is also mentioned several times in the video is W.O.R.K.A.R.O.U.N.D.

Grounding many components in order to improve performance is most certainly a workaround!

The marble  machine is a personal project of a very ambitious user that has the time to go back numerous time simplify here, change an assembly there and do a good number of adjustments and experiment with different things hoping that the project remains workable.

The thread that was referenced is about designing custom automated machinery in a productive, professional environment. In such environments experimentation just to improve performance is a luxury that no one has. Hope that it might work is not a sound business strategy. It HAS to work! Workarounds that might need to change from design to design to address other shortcomings are not an option. 

Productive, professional environments are  what I am referring to  when I say that large assemblies are not a good match for Fusion 360. That, BTW has not only to do with large assembly performance, but with lack of functionality in a number of other key areas. I'd be more than happy to elaborate in a phone/zoom/skype conference.

[nathan.skalsky]

Hi Peter,   I agree with your points.  This video (nor thread) wasn't really about how things should be (and hopefully sooner than later will be), but rather what can be done (yes with workarounds) that might surprise people who, like you, know the large assembly limitations all too well.

I could have done more to make that clear in this thread, I let the character count limit in the title push out the "Is ?" part and that's probably why it reads a bit more provocatively than I intended. 

Thanks!

-Nathan

[TrippyLighting]

@nathan.skalsky maybe we could start collecting some of these tips and workarounds in this thread. Some of them are discussed in the video, but there are a number of others that might be useful.

[nathan.skalsky]

@TrippyLighting, It would be awesome to consolidate best practices and workarounds aimed at helping folks be more successful with large assemblies.  If I had my choice, I'd prefer to see us (as a community, not just you and me) help Autodesk both improve and quantify the kind of assembly performance we are talking about (ie. modeling/solvers not GPU/rendering) and have them present this back on a technical level (ie. vs. marketing messages). 

I think there is plenty enough goodwill for this to be productive and open.  

Towards that end, not finding an idea station item that seems to capture this, I created one here >https://forums.autodesk.com/t5/fusion-360-ideastation/2018-large-assembly-performance-community-init...

Check it out, let's vote it up if this sounds like something the community would like to see prioritized. 

-Nathan

[dean.dekovic_gideon.brothers]

Hi @nathan.skalsky, I voted your idea.

Is there any continuation of the thread with large assembly tips and tricks?
I saw The Wintergatan videos, the AU handout by Trippylightning and Strater. WOuld like to know moreto show to my colleagues some easy-to-accept tips.

Best regards,
Dean

[nathan.skalsky]

Glad to hear it Dean!

I haven't heard of anything beyond the videos you mentioned.  Jeff Strater's AU handout is the latest from a proper authority on this subject that I'm aware of.

I've done a few mechatronic (Eagle FTW) projects over the summer/fall and have had good experiences, even with my older dual-core  2013 Macbook Pro (Haswell).  These are far more modest (in # of parts and # of awesomeness than Martin's MMX) however I think the general principles of managing model fidelity, parametric history, and a balanced approach to major/minor assemblies play a role (as touched on in the Wintergatan video). 

-Nathan