CAD Battle #1 - January 2026 (Best of the Worst Competition)

Derek_Van_Allen_BD · January 2

Inspired by @MichaelPascoe's render battle threads and some of the forum discussions around modeling practices I'd like to take a turn hosting one of my own competitions for a change this new year, but focused a little more on CAD theory and in a format that's a little more open to the non-pro users of Onshape that don't have Render Studio at their disposal. If this goes well I might do more of them as a regularly irregular thing. For the first battle in this hopefully series of threads, we're going to borrow a little page from the official Onshape CAD Challenges app

Specifically this page

The Twist: Malicious Modeling

I recently submitted an entry to the official app for the lifter bracket part linked above. I used logical tools (extrudes, holes, fillets, chamfers, the "correct" way to do things) and I got a middling time with a middling feature count. It got the job done. But we've all seen CAD speedruns and minimum features challenges before, and the challenges app is a great place to flex those muscles if that's what you're into. I'm looking to flex your imagination instead.

For this challenge, I want you to recreate that exact same part, but I want you to use the most illogical, inefficient, and disgustingly bad modeling practices possible. I want to see a feature tree that makes professional engineers weep. All of the features used in this studio are pretty reasonable tools to use to produce the sample geometry. This competition is looking for unreasonable solutions, so these cannot be used in your submission. Except for Sketch. Sketch can stay.

this time

🚫 The "Sanity" Ban List

To force you away from your good habits, the following features (which I used in my original model) are strictly forbidden:

Extrude & Hole: Too straightforward. Find a more difficult way to generate volume.
Fillet & Chamfer: If you want a rounded edge, you’d better find a way to "carve" it or loft it manually.
Shell: This geometry is what shell lives for, so it's gonna have to sit this one out too.
Mirror: If the part has repeating features, I want to see them modeled one... by painful... one. Differently if you're feeling spicy.
Boolean: I debated about this one for a minute but boolean operations are baked into most of the standard library tools, so banning the boolean feature explicitly should lead to some gross workarounds.

Custom features are allowed but clear derivatives of these core features will be just as off limits as the features themselves. We'll all be much more impressed if you can somehow shoehorn the Airfoil Generator into a usable workflow than if you just found and used "Plus" versions of all of the features in the ban list.

🏆 Criteria for the "Winner"

We aren't looking at the stopwatch here. We are looking for:

Creative Abuse of Tools: Using a feature for something it was never intended for (e.g., using a Sheet Metal workflow to make a solid machined block).
The "Why?" Factor: When I open your feature tree, it should look like a crime scene.
Dimensional Accuracy: Despite the intended chaos, your final geometry must match the target part perfectly. Professionals gotta have standards.

Submissions will be open for the month of January and then the first week of February we'll take a forum poll of submissions for whichever feature tree offends us the most. The worst (best) one will get mailed a Shapey.

ry_gb · January 2

Do I get bonus points for using Triad Transform? Anyway, I'm sure people will come up with worse because I still did things in a straightforward manner (loosely speaking). Here's my attempt: Most likely NOT the worst offender.

At some point, I just got lazy because I didn't want to keep creating the a cylinder over and over again, i.e. if the geometry was created once with a method, that's more than enough to prove the point. I sprinkled in a little extras at the end as well.

Lastly, all the sketches are fully defined because we're not heathens here.

EvanReese · January 2

hahaha this challenge is silly…. but I had to try. Highlights:

✅ No sketches used

✅ Over reliance on my Cone and Cube custom features

✅ Some trigonometry to get the angled face right

✅ Replace face which keeps the original corner rounds.

✅ More than double the rebuild time of the original

https://cad.onshape.com/documents/70f550586b27aa6e3bf80866/w/87953cc4b98976f216b607f2/e/9600d3bea10a053065570a7c

EvanReese · January 2

@ry_gb Those sweeps for cylinders are absolutely disgusting 🤮

Derek_Van_Allen_BD · January 2

Horrendous first submission @ry_gb. Helix wasn't even on the list of features I expected people to use, which in hindsight it should have been since I once attempted to use it as a 3d circular arc tool. Fun fact about opHelix is that it allows you to specify a spiral with zero helical pitch, and zero flat spiral pitch, but NOT both at the same time which would just be a flat circular arc. Had to write my own 3d arc utilities for that.

Caden_Armstrong · January 2

I wish I had the time to contribute, but can somebody please make one that is entirely made from offset faces nit together from standard content in an in-context part studio.

martin_kopplow · January 2

Let me have a whiskey or two first. ;0)

I had the idea of absolutely neglecting the parametric nature of everyting and only draw single unconnected curves.

EvanReese · January 2

https://forum.onshape.com/discussion/comment/125078#Comment_125078

This is awful. The only worse thing I can think of would be to model each face in a different document, and solidify them via an assembly and in-context edits.

Caden_Armstrong · January 2

https://forum.onshape.com/discussion/comment/125080#Comment_125080

This might be entirely possible with just routing curve and fill/loft

MichaelPascoe · January 2

I approve! 🤣 this is an amazing idea @Derek_Van_Allen_BD. Ty for hosting!
.

Derek_Van_Allen_BD · January 2

@EvanReese what's funny is that this feature tree is pretty much intended workflow in other non-parametric software. I just had someone in the office ask if Onshape had primitive generator tools since he comes from a more artistic background.