3005 has a unique approach to creating block CAD that produces a much higher fidelity model than most existing workflows. The advantages of this method are:
It provides more stable CAD models that are tolerant of frequent major changes throughout the season.
Large design changes in the middle of the design process are faster and more intuitive to execute.
It enables less experienced designers to work more autonomously due to the precise model better communicating the design intent for each subsystem.
Earlier access to an accurate model of the robot improves team wide understanding of early design direction and also enables planning wiring and other non mechanical systems earlier.
I am currently working on another set of CAD class sessions on more advanced topics that I plan to record and release this fall. I will make a post with more information when my schedule allows but I plan to cover the layout sketch and block CAD process in significantly more detail than I was able to in this talk. I will also cover topics such as sheet metal modeling, configurations, and BOM management.
This was a really great presentation by Orion and +1 for this type of workflow. I have taken this approach in the seasons since covid and IMO it is definitely the best way to approach designing an FRC robot, whether it be by yourself, or in a group. I work in SolidWorks and all of this same stuff can be done in almost the exact same way in SW, and probably most major CAD platforms.
This is even more high fidelity than we typically do on 254 but looks quite appealing due to the variety of benefits you highlighted.
@Nick_Coussens do you typically model to this fidelity?
I feel like going to the point of inserting the block CAD and overlaying / relying on that can lead to broken external references and such in Solidworks that I think Onshape is more resilient too.
That’s probably the main difference of my approach to Orion. I typically make a master file to do all the parts for a mechanism in, and eventually those parts all get saved out into individual files (with OnShape you keep everything within the part studio).
I also do not mate anything to the blockCAD at all. I usually mate things to the origin as that is much more robust and resilient to changes without breaking geometry and causing yellow or red rebuild errors.
This was such a great presentation. For intro exercises that students can practice with, I was trying to think about what the simplest block CAD could be - something like a gearbox, or a driver’s station. Any other ideas?
Broken references (like compile errors) are sometimes great! I’d rather see an error early than deal with a physical interference after something is built. Preventing the unnecessary errors is a matter of referencing the right geometry, which is an art (in CAD and in code). I really appreciate the tips that aim to help build up best practices in that direction.
I’m intending to simultaneously teach myself onShape (from years of Fusion 360 [never for FRC, though]) and will also be trying out the 3005 Block CAD approach to plan my garage unholy-mess-to-workshop transformation.
That garage project wouldn’t be interesting to students, but I think some pit equipment, workshop workbench, or even a simple drivebase could be additional ideas (driver’s station and gearbox are also good ideas).
Things I intend to prioritize (after trying it myself) is challenging the students with a modification exercise of “it turns out we want to make the robot 32"x28” rather than 28"x26" that we originally planned" or “we need to shrink that workbench to 48” W by 32" deep" and use that to learn what breaks when you modify the base CAD.
I used high fidelity block cad for my summer design class. It makes it really easy to communicate design ideas with other designers and I really recommend it for any team that has a large number of less experienced designers. It worked really well for the portions of the robot that have been designed so far.
I just block cadded another robot and timed myself to test the time commitment it is.
This is the time breakdown:
Item
Time taken
Master Sketches
3 Hours
Drivebase
10 Minutes
Intake
15 Minutes
Elevator
35 Minutes
Shooter
40 Minutes
Pivot
20 Minutes
Climb
20 Minutes
Total Block CAD time
Around 2.5 Hours
Total time
Around 5.5 Hours
I did have my previous block cad to guide me, but this is also a pretty complex robot. With more focus and practice, I could probably get the total time to high detail block cad robot to be <4 hours. And of course this will greatly decrease with simpler robots. I could not recommend this workflow enough.
I didn’t have too many integration issues when creating the bot, but when I saw them they were very easy to change quickly.
I’m really interested in this workflow, and the frcdesign.org team is currently looking into integrating it into the current learning course. If possible, I’m curious about your experience teaching block CAD to students. What tends to cause the most misunderstandings? What is the hardest to understand when learning it conceptually, and when do you usually introduce it in terms of student experience?
I’ve made block CAD before but it’s always just for concepting, and gets thrown in the trash when it comes time to doing detailed CAD.
I generally always do layout sketches to drive my parts.
Can’t believe I’ve never thought to use a derived part like a layout sketch. Soon I’ll be redesigning a piece of ag equipment in SOLIDWORKS and will be attempting this. My concern is broken references, we’ll see how it goes.
To be fair, I was really skeptical at first about block cad. I tried it anyway. My first intital reaction from someone who typically solo CADs robots doing this workflow for the first time is that it really decreases the amount of cognitive load needed during the design process.
Usually I’d have to hold multiple mechanisms in my head in terms of positioning, even after integrating the master sketch (and effectively CAD every single part studio at the same time), but its nice being able to offload that to CAD instead of holding it. Doing that integration work early will make it nice if I plan to design the robot in detail later.
I initially thought that it would primarily benefit teams with inexperienced designers trying to make it easy for them to get involved with cadding the actual robot, but I think the ability to integrate into 3D super early with high fidelity quickly is useful.
I’m still unsure about including the master sketch in the subassembly documents, but I think I’ll play around with it more and see how it goes.
In terms of actual CAD practices involving block CAD, I’m not really sure how to go about it, and it’s kind of uncomfortable (there is so much sketching, extruding, and mirrors), but I assume with practice, I’ll figure it out and get used to it.
