This weekend we set out a goal to do a basic robot CAD, and it turned out just to be me doing it in On-shape, I have basic knowledge after completing the self paced courses. I spent yesterday and today trying to CAD and all I did was edit the AndyMark Chassis to cut it upfront, from there I tried to build the Bellypan I spent 2 hours and got nowhere productive. I’ve watched the videos online and looked at CAD’s from teams but I have no idea where to continue and be more productive.
So my question is how do teams make a robot cad? And I know its late but how should I continue to do it.
Are you asking about organization within the file or how to use the program itself?
My team has been using Onshape so far this season and our system has been to have a single document for the whole robot and several folders for each sub assembly or mechanism (drivetrain, ball intake, etc.). Inside each folder is an assembly of the mechanism and all of the part studios needed for it. Outside of all of the folders is the assembly of the whole robot. This allows us to do all of the basic design work and clutter of dozens and parts and mates out of the final assembly. I’ve found that one of the keys to using Onshape efficiently is naming parts in part studios. Just naming the part studio doesn’t really do much, but on the bottom left corner is a list of each part in the studio. Right click there and rename. Doing that helps with organizing large assemblies. One of my favorite things about Onshape is the ability to edit and create parts in context of other parts in an assembly. Parts like mounting brackets can be created very easily by using the geometry already present in the assembly.
If you’re wanting to learn about how to use Onshape itself, the website has tons of useful video and text tutorials about all of the functions within the software. You might also check out tutorials for other software like Solidworks or Inventor. While they have their differences, many of the same practices and strategies can be applied when designing in any CAD software. If you have any specific questions about Onshape or anything else I’d be glad to try to help as best as I can.
My question is more of how is a CAD produced I defiantly do need to work on my skills its just how do teams or you guys come up with the design on paper then throw it into CAD and make the robot, that where I’m lost.
In the past my team would draw a basic geometry on the white board and just throw it together on the robot.
Learning to design is definitely not a skill you can learn overnight. IMO, there are two things that you can do (during the offseason) to learn to design FRC robots:
a. Practice! Design common FRC mechanisms (arms, elevators, shooters, intakes, drivetrains, etc) over and over again, trying to iterate your design each time. Get feedback from people you know on other teams that might have more experience. There aren’t that many unique FRC challenges where you can’t draw inspiration from previous years.
b. On that note, study the designs of past successful robots. A number of high performing teams such as 148, 1114, 971, and others, all post CAD of their robots post-season.
Well done design takes knowledge, practice, and familiarity with manufacturing and assembly methods. It is no simple task, and for a team who is inexperienced with CAD I don’t recommend trying to design your entire robot.
Remember that 2D CAD is essentially a geometry solver. Knowing that, I recommend you use some 2D sketches in order to figure out general layout and sizing over designing the entire robot. Sometimes with those simple sketches you can really get some incredible insight into how things will work out.
First of all, USE A MOUSE with a wheel. You cannot do CAD reasonably with just a trackpad.
If you don’t do the full 3D mockup in OnShape, you can still use sketches to draw out your robot layout from 3 different angles and deal with measurements. That gets you some of the benefit that CAD gives you (mainly, making sure that you see where everything is supposed to go and make sure that no two things have to be in the same place.
Like engineering, CAD is a bunch of little steps to get to the final design. A lot of the parts can be done in parallel, but for some parts you have to do things sequentially. Designing a single part is a series of sequential steps. The more you do CAD the more you know the steps. The more you do CAD the more you know what results come from the commands. You will also know that certain commands always come before other commands. You will always create a 2D profile before you extrude, but you can also incorporate holes and/or lightening data beforehand. Blends are most always last, with large blends before small blends.
Just remember when you see a video of someone creating a part in CAD they’ve probably already made every mistake you could possibly do and the result is something that looks easier than it really is.
CAD is a great tool if you know how to use it efficiently. It’s even better if your team knows how to use it as well, since more people help expedite the process. On my team (2729), we designate different mechanisms on the robot to different people and have them design and CAD that specific part. For example, we have a team of 2 people on chassis to get that done quickly, and 1-2 people are assigned to each end effector (hatch, cargo, HAB, etc). Then we have an “integrator,” who is essentially in charge of assembling the entire robot together and ensuring that no mechanisms overlap.
In your case, since you are basically a one-man team, you need to prioritize what aspects of your robot need to be designed first. I would primarily focus on creating your end effectors, since those are the more critical components to model in your CAD software. Once you have those parts assembled, retrieve the dimensions you need and sketch on a piece of paper where they would go to make sure they fit within your frame perimeter. If you have time later, you can edit your chassis and constrain everything into one master assembly if you want to include it in a presentation or technical binder. Hope this helps!
Your strategy of the white board drawing is a perfect start. What my team did this year was pretty similar to that. From our experience with past robots and other teams we had a general idea of what kind of mechanism we would use to do each task in the game. Developing an actual design out of a simple idea is a lot harder than it sounds and it really takes practice and experience more than anything else.
First, get your concept. Maybe make a few sketches and discuss how this kind of mechanism works with the game objects and with the rest of the robot design to help you score points. At this point, details aren’t super important and you’re just trying to figure out what could work, not how it could work. This discussion should really only happen on kickoff and the first week or so of build season.
Second, develop prototypes with things like plywood, easily adjustable extrusion like 8020 and versa frame, and try to see if your idea is actually feasible. If it turns out this idea doesn’t work, try something else. Once you get more practice you can incorporate CAD into this stage as well. JVN made a great blog post about this concept a few days ago you should check out.
Third, start modifying that prototype into a mechanism you can actually put on your robot. This stage is where CAD becomes really important. It helps prevent mistakes that are easy to make if you’re free-handing everything and it also serves as a quick sanity check before you’ve devoted a lot of material and resources into something that is either wrong or won’t work. CAD is a really powerful tool because it can make this stage of designing a robot so much faster than doing it by hand… If you know what you’re doing.
I’m going to caution that it might be a bit late in the season to learn CAD and implement it effectively into your design process. The steps I outlined above are over views at best and understanding how to really make it all work together is something that’s hard to learn without practice. As others have said, look at the CAD other teams have released of their robots. Try to understand why they made the decisions they did and ask yourself if you would do it any differently. 1114 and 973 have also released some tutorials on using Solidworks in FRC and go over how some common mechanisms can be designed and integrated into a robot.
Edit: I realized that step 3 was a bit of a “draw the rest of the effing owl” moment. I’ll expand on that now. Think about things like mounting holes for your frame and motors, how you’ll get power to where you need it, and what kind of material you might build the mechanism from. Do you need to use a solid block of steel? Can you use polycarbonate sheet? Is this member going to experience a lot of load during a match? How have other teams designed systems like this? These are all questions I ask myself when I’m designing something for my team. It’s important to keep in mind that not everything has to be perfect as long as it works. I’m not sure what your machining capabilities are like, but you can go a really long way with some of the versa products Vex has and you don’t need to build the sleek, polished robots teams like 148 and 254 do to be successful. This year, focus on speed and consistency. A fast, simple mechanism can go a really long way and adding much more to it might not actually make you score more points.
In my experience, you have each part in its own file. Then you assemble all the parts of a specific mechanism. After that, you have an assembly with all of the mechanisms together.
Could you and your team have measured the AndyMark chassis you built and cut out the bellypan? If so, you may want to consider using a whiteboard and sheets of paper to work out your design. A CAD model of your robot is not necessary to build it.
As BClarke stated, Build season is not the time to learn how to use CAD tools. CAD tools are a method of capturing the ideas of the designers. A good idea drawn on a napkin is still a good idea. A bad idea captured in CAD is still a bad idea. The spacecraft that took men to the moon would have been designed on equipment like this.
i see a lot of good responses here, but there seems to be a critical part of the procedure that no one has mentioned.
When you design something, it’s a little like cooking a recipe. You put stuff that already exists together. You don’t design your own ingredients.
So you’ve done your proof-of-concept models and hopefully you’ve picked your systems…the wheels, the motors, the electrics, the air cylinders and tanks (if you have 'em) .
Your CAD model needs to START with all of these PARTS in it. Only after you have all the parts can you start ‘designing’ how they hook together.
This applies no matter what CAD system.
The most useful thing about CAD is that you can tie everything together in a system and make sure it all fits, it doesn’t crash and you have something to bolt everything too.
Also, in our designs, we often work with specific systems as subassys. Just advice…not a requirement.
Actually, this is not how Onshape works and may be a bit misleading. In Onshape you model all related pieces of your assembly in one (or multiple), “Part Studio.” This is done so features can be consistent between components, for example, you can create a hole in the part studio that goes through 3 separate parts. If you change the size or location of the hole, those changes will propagate through all of the parts the holes go through.
The assembly stage is very similar to most other CAD workflows besides some quirks with mates, but part creation definitely requires a bit of rethinking of how to model things. It’s actually incredibly nice and really saves a lot of time being able to model multiple parts in one place with relating geometry and features.
So I’ve been on teams at both extremes of CAD. Merge Robotics starts CAD training in the fall and has a formal numbering system where students add specific parts to a shared drive. Every single piece is CAD’ed before making them which makes fabricating two identical robots much easier.
Where you are right now, I suggest just using it for dimensional checking for major items. See how a ball will fit in your intake. Determine range of motion/height for an arm. The main intent is to prevent cut and drill 3 times to find the right pivot. These things can be done in Sketchup or TinkerCAD or any CAD software that you are already familiar with.