"Extra! Extra! Auto-palooza delayed due to unscheduled productivity!" We just started making sweet prototypes, and couldn't stop. No talking, just learning! What did we learn? We're not sure... but there was a lot of "science" going on.
"Okay, here's what we're gonna do. We're gonna take this truck winch and hook it onto the staircase over there. Then we're going to start dragging some of our old robots up the wall to see what happens. Sounds good?"
To help us better understand what the dynamics of different types of robots hanging from this bar will be like, we hooked a winch up to the side of the staircase in our shop and just dragged some robots up the wall. No, really. You thought that quote above was a joke? No. That was the actual plan. It worked great.
Not everyone on the team can visualize how the robots will interact with the wall, and how the winch being tied to different points on the robot (relative to the center of gravity) will result in different types of climbs. So we demonstrated a few!
Since this is hazardous testing, we only selected robots which were cut out for it. Not just any Robowrangler robot can handle this type of work. We needed brave, test-pilot type robots. Robots with the Right Stuff. Robots which are familiar with climbing. Robots which are accustomed to falling. From high up. Hard.
We have a surprisingly large number of robots which meet this criteria.
We chose Rogue 1 (the 2017 practice robot) and Viper 2 (the 2013 competition robot). One is short with a low CG, lightweight and is relatively balanced. The other is a giant, top-heavy, front-heavy, (Division winning, Einstein level) Fridge. The "Fall Boys!"
Illustrating the differences CG can have on climbs was simple by showing differences between these two robots and by showing the differences you get from attaching the come-along to various spots on the robots.
After years and years of this stuff, and having worked on the design of five different FRC climber mechanisms (2004, 2010, 2013, 2016, 2017) I thought I had a pretty good model of the physics involved in my "HeadCAD". I won't lie - my HeadCAD failed me. The robots did some things on the way up which I didn't expect.
Side Note: HeadCAD is a nonsense word which I made up (or more likely stole from somewhere). We talk about it on 148 a lot. HeadCAD is a thing which most designers do where they imagine the robot in their head but pretend they're doing something more fancy. I love HeadCAD because almost everything works. It's a magical place. Why is HeadCAD different from just imagining the robot? It's not, but when a designer is trying to convince people their idea will "totally work" they just say "I've got it all planned out in my head. It definitely all fits." You've probably got people on your team who pretend HeadCAD is a thing and somehow better than imagination... now you can make fun of them for it!
The intake prototyping groups (we have 3 in-work) made some good progress. Right now we're still just trying out different materials for gripping the cubes. We have narrowed down to a "roller" style intake. No pinchy-pinchy claws being investigated right now.
The VEX EDR Curriculum has some great lessons on manipulator design. If you're working on an intake, check it out. (That was a shameless plug, I wrote a lot of those lessons in my previous life).
We'll see how consistent they are, and which ones are worth taking to the next step. Once we get a prototype we're happy with, we will have one of the CAD teams design up a version which isn't held together with 2x4s and uses wood-screws as it's bearing joints. That will let us test the specific spacings of wheels, and different motor speeds in a more rigorous setup that will be easy to transfer into the final robot design.
"Just copy the spacing from the prototype. We know it shouldn't work based on the CAD geometry, but it worked in real life... so whatever." - Something said basically every year on 148.
I just lied. I don't know why I did. That last paragraph is only partly true. That's how we SHOULD do things. In reality... the CAD group is ALREADY working on the more rigorous version of the prototype. In fact it's already done and being cut on our CNC router-table this morning for testing during the Day 8 meeting. We very frequently start working on our "best guess" design even before we complete testing / tuning with the prototypes. I imagine some of my readers are rolling their eyes right now in sympathy... "Designers... right? Always CADding new stuff. Can we just spend little more time trying to get this one to work before you do the next one?"
Day 8's meeting should be a great one. We'll have the new intake prototype. We've still got Auto-palooza lined up. We might talk about drivetrain selection and get the design team's recommendation. We may talk about Elevation subsystems, since other than Vader throwing some cubes up on there we haven't done any real work on getting the cube up from the floor into scoring position.