Team 1519 Mechanical Mayhem - 2019 Robot Lifter (Climber) Development

At this most recent weekend’s Granite State District event, where we (1519) successfully climbed to the 3rd HAB Level in 12 of 12 qualification matches (we shared L3 responsibilities in the eliminations with our 1st pick, 5687), a lot of people asked us about our lifter and if it had “just worked” from our initial design. Well, no.

What is on the robot currently is “iteration 4” of the lifter – it’s actually pretty true to our original concept, but with a lot of minor upgrades to make it work more reliably. We’re still not sure that it has sufficient durability to last the rest of the season, but we will try to address that before our next event.

The lifter uses a pair of AndyMark 325:1 Dual 775 Sport gearboxes followed by an additional 3:1 reduction for a total reduction of 975:1. It’s powered by a total of 4 RedLine motors, but we are doing our lifts at only 8.4 volts to keep the upwards acceleration down to a manageable level. :wink:

The video immediately below shows the lifter as it was functioning at the Granite State District Event:

1519’s Lifter (Climber) at Granite State, Week 1 of 2019

To get there, however, we went through lots of failures along the way. The most major of them was when we first added the upper assembly (arm tower, arm, and intake) to the robot, and the climb attempts just resulted in the robot flipping over backwards…

We ended up addressing the tipping-over-backwards issue by adding a pair of pneumatic cylinders at the back of the robot to do a “pre-lift” which prevents the lifter from imparting a big initial rotational moment to the robot when first lifting off the ground.

A playlist of about a dozen of our attempts along the way (some good, some not so much) is on our YouTube page:

If you’d like to see our robot in a complete match, the GSD eliminations would be good to watch (as a bonus, you’ll get to see 5687’s excellent robot, too), but we don’t yet have those videos on our YouTube page. Hopefully by tonight we’ll have either elimination match videos or a highlights reel available.


Love the iteration!

Nicely done! Watching GSD your climbs were reliable and looked really smooth. Goes to show how much time and work goes something that looks “simple”.

Do you mind us asking the specs of the pistons your team used? We are facing a very similar problem with our very similar climb and wanted to know if this method could be modified and implemented into our robot. Really great work this is. Well done.

Not at all. We didn’t really research exactly what we needed, given the fact that we hadn’t planned to use pneumatic pistons for this purpose. Rather, once we found we had the problem, we just grabbed some that we had acquired with the Bimba donation in a previous year, in hopes that they would eventually be useful. The ones we used are Bimba M-178-DP, which is a 1.5" bore x 8" stroke cylinder. I think a smaller bore would probably work, but that was what we had in stock!

Thanks for the kudos! It has been a lot of work, even though the end result looks pretty simple.

Adding the pistons did two VERY important, related things for our lifter: 1) it got the robot’s weight off the “tails” of the linkage bar on the floor, which reduced the torsional load on the linkage, and 2) combined with rocking the robot’s CG forward a bit, prevented the backflip.

For the long explanation on getting weight off the “tails of the floor linkage bar:” for packaging reasons we required the linkage to stow flat, which meant that instead of having the two long linkage bars inline with each other, they were spaced ~1.5" apart on opposite sides of the horizontal linkage bars. Also, due to the tight geometry constraints on stowing the linkage and getting the “lift height” required, we had to start the lift when the linkage was only about 10deg deployed (I don’t know the exact number, but it’s the right order of magnitude). This means that in order to get a >150lb vertical force, the tension/compression forces in the linkage members need to be roughly 800-1000lb (150/sin(10) = 863lb)… while we did design the linkage to be quite strong, the fact that the linkage parts weren’t all in-line meant that we also have those near 1000lb forces acting as twisting loads, not “simply supported.” Since the CG of the robot begins closer to the “tails” of horizontal linkage bar on the floor, the linkage would require incredible torsional stiffness in the joint between the two vertical bars to provide that entire lift force at the tips. We realized we needed to do something to get the weight off the “tails” of that horizontal linkage bar for the first few inches of the lift so we could get out of the “killer sine of the angle” zone, so we added the “boost” pneumatic pistons in about the only location that wouldn’t interfere with the drive train or climbing linkage - that was our “Christmas miracle” moment on Monday, just a few days before load-in at GSD.

This video ( actually shows how the torsional loads result in the tails of the linkage bar “toeing out” even with the pneumatic “boost,” but it was much more severe without the pneumatic “boost!” Unfortunately it was really hard to see exactly how severe before the pneumatic “boost,” since the robot was still too low to the ground and tried somersaulting onto this viewing angle!

Floor Linkage Tails roughly parallel at initial contact with ground:

Floor Linkage Tails “toe out” sharply as they take the weight off the pneumatic pistons:

Floor Linkage Tails spring back to approximately parallel as the lifter gets higher (less “killer sine of the angle” affect!) and the CG shifts forward:

As a side-note, it was very important that the pneumatic pistons be pushing directly on the floor (or the floor linkage bar) because our problem wasn’t that the motors couldn’t apply the necessary force, it was that the linkage couldn’t handle the load of the full robot weight (given its CG) at the low initial angle.

Do you mind us asking the specs of the pistons your team used? We are facing a very similar problem with our very similar climb and wanted to know if this method could be modified and implemented into our robot. Really great work this is. Well done.

Ken gave the specs above; they are “over-powered” for the application (capable of lifting ~100lb at 60psi per cylinder), but we didn’t have anything between 0.75" bore (only ~25 lb at 60psi per cylinder) and this 1.5" bore with a 6-10" stroke length. We could probably make do with the force of just a single cylinder, but it seemed safer to avoid balancing on a single point and the additional force gives a stable point for the robot to rock back onto without starting to drop at all. Theoretically we could have just used 6+ 3/4" cylinders, but that’s a less efficient force/weight ratio!

This handy table is from Mark Mcleod’s great pneumatics resource on the team 358 page: - Robotic Eagles - FIRST® Robotics Competition It’s simply a multiplication of the air pressure and cross-sectional area of the cylinder (pi*r^2), accounting for the presence of the piston rod on the retract. Nothing fancy, but very helpful!

I wish I had a shot of the climber linkage by itself… it’s really hard to see under the robot, but is a pretty clean unit (that is a PAIN to get in and out for maintenance!). The gears that drive the final lifting stage are monsters… 12dp, 3/4" face width gears with a 4" pitch diameter. We had flash backs to our 2014 and 2016 arm gears, so decided not to mess around with 20dp gears… we still lost gear teeth on some of the prototypes though! So, on a side-note, does anyone have recommendations on what hardening process to follow for unhardened steel gears (1144 steel, McMaster PN: 6325K32)? Bonus points for someone local that does it! :wink:

1 Like

Thanks so much, this is great information to have as we try to fix our climb over a single weekend before comp. :grimacing: Will update you guys on how it works out. This is so amazing and gracious of you guys thanks!

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.