Continuous Elevator Tips?

Hey guys!

I have been doing a lot of research about continuous elevators and I would like to know what separates the good from the best. If you know of any parts, material, or have any suggestions for how to build a fast and reliable continuous elevator, please share!

Thanks!

1 Like

I do not claim to be an expert in continuous lifts (I prefer cascade), but I designed a continuous one last year. In general make sure that there are springs or a tensioning system so that the rope does not slip, without this the lift not work. In general a lighter lift with constant force springs tends to be easier on the motor and thus faster. In general, look at some of the robot profiles by FUN for robots with lifts.

1 Like

3512 built a continuous elevator in 2018 and had a lot of success with it, although we would do it differently if we were to make another continuous elevator. Here are some lessons learned/tips we learned assuming you want to build an elevator with cable rigging:

  • Dyneema is pretty much the best rope you can you use for elevator rigging. We used 1/8" in 2018 to lift our bot and a partner (~310lbs) and had no issues, but we did double up and had 2 lifting lines. It will stretch slightly initially, but eventually it gets worn in and stays solid. Our 2018 robot’s Dyneema is still tight to this day!

  • Friction is the enemy of a smooth reliable elevator. To combat friction, I heavily recommend using bearings to allow your stages to slide past one another. Many teams use R4 bearings. In 2018 my team user UHMW bushings instead, which worked ok and simplified things, but bearings are really the best way to go. Additionally, you want smooth pulleys for your rope to lay in. Many teams use something like these from McMaster Carr. Edit 3: Added correct pulley link.

  • Speaking of friction, be cognizant of the amount of overlap between stages when the elevator is at its highest configuration. If your overlap is not enough, your high stages could start to bow due to gravity/inertia of the drivetrain and the system can bind up.

  • Your life will be much easier if you build in active tensioning (via a spring) where your rope terminates on your carriage, otherwise you will constantly need to retention your rope.

  • Use a knot that minimizes slipping. We use bowlines, although some people swear by splicing the rope. We’ve never done that since tying bowlines is just easier.

  • If you want your rope to lay nice and flat on your drum (may or may not be necessary) you need to build in something to give you a fleet angle. We accomplished this by offsetting the first pulley on both the lift and return lines from where the cord starts wrapping on the drum. I believe we started with a 15° angle, which then changed as the drum wound and unwound.

  • I’m not a software guy, but programming in set points to control your elevator can give you a much faster cycle time. You will need an encoder and or a potentiometer to count the motor rotations. It can also help to add limit switches at the extreme ends of your elevators travel.

Hope this helps!

6 Likes

I have seen several of those robots. They are pretty amazing.

Constant force springs. These are pretty good. I saw on a tba blog about 846 The Funky Monkeys in 2015 where they used a Constant torque spring or something (not really sure how this works).
https://www.google.com/amp/s/blog.thebluealliance.com/2015/12/02/behind-the-design-2015-funky-monkeys-846/amp/

If you go for a t slot extrusion elevator like 195 in 2018, i believe they used some sort of delrin as the slide things that hold the elevator together.

Sometimes if you only do 2 runs of rigging, it can pull the elevator sideways and add friction. Sometimes if theres a gap in between the tubes and bearings the same can happen. If you do 4 runs and minimize the gap, this is less of an issue and it has the bonus for pass through mechs. Basically, that madtown had this year.

I’d say that 2910’s 2018 is a simple continuous elevator to look at for fun.

3 Likes

A few thoughts, I’m sure I missed a bunch of stuff.

Plan your wiring from the beginning. Many teams like to use the energy chain, but it’s really hard to add wires afterwards. It’s also really hard to package the chain after the fact.

Consider center of gravity. Heavy elevators with heavy end effectors generally lead to tippy robots.

Mount your gearbox / drum somewhere stiff in the chassis. Remember the gearbox mounting is reacting all the lifting load so it needs to be in a spot that you can really pull on.

There are several drive options. Most teams use either a belt, a chain, or a winch. Winches can get weird if you don’t manage spool wrap properly - need to make sure there is enough width to the spool that it doesn’t wrap over itself too much since this changes your effective pulley diameter (and therefore the ability to tension your cord). Also for linearity reasons - aim to have straight vertical runs to the best of your ability. Belt and chain need proper tension so design something in to do that.

Support bearings are generally the determining factor in how high you can go with each stage. I wouldn’t go much much less than an 8" span between the lowest bearing on the first stage and the top of the supports on the frame at the peak height. You can cheat it a little more but lose some stiffness / stability.

Establish a stall / hold strategy early on. Constant force springs work, but they add weight up high. If you build it light and over power it you can generally get away with stalling the powered motor / using brake mode. Do the math on this and look at Vex’s stalled rotor tests. You can also use a brake to move the weight down low.

Think about stage stiffness in the cross elevator direction. Often teams have trouble with their stages flexing apart and in the worst case scenario falling out. This is much more relevant on a continuous elevator since the stages aren’t supported by the frame as often

Plan how you’re going to control it. if you’re not using an encoder or potentiometer you’re probably not going to have a smooth elevator.

Watch 973 RAMP on elevators.

Don’t reinvent the wheel. Benchmark. There’s lot of elevators to look at.

9 Likes

Thanks guys! All this is super helpful!

If you plan to use aircraft cable on a winch, expect splinters of the most painful variety… it’s been two years and I can still feel my hands tearing :grimacing:

Adding some detail: We used Dyneema this year, and had to make a rope swap at Chezy this weekend. The spare we brought ended up being too short (whoops) so we ended up asking around and getting some rope from 972. They use kevlar instead of dyneema. We noticed that the elevator was running more smoothly with the kevlar line (totally subjective, but everyone commented on it). However, after 2 days of competition it’s already looking a little ragged in places. So, kevlar seems to be a good alternate option, but use thicker line (>=1/8") to be safe.

Also, on CF springs for counterbalancing: this can be tricky on a continuous elevator. Each stage needs to counterbalance every stage after it. So, for a simple 2 stage elevator, the first set of springs mounted to the frame needs to counterbalance the first and second stages, and a second set of springs between the first and second stage needs to counterbalance just the second stage. Three stages makes this worse.

1 Like

I’m curious why? Are you having abrasion issues with your rope? What’s your pulley setup look like? Rope shouldn’t really look ragged after 1 event if it’s sized correctly and not seeing abrasion.

At a guess, our pulleys are a bit small, and there’s quite a bit of tension. We weren’t 100% sure that it needed to be swapped, but it was on our list of “stuff that hasn’t failed yet, but might fail, and would be hard to swap out between matches.” (I wanted to replace the rope at our shop once we wrapped up our last drive practice, but we ended up running out of time.)

We also have a rope guide that might be rubbing during aggresive driving. Whatever it was, the dyneema seemed fine with it but the kevlar is fraying a little.

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