Elevator Design Questions

Technical Technical Discussion
1

As an offseason project, our team is designing and making a telescoping arm, like 233. We’ve done one in the past, but maintenance was a big issue. Our goal is to make one that is really strong, so we don’t have to keep taking the stupid thing apart.

After our first brainstorming session, we have a few questions about the best way to make this thing. Our first question deals with the belt/wire/chain that raises and lowers the arm. We really liked our chain system, which worked the same way that 254’s belts from 2013 worked. We’d like to implement a very similar system, but use some type of cable instead. We’d have a drum with two separate cables wound in different directions so that as the drum rotated, one cable would get pulled in, while the other would get let out. Our question is, what type of cable can we use? We’re looking at cable designed to be used on pulleys and drums, rated for 1,000 lbs with a .125 diameter. We’re not sure what we will end up doing with our arm, but lets say we lifted our robot. Is this strong enough?

Our second question deals with the drum itself. Is winding up the cable in a nice way something to be concerned about, or will it be okay to just let it wind up any way it wants?

Our final question has to do with the bearing setup. We plan on having ball bearings roll along the inside wall of the tubes, but we’re afraid that inconsistencies in the wall thickness could be an issue. Is it?

2

We have used both wire cable and Spectra cord in our elevators, with dual reels so it is powered both in and out. We have found the Spectra cord to be superior in that it allows small radius (1/4") returns allowing for a more compact design. It is light weight with high strength. We have used springs and bungees to take up slack produced as the cord winds over layers and changes spool diameter. Winding on one drum surface will minimize this problem.

Extruded tubing does vary in dimension, and if you are expecting a tight fit, it may bind. We have used square tubing and nylon bearings made to slide within the tubing, and allow for dimension variances. We have done two stages with no issues. :slight_smile: :slight_smile:

3

3467 has used steel cables for the last two years on major mechanisms. Last year, we used a single .125" (I think) steel cable to lift the robot in our climbing arm, so .125" should be strong enough for use in a telescoping arm. We haven’t really had a problem with cables winding on the winch spools; they seem to naturally spool in a reasonably orderly fassion (we do use pulleys to make sure the cable comes in at a consistent angle).

4

Take a look at the 1100 entry in the 2007 FIRST design book. It is entirely devoted to highly robust elevators.

The implementation uses chain and 80/20 but could be easily adapted. The best part is that it is highly deterministic in its positioning and is modular in its assembly.

I have personally used its construction to lift 2 robots off the ground simultaneously and played with in 4 different FRC games and never had it fail. It is not the lightest weighing design but it is quite robust and a great option if your team is looking to begin working with linear motion.

5

Please, for the sake of your hands, do not use steel cable. I’m prepared to say on anything, but especially where you want to use it.

First, steel cable (especially 1/8" stuff, which is monstrously big) hates to bend around small radii. I’m sure that in a telescoping arm, you’ll want the cable to do a full 180 degree bend with about a 1/4" radius or so. I’d bet that a 1/8" 7x7 steel cable is rated to bend with about a 4-5" diameter, and that if you tried to bend it with a 1/4" radius, you’d almost immediately crimp the cable. If it was under robot-lifting load, you’d probably break it within about 3-5 cycles, even with that 1000lb rating. Because you’d be bending it around such a small radius, you’d essentially put little creases in the steel cables that are just begging to break as soon as you put load on them.

Second, it’s a pain to design spools that are small and wind up steel cable. You have to out a lot of thought into how you terminate the steel cable on the reel, especially if it’s under load.

Third, steel cable is a pain in the… Err… hands to work with. I can’t count the times I got steel cable stuck in my hands is season. It’ll happen basically any time you touch frayed cable (which would probably happen a ton in your design) or the ends of a cut cable. Trust me when I say it’s very, very painful.

This experience came to us the hard way this season. We used steel cable to pull back our shooter. We tried 1/8", 1/4", 3/32", steel, galvanized steel, stainless, vinyl covered, 7x19 cable, and 7x7 cable before settling on the cable that seemed to take the longest for us to break: 3/32" 7x7 stainless steel cable with stainless steel crimps. Even with so much iteration, we’d break a cable very 50 shots or so, because the cable would wear over time and eventually break as it wound around a 3/4" diameter idler pulley. This is where my concern about you trying to wind a larger cable around a much smaller idler comes from. I think I must have replaced the cable 10 times on the practice bot alone, getting steel spinsters in my hands each time. The cable also broke before our first match at SAC, essentially keeping us from playing offense almost all of the regional.

We switched to 1" wide military spec webbing for SVR, and had zero problems. I highly recommend it. We got it at REI, but honestly the stuff on the tie down straps on cars that you can find in any hardware store would probably be just as good. While it is larger, it will not break (except under truly extreme load and with significant abrasions). It also seems like the Gates belts that 254 used would be particularly good in this application (as you don’t have to worry about a spool, you just use a timing belt pulley). You could also look into a leadscrew (I believed team RUSH used one in 2011) or rack gears (which I think 233 themselves used in their 2013 shooter elevation).

So, basically, it sounds like a cool project. Just please, please don’t use steel cable.

6

Rope is very good for these kinds of things. Any high performance rope is going to have more then enough strength at 1/8" diameter with very limited stretching. Here is a mcmaster-carr source for rope however you will probably get a better deal at a marine supply store. Make sure you get someone good with knots though as that is important when using rope.

7

I’ll bite on this one. I think my brother put it best when he described cable as the solution that you go to when all other solutions won’t work. I would work very hard to avoid cable. It may seem cool, but you have to pay attention to a lot of stuff when designing a cable system.

When you spool the cable on and off the 2 spools, how are you going to design it so that the cable length doesn’t grow or shrink? I have some not very fond memories of dealing with cable tension on 254’s 2011 robot, and all the controls headaches that the solution caused. If you naively spool cable onto a spool without proper cable management, the cable can wrap over it’s self, and take up more length when spooling than unspooling, which is bad. If the cable doesn’t wrap over it’s self, it will walk horizontally across the spool, causing the cable length to also either grow or shrink. If you then decide to ignore this problem and brute-force a solution, you need to either leave slack in the cable and prevent it from falling off of all the pulleys or tangling on the spool and/or add a spring tensioner in the system.

In 2011, 971 used timing belt the same way that 254 did in 2013. The really nice part about the system was that there was no spool and the belt run was designed so that it didn’t change length at all while the elevator moved. You can do the same thing with 25 chain if you don’t like timing belt. The end result was very compact and clean, and I would highly suggest looking at doing something similar.

8

A noteworthy elevator in 2011, if I do say so myself, was 469. If you can find some video with sound, it’s amazing the abuse that was put through and it didn’t break… :slight_smile:

9

We are an FTC team, but we can help too! Our linear slide is of a similar design, it uses slide blocks connected together, with a pulley on the top and bottom of each block, and a spool with webbing to lift it and use gravity and oil to lower it, but if you had pulleys and another spool it would work just the same (only better, but we only get 8 motors :)) We use webbing around the pulleys and spool to lift it, and we have killed motors before we have killed the webbing, so something like this webbing here would probably work well!

10

Another simple and elegant way to make a elevator is by using a rod-less cylinder.

Take a look at Festo rodless linear drive cylinder pneumatic actuator DGP-25-457-PPVA-B

They come in a variety of lengths and can be fast and accurate. They are used in material handling equipment.