Uniformly Tensioned Rubber Band System Analysis/ Guide

VEX V5 General Discussion General Forum
1

WARNING: LARGE IMAGES

Hello vexforum,
Due to popular demand, I thought I’d explain a little about uniformly tensioned lifts and why they are useful. I’ve explained the uniformly tensioned rubber band system in my robot reveal, but as the pictures are not working and editing does not work, I thought I’d go more in-depth.

As you all know, adding a rubber band tensioning system to a lift is very important to increase the amount of game objects the lift can carry, to reduce strain on the motors, and even to change the lift’s gear ratio to a faster one.

Most teams use this sort of rubber band system on their lift (4 bar, 6 bar, 8 bar, etc.):

This system is great due to complexity vs. effectiveness. Just by putting a couple of pieces of metal (in this case, standoffs and screws) and rubber bands, the stress on the motors is greatly reduced.

However, some of you may have come across some flaws with this sort of tensioning system. The rubber bands have a greater upward force when the lift is at the bottom, while the upward force starts to diminish while the lift rises. What this means is that as you lift upwards, the assistive force of the rubber bands start to disappear, as the motors begin to have to do more work. The lift then reaches a point where the rubber bands are completely slack and has no assistive force on the lift.

I created a graph to better illustrate the situation:

Obviously, the numerical values are not accurate as the rubber bands won’t be producing a 100N of force on the lift, but the concept is the same. The Graph also may not even be linear, but the point is that the force diminishes as the lift angle increases.

In addition to the fact that the rubber bands do not work effectively when the lift is raised, high upwards force near the bottom of the lift is also not ideal. With no game pieces in your robot, the lift may start to rise on its own because of the strong upwards force of the rubber bands. One may combat this problem by taking off rubber bands, further reducing the effectiveness of the rubber bands. Others may put a small constant motor value in the code ( while (lift is down) motor value = -10; ) but this is still not ideal.

An ideal graph would look like so:

Constant upwards force regardless of the position of the lift.

My discovery of the uniformly tensioned rubber band system (UTRB for the sake of the phrase’s length) made its way during the 2012 gateway season. While creating a pneumatic lift, I discovered that a UTRB was absolutely necessary due to the fact that pneumatic pistons are quite weak, but also have a constant force along its stroke (save for minimal amount of loss force due to the fact that air gets released every time you activate one). After the robots completion, the robot was able to lift 6 game pieces (3 pounds) to 30” and still be able to bring the lift down with 0 pieces (no weight) all while having a pneumatic lift. To put this in perspective, the pistons could not lift the manipulator up at all without any type of rubber band system. This should demonstrate the effectiveness of the UTRB.

How it works:

Let’s look at the UTRB when one’s lift is all the way at the ground.

If you look carefully, you can see that the rubber bands almost cross the pivot point of the lift.

This means that the force acting on the lift by the rubber bands is very small.
R (crossproduct) F = Torque
Or
R sin(theta) F = Torque.
When theta approaches 180, sin(theta) approaches 0.

CONTINUED…

10 Likes
2

Now when the lift is halfway up:

sin(theta) will be greater

When the lift is at the top:

Sin(theta) is the greatest here with a value of 1

Now coupled with the fact that the force of a rubber band decreases as the length decreases and the value of sin(theta) increases as the lift increases, what we get is a(n almost) uniform torque (upwards force acting on the lift).

Here is a video demonstrating the effectiveness of the lift:
https://www.youtube.com/watch?v=ertKM1E2ejE&spfreload=1

Interestingly enough, during this comp, I noticed that another team had come up with this type of tensioning system. One of 1437’s viewpoint robots had this tensioning system in the back of their robot. (If you look closely enough on the red side, you can see that the back of their robot has a bar jutting out of the back of their lift, which is the same UTRB, but just located in the back).

Now in 2013, John Ma and Kevin Li from WASABI used my tensioning system but in a much simpler way

And here is a video demonstrating their lift:
https://www.youtube.com/watch?v=yngtgmTHRRU

The picture, being the more updated version, demonstrates how they only used a couple pieces of metal rather than my original heavy structure (made with a bunch of straight bars and standoffs). Their structure reduces the weight of the structure and I believe this is the simplest it could become. (Use this structure is what im saying basically).

By the way, if anybody is interested, here’s a link of their reveal and pictures of their amazing robots which they go more in-depth to.
https://vexforum.com/t/1492-wasabi-post-worlds-reveal/23791/1

The amount of rubber bands one should use in my opinion should be enough to counter act the weight of the manipulator in addition to some game pieces. For instance, in our gateway robot, when all the air was released from the pneumatics, the lift would start to rise on its own due to the tensioning of the rubber bands. In the case of a motor lift, the idle torque of the motors should be the factor that holds the lift down. In this way, more game pieces could be lifted up (great for high load games like sack attack).

Granted, this tensioning system may not be exactly uniform, as the value of the force provided by the UTRB may waver, but I believe that the deviance is negligible. Some members may do some physics to figure out the exact point to mount the rubber bands to achieve the most uniform tension, but I think it is quite difficult as it requires maybe an upwards of 4-5 variables to solve for.

Heres some more pictures of the UTRB if anyone is interested:
http://imgur.com/a/47gnY

Sorry about the quality of the pictures, as they were uploaded to facebook, then uploaded to imgur for the purpose of this guide

Alright! That’s all there is to know about UTRB’s. Thanks for reading, and I hope this helps teams in their quest for awards or yearning for knowledge. If you have any questions, feel free to post below.

8 Likes
3

Great and informative right up. Definitely going to look at adding rubber bands to our lift.

4

Great explanation. We used something similar earlier in the year, using the save principals used in your pneumatic lift.

5

I’ve been interested in similar UTRB ideas for a while, but I’m looking for ways to measure the actual uniform effectiveness. It seems like the best way is too invasive.

  • replace lift motors with a sprocket and use a spring scale on the chain to measure torque for each lift height.

Can you think of an easier way to measure the actual uniformity?
Bonus points if the measurement is at the motor shaft.

I think of rubber band spring lift assist as an exchange of Potential Energy.
Lifting mass of the arm increases its PE, while the spring relaxes decreasing its PE.

6

A method one can use without removing he motor is to disengage the gears linked to the lift, adding a rotation sensor to the lift shaft and putting a force sensor at the front of your lift (where the manipulator and lift are linked). Then by pulling down slowly on the force sensor, one can get data points and plot theta vs. Force.

However, I think that this graph wouldn’t be very practical in the real world. If everything was frictionless, then you would get an accurate graph, but all the friction between the shafts and joints would smooth out the graph.

7

Paraphrasing: add cargo to the manipulator, disengage the motor, graph Y= the apparent weight of manipulator vs X as input lift axle angle.

OK, that should work. Then if you have two rubber band placements, and you measure them both, you can see which one has the flatter graph == more uniform tension.

For a rotation sensor, you can just mark angle on the input gear.
5-10 points measured manually across the height of the lift is enough to get an idea.