# Elevator Rigging Tensioning + Motor Load Confusion

I’m trying to pick a spring to tension the rigging on our continuous elevator, like I’ve seen many teams do before (254 and 973, at least). If anyone knows how I should select a spring, just write it and skip the rest of the post, where I explain what I’ve done so far and why I’m lost.

The factors I’m examining are primarily min load, max load, and rate, as McMaster calls them. When designing for how much the rope will “slack” (and safety factor), I need to know the max load the rigging (and thus the spring) will experience.

The load is dependent on the elevator’s acceleration. 0 acceleration and the load is the weight of the carriage. 9.81 m/s^2 of acceleration and the load is twice the weight of the carriage. So I need to determine the max acceleration the elevator will experience. When I use popular FRC calculators, they return two numbers that are “like” load.

The first is “Load Current Draw” or “Current Draw Per Motor”, depending on which one you use. When I run this back through motor constants (K_\tau), the number of motors, gearbox, and spool, I arrive at the input “Load” I gave the calculator, plus some extra which comes from efficiency factors, I believe. So I guess this number has been calculated the same way, but in reverse, which doesn’t help me. Per my above intuition about load on the spring, the load will be larger than the weight of the carriage.

The second is “Stall Drag Load” or “Stall Load”, but this number is far too high, and I believe it to simply be the load the motors will create at stall torque, accounting for the gearbox and spool. However, I don’t believe the elevator will ever stall. Also, this number is 200+ lbs, and springs that can withstand 100+ lbs are hard to find and expensive.

Thus, I am left with two questions:

1. How do I decide which spring to use for inline cable tensioning? Do I need to account for the weight of the second stage at all? Should I make sure the motors physically cannot break the springs (I was planning on solving this in code)? How much should the spring extend at the accelerations we want or at 0 acceleration?
2. How do can I calculate how fast a motorized mechanism will accelerate? Is the only answer to use a differential equation motor model (like in Austin Schuh’s System Modelling talk)?

3x 775pro, 14.12 : 1, 2" dia. spool, continuous rigging.,15-25ish lb carriage.

Unless using a constant force spring, I find it far more convenient to work with the distance the spring will stretch (delta from extended to retracted), and the energy it will need to store on that stretch. I have copied tables of springs from vendors (e.g. Lee Spring), copied these into spreadsheets, then calculated extra columns representing those values, and selecting from there. Edit: With the lift arm below, I did not actually worry about the load of accelerating and moving the load, merely providing the necessary change in potential energy from one end of the run to the other, because the intent was to bring the load to a stop at different altitudes. With the launcher, yes, it was necessary to calculate the kinetic energy of the ball and arm.

Edits, now that I’m reading the rest of the post: If you need large spring forces, the easiest way to do this is to gang multiple springs together in parallel. Not only does this save money on the initial purchase, it is far more scalable - if you build the thing and find it’s pulling a little to hard or a little too light, just remove or add another spring! (Be sure to design your gang mechanism for the largest value you expect.) 3946 used this technique in 2016 Stronghold for the catapult (wound up with about 16 or 18 springs) and in 2018 Power Up as a counterspring for the arm (12 or 14 springs).

With my method, yes. The first stage will rise half way, the second stage will rise the full distance. Note that constant force springs work better with continuous cascading elevators, and normal springs with cascades continuous, because at the high end less force is needed to continue the lift.

Absolutely. Unless you’ve done this enough times to really know what you’re doing (and if you have, why are you asking here?), remember that you only have a few weeks. Don’t depend on software for the integrity of your robot.

Not sure I understand the question. If I do understand, I never plan to exceed (or preferably reach) the high-tension end of a spring, and find that having the low-tension end at a force between 1/2 and 1/3 of the high tension end. With a cascade, you may be able to take that down to 1/4.

F=ma. F includes the spring and the motor torque divided by the moment arm.

For FIRST

, I’ve usually found doing close order of magnitude calculations (+/- 25%) is as close as I’ll get. There are too many unknowns to get it any better. As such, it’s more expedient to do Q&D math, build, and iterate than to spend a lot of time on the math for kinematics. (And I’m the guy who figured out the integral to determine the shape of the scoosh.)

2 Likes

Here’s an example of what I mean by “ganging” springs. This isthe power point sketch of the risers for Koopa Troopa (3946’s 2018 competition robot). [Further edit: As I look at it again, I realize that this is not what we competed with, but a later iteration we never completely built. The competition version was similar in terms of the ganged springs, however.] The views at either end are from outside left and right of the robot, and those just inboard are inboard views of the left and right risers, from the centerline. The thing in the middle is a portion of the arm. There is a bank of 6 springs outboard of each riser which serve as counter-load to the main arm, which pivots on the arm about 80% of the way up the side channels. There is a 1/4" bolt running through the loops at the ends of the springs. The bottom one is bracketed by a short length (3/4"?) of 3x1" c-channel, and the top one is free floating, with a length of 550 paracord coming of the middle of it. There are also various spacers/washers on those bolts keeping the springs from getting entangled in each other.

For 2016, the ganging was accomplished with a couple of VersaFrame angle pieces. I don’t think I have any drawings. One was mounted on the chassis, another on the arm, parallel to each other. The springs connected to the hole on the chassis side VF angle and the matching hole on the VF angle on the arm side. The arm was on the centerline of the robot; we just had to keep the loads balanced left-to-right as we added and removed springs.