Breaking Cable

Our team built a winch with spectra cable to lift our robot this year. I'm wondering how other teams determine what strength cable to use. Here's what we did: I figured that the greatest weight our robot would need to lift is 260 pounds (our own robot plus one trying to pull us down). So we ordered 468-pound test.

To make sure this was adequate, we used the line to lift a student who weighed over 200 pounds. Lifting was no problem, but a little bouncing would break the line. Just to be safe, we decided to go with the 876-pound test cable instead. Since an active competition might involve more than just lifting dead weight, I'm wondering if there's a good way to calculate the correct cable size, rather than just testing it like we did...

[patrickrd]

You have to think about the when the maximum force is applied to the cable. It might not be while the robot is lifting off the ground. It might be when it is in the air, and impacted by another robot. If there are other robots next to you while lifting, you might have to carry a large amount of the weight of other robots in addition to yours. I would take the worst case of all these considerations (i'm sure there are others) and use a safety factor of two. So something 600-800 lbs would probably be safe.

Another idea: the motor can only provide so much force to the string. You might multiply the stall torque of the motor by the radius of the winching mechanism. [EDIT]stall torque divided by radius, not multiplied[/EDIT]. This is a pretty good estimate of the maximum force that CAN be applied to the cable... However, impacts can instantaneously put larger loads on the cable.

Patrick

[team222badbrad]

Um yeah we accually broke cable during testing in the 6 week build time. We were lifting our bot up to hang and at the time we didnt have any limit switches on the lift mechanism. This caused me to overdrive the arm and SNAP, Then .25 seconds later the robot was on the ground....

And yes we had 2 drill motors lifting us up... Once during a regional we were in a hurry and time was running out. There was very little space on the bar and we were lifting up our robot and all the sudden I see team 316 on the ground.

Then after the match was over we found out the score would have been a tie if they wouldnt have came off.. 130 to 80 eek

Sorry 316 we didnt mean to take you off the bar!

Not sure of the exact cable size we used but I thinks its over 1000 pound test.
3/16? In other words our lifting mechanism is stronger than last years drive train....

[ChrisH]

Originally Posted by WebWader125

To make sure this was adequate, we used the line to lift a student who weighed over 200 pounds. Lifting was no problem, but a little bouncing would break the line. Just to be safe, we decided to go with the 876-pound test cable instead. Since an active competition might involve more than just lifting dead weight, I'm wondering if there's a good way to calculate the correct cable size, rather than just testing it like we did...

Generally static loads, those that do not change with time (like a hanging robot), are much easier to calculate than dynamic loads, that do change with time (like a bouncing robot). Impacts and bouncing are among the hardest to determine.

Often a simple test such as you conducted is the quickest and easiest method for determining how to handle a dynamic load. Another method is to take the static load and use a safety factor of 4. That will generally cover FIRST robots for all but the most severe dynamic effects. That's what I do and I haven't had anything break yet.

BTW kudos for testing the bouncing condition. Most people wouldn't even think about "bounce", let alone test it. At least not until after something broke.

[Gary Dillard]

Here's a couple other things to consider:

The 260# you mentioned is the vertical component of 2 robots static weight. If you are pulling at an angle, you will need a higher load to get that amount (divide 260 by the cosine of the angle).

That load is also the tension in the cable between the hook and your robot. If the cable passes over a sheave or pulley anywhere then you will have an additional friction load in the cable leading to your winch.

The "bouncing" load is almost entirely a function of the stiffness of the cable - this is an energy balance where .5mv^2 input equals .5kx^2 stored in the cable. A cable without much flex (such as spectra, which creeps over time but doesn't really stretch statically) will take very high loads dynamically versus something like sailing cord that we used. It's hard to quantify but it can be done - physically measure the spring rate of the cord by measuring the deflection under load, assume some distance of drop, calculate the velocity at that distance and you'll find the g-load by velocity*sqrt(k/m)/g.

You know, you gotta admit that watching any match where the cord broke and a robot fell to the floor with a crash was pretty exciting - kinda like going to NASCAR and waiting for a wreck. Lots of moans and groans, but I was impressed with how many robots came right back after a big fall. Some pretty resilient designs.

[dubious elise]

Originally Posted by WebWader125

I figured that the greatest weight our robot would need to lift is 260 pounds (our own robot plus one trying to pull us down). So we ordered 468-pound test.

To make sure this was adequate, we used the line to lift a student who weighed over 200 pounds. Lifting was no problem, but a little bouncing would break the line.

My team also employed the use of SPECTRA with about a 400 lb test. We ordered 100ft and divided it three ways so that I could braid all of it into one very long and extremely durable line. We had a similar fear that the cable might break, so we purchased another 100ft to duplicate the process. Luckily, the original line lasted through two regionals, nationals, and a number of demonstrations.

My assumption is that braiding the spectra, when possible can only add strength. Even 200 lbs can create quite a force when bouncing on just one line. I strongly suggest spreading the weight to two or more lines if you can, just to be safe!

[Max Lobovsky]

Well, i think everyone has been assuming that "test" means the break strength of a cable or rope, but i am almost definitley sure that it already includes a safety factor. I think a certain test is actually expected to "carry" that load in some application.

[Matt D]

We used either 1200 or 1300 lb test cable. We also had trouble with our cable snapping. After two regionals, we determined that it was friction (and a 90 degree forced bend) that was wearing down the cable which caused it to snap. It just goes to show that for all the calculations and safety margin, little dynamic things can mess you up (they messed us up). We tried spectra in some testing before nationals, but we had trouble affixing it to our hook (the sharp edges of the hook wore it out so it would still snap) so we went back to cable.

[Adam Y.]

We tried spectra in some testing before nationals, but we had trouble affixing it to our hook (the sharp edges of the hook wore it out so it would still snap) so we went back to cable.

From what I have read though Spectra has the same problems. If it wears out it will break.

[Andy A.]

I've heard a few cases of teams breaking Spectra lines that, on paper, they should not have been able to break. This leads me to belive that Spectra has some quirks in how it handles 'bounce'. My guess is that it's just plain to stiff and doesn't stretch nearly enough to handle much in the way of sudden high loads.

I would have opted for some climbing line in a lift. It is heavier, but has a good deal more stretch and is designed to deal with very high dynamic loads (falling climbers). So, you just match the static load test to the wieght you need to lift and you should not have to worry about bouncing. Also, it deals with abrasion very well.

*shrug* I still think it was more fun to stay on the ground then deal with that silly bar.

-Andy A.

[Bcahn836]

we got some steel cable that one of our mentors uses for his prosthetic arms. He has a 2000 ft roll of it and let us use about 20 ft of it. However we didn't use it to hang from the bar but to open and close our arm to grab the 2X ball, and we havn't broken it yet. He claims that he has broken the pieces on his arms lifting an extreme amount.

[BillyGoats]

he have used cable for the past three years

2002 - held our claw... we broke this a lot....
http://www.valleytech.k12.ma.us/robo...002%20Claw.jpg

2003 - push pull system for out arms.. never broke it never became tagled! great stuff
http://www.valleytech.k12.ma.us/robo...botonField.JPG

2004 - holds us up...
starts - http://www.valleytech.k12.ma.us/robo...UCFReg3300.JPG
on bar - http://www.valleytech.k12.ma.us/robo...UCFReg5400.JPG
can hang this high - http://www.valleytech.k12.ma.us/robo..._UCFReg112.jpg

we had held our self plus another robot many times! great stuff!

I appreciate all of the comments. What I'm really wondering is, why does cable (spectra, fishing cable, etc.) quote a test strength in pounds, rather than foot-pounds? Obviously the same weight bouncing on the cable will break it easier than a static load. And it's the dynamic breaking strength that I'm really interested in.

In our tests, we also noticed that the cable tended to break where it was tied to something else.

[Bcahn836]

I think it all depends on how the cable is tested and rated. Like you said it might hold a certain weight still but not bouncing. In our test we stretched some nylon cable across a jig we made and but weights on it. The cable was rated at 150 lbs but broke at 60 lbs. So as i mentioned before we used some steel cable that we haven't been able to break so far. I think the motors would go before the cable. LOL