#aircraftcable

(I’ve been meaning to post this to the web forever.  Sorry for taking so long!)

Last year, we created this sculpture called Breaking Wave, a collaboration with Jeff Lieberman from Plebian Design.  Check it out:

http://www.hypersonic.cc/projects/breakingwave

A lot of work went into conceiving of the piece, designing it, engineering it, testing it, and installing it.  We have tons of documentation of that process, although we haven’t published much to the web.  This is step towards getting some of that out there. 

This blog post is about one specific little thing with the Breaking Wave sculpture: aircraft cables that failed, specifically by tangling up and “birds nesting.”   A birds nest looks like this (just off that center pulley, opposite the sharpie’d arrow):

wanna see another one?  Ohh, it pains me just to look at them. (right there, in the middle of that winch drum)

How did this happen?  After the sculpture was installed and running, we started encountering problems with the 36 drive cables (.036″ dia, 7x19, nylon coated SS pre-stretched), where we would find “birds nests” after several thousand cycles.  As far as I knew to that point, we weren’t violating any engineering rules on using these cable, our safety factors and bend radii were all well over-margin. I did a lot of research, but I never found much of an answer either on the web or in book.

Here’s how it happened:  our drive cables are fixed at both ends - there is no swivel or ability for the cable to rotate.  That can be ok, but not in this case.  Our drive cables also went around 2 pulleys.   In the path around the second pulley, the drive cable comes off at a slight angle of somewhere between 0 and 5 degrees, that’s where the problem is.  (this drawing shows the drive cable path)

When you have a wire rope that cannot spin at either end, passing over a pulley where the wire rope is not perfectly parallel with the line of action of the pulley, it’s bad.   Wire rope has a “lay”, this is the direction that it is internally twisted when it’s made.  As the wire rope passes over a pulley with a slight angle, the wire rope tends to slightly climb up one side of the pulley, and then rolls back down - because of the “lay” that’s already in the cable.  This makes the cable twist even more, and there’s no where for that twist to go since the cable cannot swivel at either end - so it turns into a birds nest.  

How did we solve it?

We built a cycle test machine in the studio that replicated the drive system of the sculpture.  We set it up to run 10x faster than the actual sculpture so that we could test solutions quickly.  The first thing we tested is the actual cables to verify the problem.  It happened as expected on the test rig.  Then we tested a bunch of different cable solutions.  

We tested a lot of different types of synthetic cable replacements:  vectran, dacron, spectra, dyneema, poly, aramid...everything.  Here’s what we found:

Dacron is really high stretch, it’s not a even close to a decent replacement.  I’m surprised people even recommend it.

Spectra is really low stretch and quite strong, but because of it’s low stretch, it breaks easily under and impulse load.  As well, all Spectra has to have a nylong (or other) sheath around it because of UV.  We found that the nylon sheath doesn’t stand up to repeated cycles over pulleys.

Vectran is pretty good stuff.  Pricey and hard to find, but it did well in strength, stretch, and wear resistance.  2nd place.

Dyneema was the all around winner.  It has great strength, relatively low stretch (although it does have an initial break-in stretch), and excellent wear resistance.  I highly recommend it.  

We replaced all of these drive cables with 1mm dyneema (synthetic) rope.  Dyneema is braided, so that there isn’t a predominant twist in either wrapping direction, so it doesn’t try to twist.  

So, why am I posting all this?  I searched the internet for days wondering what was happening when we first encountered this problem.  My hope is that maybe if someone else is facing a similar issue, they might stumble on this post and it could help them.  Hopefully I got enough key words in here to search for.