DrJB Posted February 22, 2014 Posted February 22, 2014 (edited) This appears to be a puzzle and I have seen few contraptions (or proposals for new parts) that various people have attempted. Picture this: You want to transfer an air-line (or electric wire) between the bottom of a crawler (track-drive) to the top superstructure ... and you want to do this without twisting the air-hose ... Most will say 'impossible'. Well, not quite (and hence the reason behind this thread). This is a classical problem in mathematical topology, and was resolved a while back. In the picture to the left, the wire will twist if the blue part rotates. In contrast, in the right figure, the air-hose will not twist. It will however bend back and forth once per revolution and is thus a good idea to make the radius of curvature a bit large so as not to damage the wiring after multiple cycles. Edited February 22, 2014 by DrJB Quote
efferman Posted February 22, 2014 Posted February 22, 2014 (edited) i am working in real live with hydraulik hoses and this doesnt work. take a hose and put two T pieces in the end of the hose. now do what you postet above and see what happens. the hose will twist, your method let the hose only let later die Edited February 22, 2014 by efferman Quote
Moz Posted February 22, 2014 Posted February 22, 2014 Will definitely twist up the cable that runs through it. This is exactly how rope makers work - and there are Technic examples around if you want to look at them. The easiest way would just be to make one and see what happens. Quote
DrJB Posted February 22, 2014 Author Posted February 22, 2014 (edited) Well ... I tried it, and it does work ... though not with Lego yet. I described that such problem has a solution (mathematically). Whether the solution can be implemented with regular lego parts ... that requires some skills. Will definitely twist up the cable that runs through it. This is exactly how rope makers work - and there are Technic examples around if you want to look at them. The easiest way would just be to make one and see what happens. Will absolutely/definitely/positively NOT twist the cable ... You should try it, it's one of those not-so-trivial solutions. The trick is that the cable has to be soft in bending, so that it can bend in a circular fashion repeatedly back and forth, without breaking. Try it first, you'll be surprised! Edited February 22, 2014 by DrJB Quote
Darth Dino Posted February 22, 2014 Posted February 22, 2014 Hi i worked in real life with hydraulic hoses. So i am a supporter for "IT WILL TWIST". If you just solved your problem mathematical, then you did something wrong in the formula. There are special couplings for pneumatic, hydraulic and electric hoses/wires to support rotation. Dino Quote
Heppeng Posted February 22, 2014 Posted February 22, 2014 Agree, if it was that simple that's what would be used in real life. People have tried this and it does not work! Quote
Saberwing40k Posted February 22, 2014 Posted February 22, 2014 This method does not work, but I have found one that does. It only uses 100% Lego parts, but it could potentially damage them. It involves connecting 2 m motors output to output, through a turntable. ! motor drives the other, making it act like a generator, and thus supplying power to the superstructure. Quote
Alasdair Ryan Posted February 23, 2014 Posted February 23, 2014 This method does not work, but I have found one that does. It only uses 100% Lego parts, but it could potentially damage them. It involves connecting 2 m motors output to output, through a turntable. ! motor drives the other, making it act like a generator, and thus supplying power to the superstructure. What about using the E motor/generator from the energy sets? Quote
DrJB Posted February 23, 2014 Author Posted February 23, 2014 This method does not work, but I have found one that does. It only uses 100% Lego parts, but it could potentially damage them. It involves connecting 2 m motors output to output, through a turntable. ! motor drives the other, making it act like a generator, and thus supplying power to the superstructure. Again, we have a lot of opinions, based purely on 'gut' feeling ... but no absolute proof ... That's ok. Quote
Saberwing40k Posted February 23, 2014 Posted February 23, 2014 What about using the E motor/generator from the energy sets? I don't have any, but that's a great idea. Again, we have a lot of opinions, based purely on 'gut' feeling ... but no absolute proof ... That's ok. Oh, I have absolute proof. I have built this, and it does work, I'm just not able to film it at this point in time. Quote
Alasdair Ryan Posted February 23, 2014 Posted February 23, 2014 I don't have any, but that's a great idea. I guess that's me showing off my slightly clever side. Quote
DrJB Posted February 23, 2014 Author Posted February 23, 2014 Oh, I have absolute proof. I have built this, and it does work, I'm just not able to film it at this point in time. LOL ... my comment was about people 'bashing' my rotating hose connection ... not your motor-generator combination ... lol Quote
Moz Posted February 23, 2014 Posted February 23, 2014 (edited) The trick is that the cable has to be soft in bending, so that it can bend in a circular fashion repeatedly back and forth, without breaking. Try it first, you'll be surprised! OK, I just grabbed a turntable and some ribbon cable (so it's easier to see it twisting), and made a little video. Can you point out what I did wrong? Edited February 23, 2014 by Moz Quote
DrJB Posted February 23, 2014 Author Posted February 23, 2014 (edited) Great video, I think we're getting close ... If you look carefully, you'll see that the flat wire twists one way from the bottom to the top (long segment), and then twists the other way from the top to the rotating turntable. The 'weakness' in your implementation is that a flat cable does not lend itself very well to circular bending. Try it again, but with a circular wire and you can put some white marking ALONG the length of the cable ... Make sure the wire though is soft in bending. Edited February 23, 2014 by DrJB Quote
Moz Posted February 23, 2014 Posted February 23, 2014 (edited) ok, sorry about my hands in the way but I had to twist this one about 3 full rotations to get it as twisted as it is. I've used very flexible electrical cable that's about the same size as pneumatic tubing, since I have two different colours available. This stuff is very flexible silicone-insulated stuff, I can tie a thumb knot in it that's ~15mm at the widest point. I tried to keep the top section untwisted so it's obvious that all the twisting is happening on one side of the bend that pokes up. I still don't know how this is supposed to work, but I'm willing to upload another video so you can explain what's supposed to happen and exactly what I'm doing wrong to get the hideous mess that I and other expect. Are you expecting the single pneumatic tube to slip around on one end of the attachment so that rather than twisting it has effectively got the slip ring that Efferman talks about? Edited February 23, 2014 by Moz Quote
EdmanZA Posted February 23, 2014 Posted February 23, 2014 This is a classical problem in mathematical topology, and was resolved a while back. I'm intrigued by the maths of this. Would you be able to post a link to somewhere where I can do some more reading about it? The other thing I'm not sure about is whether the ends of the tube are assume to be fixed or can rotate freely. Quote
skppo Posted February 23, 2014 Posted February 23, 2014 You ask for proof but provide none yourself. Tried it, did not work. The hose has twists in it. Only way I can see this working is with a slip ring. This is topologically equivalent to a hose with no bend. It just has a bit more slack. If you think like this: mark a spot on the inner side of the hose on the rotating part. Then mark the same spot on the stationary part of the hose. As the hose goes around the spot will always point at the center. If the hose wouldn't twist the other spot should always face the spot on the rotating part it. It however does not, as it does not rotate with the rest. Quote
Moz Posted February 23, 2014 Posted February 23, 2014 The thing I'm not sure about is whether the ends of the tube are assume to be fixed or can rotate freely. One end is fixed, the other rotates. Imagine you have a battery under the turntable and a motor above it, and you want to connect them. In the first video I used coloured ribbon cable so it's easier to trace it through the turntable, but Lego motors use ribbon cable too, so it's a perfect example. I'm also more willing to twist a short length of $3/m ribbon cable than a $10 Lego electrical extension cable. If DrJB can't make a video of his demonstration, or explain why my demonstration doesn't match the situation he has at his end, I think we might have to give up and just accept that what everyone else experiences is definitive. FWIW, I'm using my cellphone and out $10 worth of Lego on my living room floor above, so I do wonder why DrJB doesn't do something similar and just produce a video of his own. Ideally he'd motorise it, so we could see a battery box underneath powering an M motor spinning itself round on top. Quote
EdmanZA Posted February 23, 2014 Posted February 23, 2014 One end is fixed, the other rotates. That's what I assumed - one end fixed relative to ground, the other fixed relative to the rotating part i.e. the ends can't swivel at their fixing points. I'm just wondering what assumptions that maths makes regarding these points. Assuming the maths is right, I find it very interesting seeing how the real world can violate mathematically ideal assumptions. Quote
NORAN Posted February 23, 2014 Posted February 23, 2014 Hi, if you need to pass a cable or hose from a static to a rotating surface, the best way is a slip ring, whether electrical, mechanical, HYD or FO. When I was riding my ROV had to use 2, one in the cage for theter and in the umbilical winch, follows a picture of the cage and slipring video running in the first test system has a signal loss but to the improving quality and making it easy. https://www.flickr.com/photos/118485053@N02/sets/72157641410773064/ There is another way of winding and unwinding a cable without the use of slipring, it is used in a satellite SEAMAXX ROV works somewhat like an inverted drum, the cable is played inside the drum and pulley is holding tension the tether. Quote
Boxerlego Posted February 24, 2014 Posted February 24, 2014 Try using a ribbed hose connected with a friction-less pin to the rotating platform. Quote
Rockbrick Posted February 24, 2014 Posted February 24, 2014 Like this: My Brain just exploded not in a dissimilar way to the eversion of the sphere.... Quote
Bruce Headley Posted December 26, 2017 Posted December 26, 2017 I was playing around with construction of a 'radar', using Lego Mindstorms NXT bricks. It is simply a Tach Motor coupled to the Ultrasonic Sensor, that twists back and forth 180 degrees. The problem was with the supplied cables .. they are very stiff and kept getting hung up on everything and anything nearby, plus kept flopping in front of the sensors path. So, falling back on my years working in aerospace, I decided to make a rudimentary version of a 'twist capsule' to help with the problem. Basically, it's a fairly long length of flat ribbon cable (like a MindSensors 3m Flexi Cable), connectors crimped on both sides, then rolled up like a watch spring. The cable exits the center of the coil at a right angle and from the outer edge. A pancake coil. I cut out a couple of stiff cardboard discs and punched holes in the centers to act as retainers. To assemble: A long cross shaft is pressed into the Tach Motor hub, sticking upwards. One of the discs slides over the shaft. The cross shaft pokes through the center of the pancake coil, along with the center cable & connector. Enough cable needs to be pulled out to connect to the sensor. The other disc slides over the shaft. This helps with keeping the pancake coil in shape. The Ultrasonic Sensor (with a few additional brick pieces) attaches to the end of the cross shaft. The outer coil cable was loosely attached (tie wrap or bread tie) to one of the mounting holes on the Tach Motor. It worked! The pancake coil wound and unwound just like a watch spring. I also found out there was enough free 'twist' to allow for scanning in 360 degrees without any issues. Granted, this does not allow for continuous rotation, but for my purposes it works extremely well ! The other option in mind was to point the Ultrasonic Sensor (or the IR Sensor) upwards and construct a motor/reflector system, but ... Hope this idea helps ! Quote
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.