View Full Version : How to create bending motion?

Mike C
05-23-2005, 11:23 AM
Hi All, http://forums.parallax.com/images/smilies/blush.gif

I have a BS2 Homework Board with a couple continous rotation servos and standard servos from parallax. I wanted to create movement in a way that bends a tube or pipe...I thought about using pulleys but I'm not a mechanical engineer, thus it is way over my head.

In short, the project will be using a tube with a photoresistor at the tip to find the darkest/lightest area and the tube would bend towards that direction.


05-23-2005, 07:45 PM
Like a robotic arm with a tube attached to it?

05-23-2005, 07:45 PM
Do you have to have the "bending motion". Can you just have this tube with the photo-resistor rotate ?

Have 2 motors: one that rotates the tube horizontal to the ground and the other to raise it up and down.

Pretty much a cylindrical coordinate system.

Mike C
05-23-2005, 11:28 PM
I would prefer having the bending motion....
kinda like a tube bending like a fiber optic cable http://forums.parallax.com/images/smilies/tongue.gif

But Boris, how would i move the tube up and down?

Thanks again.

05-23-2005, 11:50 PM
Of the top of my head I can come up with two different setups.

1) Have the horizontal motion motor with a plate attached to it. Then on that plate position another motor (Stepper or Servo)·with its axis being parallel to the ground.


2) Same horizontal motor setup but have a worm drive type rod pusing up/down on the tube for the up/down movement. The worm drive rod can be positioned closer or further away from the axis of rotation of the tube in order to control rates· and range of motion.

······ ·······/
··········· / |
········· /·· |
········ /····|· /\
······ /····· |·
···· O······ |· \/

There is my crappy drawing for the 2nd setup. Hope it stays the same once i submit the post.

kelvin james
05-24-2005, 12:10 AM
If you are using flexible tubing that does not require a lot of force to bend it, this would be fairly simple with a standard servo. One end of the tube would be fixed and the other end would be able move / slide to accomadate the difference of the bending. The motor horn ( you could just screw on a small shaft to horn to lift the tube) would be centered under the length of tube to provide the bending motion.
You may also want to consider physically attatching the servo horn to the tube with a ring, as depending on the type of material the tube is made of, it may want to retain its' " bent position", thus not wanting to lay straight again.


Mike C
05-24-2005, 04:42 AM
Do you mean to have one motor take care of one rotation and another motor to rotate the 1st rotor which rotates in another direction?

I thought about stacking one motor to turn another motor with the tube attached to the top motor. However, the movement would be based on the base of the tube and really would not be the same kind of movement I'm hoping for. But it would be able to cover an area of a hemisphere.

I was going to use plastic straws or find something else from a hardware store that would better retain its shape after several bends.


This is what I'll try - One motor would be attached to a 2nd motor (probably need to find something to attach them) and then the 2nd motor would be attached with the tube. Gonna try to see if that works first.


P.S. This is an aside but I was wondering if Parallax sells a housing unit for the PING)))) sensor to attach it to its servos?

Thanks for all your suggestions and inputs :)

05-24-2005, 10:06 AM
Typically tubing is bent over a mandrel to avoid kinking.· To get an idea of what it looks like, go to www.mcmaster.com (http://www.mcmaster.com) and search "tube bender".· You'll get three pages.·

The forces required to bend tube vary significantly depending upon diameter, wall thickness, material, etc. as you can well imagine.· Small diameter soft copper tubing bends fairly easily and retains its shape well.· Hand held tube benders for copper are inexpensive and I imagine could be modified fairly simply to be automated.

Chris I.

Mike C
05-24-2005, 10:49 AM

Thanks for ur response, but I think that the tube bender u are showing me is too big for what I was imagining.

Ken Gracey
05-24-2005, 11:59 AM
Mike C:

Yes, a servo-mountable PING)))/HM55B Compass mount is in design and will be in production in a week.

Ken Gracey
Parallax, Inc.

Tom Walker
05-24-2005, 08:25 PM

My family and I are supposed to be going on vacation around the end of June and I would love to get some 'bot experimentation time in (my wife's not aware of that fact yet <g>), and this would be very nifty if it's not too expensive...see previous reference to "vacation"...<bg>

Truly Understand the Fundamentals and the Path will be so much easier...

Paul Baker
05-25-2005, 12:58 AM
Ive thought about this a couple of days, and heres my best idea: use a flexible structure which has some amount self stiffness, the closest thing I can think of is the spring tube benders on the mcmaster-carr website (easily bendable but not compressable). Solder eyelets every inch or so along the tube one in each cardinal directions like this:

/ \
O| |O

Do this roughly every inch along the tubing, tie a cable to the eyelet on the flexing end of the tube, thread the cable through the remaining eyelets of that side around a servo driven pulley and back up through the opposite side eyelets and tied off to the final eyelet. Do the same with the other two directions, so that you have one cable/pulley controlling the east west direction and another cable/pulley controlling the north south direction. The cables should be tight enough to prevent gravity from acting upon the tubing. To make the tubing flex upward you would turn the servo/pulley so the northern cable tightens and the southern cable loosens, this will cause the spring tubing to flex upwards.

Mike C
05-25-2005, 08:50 AM
Hi Paul,

Thanks for your explanation. I checked out the spring tube bender, it sounds like a good idea. I want to just reiterate what you said to make sure I am picturing it right. So for north and south movement - I'd have one string from north end going into the tube from the botttom to the top in increments of ~an inch, then end with at the bottom end of the south side. Same method with the west-east movement.

My question would be how to use the string to tighten on the tube so that it won't bend to gravitational pull.

Thanks again.

05-25-2005, 09:27 AM
Actually, the cables through eyelets is a very elegant solution, it can even be inside the flower "stem" so you can't see it. Now all you need to do is use CDS cells as "petals" to see where the strongest light source is coming from and turn the servos to point till the center of the "flower" receives the most light!

"Never create anything you can't control"
"The amount of intelligence on the planet is fixed... the population is growing"

Paul Baker
05-25-2005, 10:05 AM
In this instance a picture best describes it so I attached one. The black curved tube is the spring tube, the green loops are the cable guides. There will be a limit to how long you make the tube, the longer it is the higher the tension required to keep it stationary. The servo will naturally resist the tension in the cable due to weight, but there is a limit to what it can do so youll have to experiment with length.

Post Edited (Paul Baker) : 5/25/2005 3:16:36 AM GMT

Mike C
05-25-2005, 11:25 AM
very very pretty. Thanks for the visual aid Paul :)
The layout of the pulleys/motor is also better than i pictured.
I'll get my hands on the spring tube bender and experiment with the layout a little more.
Thanks for your effort and wisdom!