need rollers that conform to cable

[Anonymous]

I have 2 rollers with a cable squeezed between them. They will drive the cable up and down under load.  The cable is a coiled wire, like a plumber's drain auger snake.  The roller grooves need to be shaped to conform to the cable's outer surface to prevent slippage.  

Attached is a screen shot of the arrangement.  I tried cutting the groove in the rollers using a cable formed into a circle, but the cable shape changes as it's bent in that scenario, where in real life it's straight as it goes thru the rollers.  Any thoughts on how to imprint the cable shape into the roller grooves?

[davebYYPCU]

Not a simple job, research the helical gear, 

circular pattern of one tooth, dependant on how accurate / strong it has to be.

Might help.....

[etfrench]

@Beyondforce did a tutorial on creating a worm gear which may show you the way to create your cable pulleys.

[chrisplyler]

May I humbly suggest something along these lines? If you attempt full engagement all the way around, either the roller or the cable will be damaged during engagement/disengagement of the "teeth."

[MichaelT_123]

Hi Mr Chris,

The cable you described can be considered as a screw and as such many known mechanism ideas can be applied here.

One of possible driver idea is represented by the bellow sketch.

Any way if you want to check the configuration you presented, make rollers from a dense patty/elastomer and run your cable between them. The resulting pattern could give you some additional fuel for thoughts.

Regards

MichaelT

[Anonymous]

Original poster here.  Thank you all for your replies.  

a worm gear is less favored than a roller due to friction and speed limitations....I want to move the cable 12 ft in 5 seconds or so.  If the roller idea doesn’t work out, I may go back to worm gear idea.

Replyers rightfully pointed out roller gear engagement/disengagement concerns.  I really like the idea about using a “forming plastic” on the roller and rolling the cable through to get an impression.  That would inherently form roller teeth with proper engage/disengage angles .  I plan to use epoxy putty and try that out.  Thanks again, good luck to all.

[TheCADWhisperer]

Can you File>Export and then Attach your *.f3d file here?

[Anonymous]

here's the drawing.  the cable is 13/32" diameter, and the wire is 3mm diameter.

[MichaelT_123 · Accepted solution]

Hi Mr. Fisher

Here is a driving roll :

• module 1.23472
• pitch radius 44.45 mm
• number of teeth - 36

It is almost ready for casting in a tough plumber grade urethane elastomer.

How to make/design a new one?...

• select gear modulus (pitch diameter/number of teeth) as per cable pitch and desired reel size
• cut cable body into roll cylinder
• having the first cut,  create circular pattern taking number of teeth into consideration
• intersect the resulting bodies using combine tool. 

Regards

MichaelT

[Anonymous]

Wow, wow WOW!  I am astounded at the skill of the people on this forum.

MichaelT_123, thank you for your reply.  Just amazing.  I will attempt to design one given your instructions, but being a newbie I'll probably stumble. 

A couple questions:

Your roller design has lots of little facets for each "cog":  do these provide the clearance for smooth engagement/disengagement ?

I assume this roller gear is equally effective (smooth operation and "gripping strength") turning in either direction...correct?   i.e., no bias in the design as to rotation direction

[Anonymous]

I was able to duplicate Michael T’s solution.  This solution cuts teeth into the roller based on cutting the roller each time the cable comes tangent with a roller tooth, I.e, my roller has 36 teeth, so there are 36 cuts made to form the roller teeth.  So each tooth is cut with the cable tangent to that tooth, then again as the cable is set tangent to the other teeth around that tooth.  This produces little facets in each tooth as the cable is set in increments of +/-10 degrees from tangent with that tooth to cut the teeth surrounding that tooth (36 teeth, each one is 10 degrees from the next). Sorry for over explaining.

In real life, the cable will continuously move from tangent with tooth 1, then on to tangent w tooth 2, etc....the little facets that are produced in the roller teeth from the above solution will actually be more of a smooth curve.  Is there a way in Fusion 360 to get more cuts than just every 10 degrees to smooth out these facets?

[TheCADWhisperer]

Can you File>Export and then Attach your *.f3d attempt here?

[Anonymous]

File of Roller attached. Appreciate any help.

Incidentally, Im 3d printing the roller now and the small facets dont seem to be significant.  This is a small roller and such little details kind of get lost.  Im thinking it may not be a problem, but just for my edification would still like to know if there is a smoothing function or some other way to solve this issue. 

[MichaelT_123]

Hi Mr. Fisher,

... these little facets are the result of building up the teeth profile by cutting the cable into reel while rotating it by a certain degree. Selecting smaller rotation steps will result in more, but smaller ripples. As per my intuitive estimation height of ripple is a function  ~wireRadius*cablePitch*sin(angleRotationStep), so an increasing number of cut-rotation steps will quickly decrease the wavelets' height, but also increase the number of facets. I think it would be reasonable to select the step's quantity equal to the number of teeth on the wheel. It would be nothing wrong to double/quadruple/... it, however successive operations on the model (milling, 3D printing, etc.) might suffer performance issue. Anyway, probably in the application in hand, wrinkles will be smoothed out in no time. 

The analytical/mathematical formula describing tooth surface/profile could be probably calculated ( with some efforts though), but for now, the described approximation should suffice.

It would be also worth to consider forces vs materials vs design calculations beforehand.

I personally debt that the single tooth strength can withstand operation you briefly described.

Why not try simulating the design in SIMULATION environment of F360, 

as it is, by all means, the interesting example of:

"Simulation of a springy long body interacting conformally with a squishy, curvy and continually rotating one"

Consider checking universities around, they might be interested to endow such Ph.D. dissertation? 🙂

If not, movie studios might be interested!

Regarding:

"I assume this roller gear is equally effective (smooth operation and "gripping strength") turning in either direction...correct?"  

Yes, personally it would be my choice. Yes, it is bidirectional/symmetric by design with forces nicely distributed and offering limitless design variations. Such design concept functions in many technical corners of technology.

With Regards

MichaelT

Now seriously, about the simulation!

It would be a valuable educational lesson if somebody with some time in hand takes on the challenge!

[Anonymous]

Michael T, you’re a funny fellow!  And a smart son of a gun too.

im gonna try the rollers first.  If they can’t handle the load, I’ll consider worm gears that are larger diameter than the cable in order to minimize worm gear revolutions per length of cable travel.  The thread pattern on the worm gears will need to have multi-starts, which could get tricky, but it will decrease the rpms needed for required travel speed.  Probably use 2 worm gears with some sort of cable guide to keep the cable between the gears.  Easier said than done.