Three Section Telescopic Driveshaft?

[Cotnginny]

Three Section Telescopic Driveshaft?

I have two gearboxes I need to connect. In operation they will move laterally in relation to each other. They move in one plane only (no offset or angular movement). The present setup just uses a two part driveshave with a U-Joint at each end to allow for any slight movement caused by flex in the machine's frame. This shaft allows for movement of the gearboxes of about 8 inches.

Here is the problem. I need movement of 17 inches and it has to be a range of 23" to 40". So ANY length two part driveshaft will not work. (To reach 40" it would be too long to collapse back down to 23")

Confused? Just picture two gearboxes sitting next to each other and a driveshaft connecting them. I need to be able to push them together until the driveshaft is collasped to 23". Then be able to move them apart until the driveshaft is 40" extended.

So, where can I find or how can I build a 3 part driveshaft that can telescope down really short (23") AND extend really long (40")? Is this possible? What problems would this have?

HP is about 150, RPM is 1500. There is no need for CV Joint, UJoints are fine. Everybody at the Driveshaft Shops look at me like I'm a nut!

 

[Danger]

Welcome to PF, Cotnginny.
My first thought would be to use square tubing. It can be stacked almost endlessly, without the need for splines.

 

[AlephZero]

Does the shaft need to change length while the machine is running, or it is an adjustment with the power off?

 

[Cotnginny]

The shaft will be changing length while under load. One problem I have run up against is the sections must be "pinned" somehow so everything telescopes properly and does no just pull apart.

And square tubing could work if matching O.D.s and I.D.s could be obtained.

 

[FredGarvin]

I am amazed that you have that much movement in this set up. Anyways, I would look at having a very long round tube with female splines on the ID and a sliding shaft with male splines on the OD. Even with your shaft fully extended, i.e. where your spline engagement will be at it's least amount I think you'd be fine since you're not running any real power through the shaft. Having something that long will still be a bit of a shaft dynamics nightmare though. You'll need to look at having, at least, some kind of mid span pillow block bearing to control deflections.

 

[Q_Goest]

Splines would be nice, though I kinda like the square tube idea. I believe there are manufacturers that make sections of splined shafting. Check with ThomasNet.

Some additional concerns include balancing and wear on the telescoping pieces. 1500 RPM is about the same as automobile drive shafts. Check to see how they're balanced and who can do it. Regarding wear, metal to metal will require grease as a minimum. If the sections go in and out quickly, you'll need to consider ways of maintaining lubricant and different types of materials that don't wear as much.

 

[AlephZero]

Ideally you would want something like a keyed shaft (or the equivalent with splines) with only a short section actually in contact and carrying the torque. Otherwise it will jam up when you want it to slide, because it won't be perfectly aligned.

It wonder if the you could use a 3 piece system. A hollow sleeve with with internal splines along the full length and a central location bearing in the middle, and two drive shafts with a short section of spline on the end (maybe an inch long) connected to the gearbox.

Say two 11in shafts and a sleeve about 20in long. The center bearing support/location would be outside the sleeve.

Machining those 20 in long splines will cost you a few dollars though!

 

[Q_Goest]

Try looking in ThomanNet for splined shafts or drive shafts and have someone make it for you:
http://www.thomasnet.com/nsearch.ht...eading=74080409&navsec=prodsearch&oplisting=4
http://www.thomasnet.com/webresults.html?cov=NA&which=web&what=drive+shaft

 

[FredGarvin]

We get them custom made all the time for various test rigs. If they designed well, with the machining in mind, they can be well within the realm of affordable. However, I can not say that I have ever designed a 20 inch splined section before!

I think you're right Aleph that the 3 section approach would be the way to go. You would have some issues depending on alignment and lock up though. I don't see any way around that without using flex couplers that have more compliance than the splines do. It may still be a vibration mess.

 

[Cotnginny]

Good Points

Thanks guys. I have about narrowed this down to a 3 section approach. 17" of travel is not really a lot for Ag equipment (which this is). Its just that I have to be really short AND long. Good machining and balance looks like the key.

And thanks to those running this site. This is a great forum I've stumbled upon!

 

[Danger]

A hollow sleeve with with internal splines along the full length and a central location bearing in the middle, and two drive shafts with a short section of spline on the end

Machining those 20 in long splines will cost you a few dollars though!

How about a compromise? A piece of key-stock could be welded down one side of the tube, with one corresponding groove milled into each shaft. I'd consider putting seals on the ends and packing it with grease.

edit: I don't mean really packed; there'd be a vacuum trying to form as the rods separate. I just mean a fair slathering, with rubber seals to keep dirt out.

 

[AlephZero]

How about a compromise? A piece of key-stock could be welded down one side of the tube, with one corresponding groove milled into each shaft.

Torque in the shaft = 150 * 550 * (60/2pi) / 1500 = 525 ft lb.

If this is a fairly small diameter shaft, applying a few tons of force just at one side could be a problem in terms of bending the shaft. I would be happier applying the torque load at several points round the circumference (3 points minimum) so (in theory at least!) it won't bend.

 

[Danger]

You kinda lost me there, Aleph. I can see it wanting to twist if it's telescoped out enough, but where would the bending force come from?

 

[AlephZero]

If I understaod what you were proposing correctly:

Imagine the shaft is a circular section beam fixed at one end. You are applying a force to the other end, from the keyway, just at one point on the circumference. That is both twisting and bending the beam.

It also puts a reaction force and moment on the "fixed" end of the beam. For a shaft in bearings, that will be taken out as rotating radial loads on the bearings (of similar size to the force providing the torque - i.e. a few tons, a lot bigger than the loads you would get from "normal" unbalance in the shaft.

If you put two keyways on opposite sides, you apply two forces in opposite directions. That gives twist but no bending, and no radial load at the gearbox end.

I would prefer more than two keyways (or splines), but possibly for irrational reasons in this context. The dynamics of "two bladed" rotating structures (propellors, helicopter rotors, etc) is a lot more complicated than for more that two blades, my instinctive reaction to that situation is "not doing it at all is simpler than working out what might happen if you have to do it".

 

[Danger]

Okay, I think that I see what you're talking about. Wouldn't that tendency to bend, however, be negated by the circumferential support of the outer sleeve and centre bearing?

 

[AlephZero]

I was assuming the only contact between the shafts would be where the drive was taking place. I didn't think about the center shaft till your last question, but that also has a big problem with one keyway - see the attached free body diagrams.

I guess the only way it would work with one keyway was if the shafts DID support each other somehow - but personally I would still prefer killing off all these issues with a "proper" set of splines.

 

[Danger]

I'm afraid that I don't understand free-body diagrams. I never even heard of one until I joined PF, and then Fred was kind enough to explain the idea to me by PM. I could probably make one if I knew what I was starting with, but I can't interpret one that somebody else makes.
I totally agree that somewhere in the vicinity of 16-23 splines would be appropriate; I was just trying to look for the low-dollar alternatives. It really makes a big difference if this is to be a single-purpose personal project or a prototype for something that is intended to be patented for mass production. As a fairly economical compromise, 4 splines might deal with our mutual concerns.

 

[AlephZero]

"Free body diagram " is just a fancy name for a picture showing all the forces acting on an object. The important thing is being consistent about showing the forces acting ON the object, not the equal-and-opposite reactions, so it's clear how the forces balance, (or "are in equilibrium", to use more fancy words).

 

[Danger]

Yeah, that's how Fred explained it to me. Yours doesn't look even remotely like one that I would do, though. Apparently I misunderstood when he was telling me what is involved. I would have a more 'pictoral' approach, so that what's what is more apparent; I can't make heads nor tails of yours, despite the labels. I'm not faulting your diagram, only my ignorance of such things.