curved lead screw 2

[njlee]

Hi All,

Has anyone seen a custom thread design that allows a threaded nut to follow a curved threaded rod (true arc trajectory), like bending a lead screw. can it be done?

I think it might be possible if the nut only has one thread inside.

has eveyone seen something like that? and where?

thanks,
Lee

 

[drawoh]

njlee,

Can you cut the nut with threads that grossly clear the threaded rod? If you use a thin ACME thread, it will have a full thread profile, and enough clearance to handle the curved rod.

JHG

 

[MikeHalloran]

I don't recall ever seeing it done, but that doesn't make it impossible.

The nut could follow a curved screw if it has only one internal thread, as you say, or if the internal thread actually comprises two wheels with thread- profile shaped tires. But there's not much load capacity either way because of the small pressure area.

If the nut is relatively long and relatively thin and of uniform thickness, you could bend the nut to the same radius as the screw is bent after the nut is threaded. That should be fairly strong.

There's also no reason you couldn't make the nut in multiple pieces, with axial splits, sort of like the half- nuts used for threading on lathes. You could probably EDM the working faces of the nut segments, using a piece of bent leadscrew as an electrode.

Recognize that the leadscrew has to be skinny and spring- tempered in order to allow any rotation, and the curvature and slender geometry greatly reduces its column stiffness, so that may limit the load you can transmit with it. You also should expect vibration resulting from coupling between the windup of the leadscrew, extra bending and lost displacement from the 'axial' load, and stick- slip in the nut and whatever you use for thrust bearings.

Find a chassis service manual for a '60..63 Pontiac Tempest for an example of a (not threaded) shaft designed to carry considerable torque (and no thrust) and be rotated while bent.

-Mike-

 

[MikeHalloran]

Oh, yeah. One other fly in the ointment. Like the Pontiac, the shaft requires two radial bearings at each end to induce the curvature. But in between the bearing pairs, i.e., in the part where you are inducing curvature, I think the shaft will take the shape of a catenary, not an arc.

-Mike-

 

[drawoh]

MikeHalloran,

I was not referring to a thin nut. I meant a regular shaped nut with a thin thread profile. You are correct though that the load is concentrated and the stresses are high.

Bending the nut could not work because you would be unable to rotate it around hte curved shaft. On the other hand, if the nut were made of some flexible plastic, you would have full thread contact. The stress analysis for this would be a whole lot of fun.

JHG

 

[arto]

How about something like a pair of rings with small cam-rollers, spring-connected to take up the backlash/slop?

 

[notnats]

How about an hourglass section nut? Large bore at the ends with a narrow waist.
Don't ask me how to machine it, that's not my area.

Jeff

 

[MikeHalloran]

Ah. I thought you were intending to rotate the shaft.

If you're intending to rotate the nut instead, then the shaft could be made with an interrupted thread, i.e., with 'teeth' on top and bottom, and the remainder relieved below the minor diameter of the thread, or milled to a diamond shape like locator pins, or made from bent sheet.

-Mike-

 

[Barry1961]

I think the Rho'Lix type nuts can travel on a bent smooth rods but are only rated for 889N (200lb). There are some other nuts that run on smooth rods that may have higher ratings.

Barry1961

 

[Cockroach]

Gentlemen, for my two cents worth I can't see it working in the conventional sense for the reason(s) specified. Most importantly I concur with the observation that the nut would need to rotate and follow the curved screw. It just doesn't seem right, but making the nut thinner and cutting back on tooth thickness would definitely introduce high stress concentrations.

I like the idea of a flexible nut and would offer the suggestion that operation be similar to the rotating ball screw type design. Then you would have an outside change of providing the necessary clearances since the teeth don't converge towards eachother going around the bend, if you know what I mean. The profiles must definitely be small otherwise the crests of the teeth converge and the nut would not pass. A ball screw would solve this issue.

I can't seem to wrap my brain around usage. Why would a guy need a "curved screw"? Also, I would hate to be the machinist trying to cut or roll the product. The setup(s) would truly be a novel manufacturing concept!

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[EnglishMuffin]

Well, it all depends on the degree of precision and the capacity required. For a relatively crude device, you could make the screw by bending a straight one. To create a crude flexible nut, if it were a light duty application and you were not much concerned about friction, you could perhaps use one or more short helical springs mounted inside a carrier, with their ends anchored into it by some means.

 

[Dennyd]

A star for EnglishMuffin for the helical spring as the nut idea! Few think outside the box this way and this is a very plausible solution depending on the needs. It might also be possible to use a similar spring around a bent shaft for the bent screw. What are the needs in this case? It is curious that NJLEE has not responded to any of these posts. We could certainly use a bit more information in order for our suggestions to be more on target.

- - -Dennyd, P.E.

 

[RebelBrill]

Another possible approach, depending upon loads, minimum bending radius of the screw, etc, may be to straighten the screw as it enters the nut and then allow it to recover its bent shape as it leaves the nut. I don't believe this is difficult to do for large bend radii and small loads. Small bend radii or large loads, however, will be a little more difficult.

 

[sacem1]

I agree with dennyd in that the originator of this string has not been in contact, would be nice to have some feedback or, is it just someone who takes a shot at something he belives will break our minds and then sits back and watches the show, that would not be nice to all of us, hope he just had to go out of town and not able to get in touch.

A star for EnglishMuffin for having saved us from roting in hell.

Cheers

SACEM1

 

[njlee]

Hi all,

Thanks for the replies. Long story short. I was challenged by an industrial designer to come out with a working plastic curved lead screw design for a medical device project (apparently, designers prefer arc over line). I agreed with him because of the fact that the payoff will be big, it will make the device follow human shape better. The nut needs to be rotated. load is not a major concern.

Initially, I was looking for some mathematical equations for the threads so that I can model it in CAD, make a prototype and try it out. After doing some research, I was surprised to find out that there are little information about this subject. maybe it is an opportunity.

Due to high volume, my client doesn't want to use any flexible parts or springs. Everything has to be rigid plastic for low cost. He would prefer to pay for 300 thousand dollars injection molding tool over springs that might only cost a few pennies each.

I still believe it is possible to do it. I think the main problem is the fact that the pitch will not be constant between the inner radius and outer radius along the rod. If the clearance is big, the nut might rock forward and backward which I don't like. To simplfy the problem, I have decided to consider true arc trajectory (constant radius) instead of random curves (splines or conics). The rod will have partial threads on opposite sides (no full rotation). I'm thinking of using a straight thread profile (constant pitch)....I'm thinking of using a collar (no threads inside) which can move smoothly along the curved rod to guide the nut......


Thanks for all the responses, it is nice to learn a few possible creative ways from you. I didn't respond because I don't have anything to contribute so far. I will definitely share my experiences after I have time to work on it a little bit more and have something to contribute.

It turns out to be a fun problem.


thanks,
Lee

 

[MadMango]

Plastics... a whole different approach is possible now. It could be as simple as a molded ABS lead screw with a comforming urethane nut.

Ray Reynolds
"Computers in the future may weigh no more than 1.5 tons."
Popular Mechanics, forecasting the relentless march of science, 1949
Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[unclesyd]

Am I missing something?
Is this lead screw or guide?
If you turn a curved lead screw by one end it swing in an orbit equal to the radius of curvature, unless the lead screw is . A guide could be used say with belt or chain driver.

 

[griffengm]

If one cuts a continuous helix on a straight shaft and bends the shaft, the thread spacing changes on the IR and OR. Depending on the application, the speed changes through one revolution could be significant yielding a sinusoidal speed pattern, not a uniform speed.
On the other hand, a steady feed might be developed but the machining of the mold would be interesting to say the least. The rights to the drive might be worth more than the device it goes on.
Does this device look or operate anything like an orbital welding head? Is the radius fixed or variable?


Griffy

 

[EnglishMuffin]

unclesyd : You can have lead screws with rotating nuts - in fact it's sometimes essential with very long screws in the machine tool world. One has to assume of course that this is a rotating nut application.

 

[Cockroach]

EnglishMuffin, I heated up a piece of Ready Rod and put a gradual bend in it, quarter circle of random radius. A tightly coiled extension spring was cut down, and a washer glued onto one end to serve as a knob. The threading of the Ready Rod naturally extended the helical spring making it act as a nut; any backlash was removed.

I'll be damned, the spring nut could indeed travel the entire length of bent screw, back and forth many times! I agree with Dennyd, you get the Star Award on this one!

Out of curiosity...with a name like "EnglishMuffin", any coaching suggestions for England in the European Cup 2004? I'm pretty sure after that game against France, beyond the box thinking would be welcomed!

Absolutely excellent idea, don't know what I can use this for, but excellent idea. Congratulations!

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada