Threaded Bolt Selection 3

[ChrisConley]

I have a general threading question which I will attempt to describe as accurately as I can.

I need a device that will allow small adjustments in a table's orientation. This strut will allow for a one time adjustment of a large magnet. To this end a threaded shaft inserted into sockets, one socket is attached to a fixed support, the other is . As the shaft is rotated small movement in the shaft allows for fine tuning pitch,rotation, etc.

^^^^^^^^ ^^^^^^^^
---------------------------------------------------------
vvvvvvvvvvvv vvvvvvvvvvvvv

This is a quick representation of the shaft with two threaded ends. My questions are: what combination of handing will allow for the greatest accuracy of adjustment (RH-RH,RH-LH, etc.), Secondly is there a formula for calculating the amount of torque required to turn a shaft such as this one through one rotation. And finally is there a way to determine (once the handing is determined) how much distance will be covered by one rotation of the shaft.

I hope that I have been clear enough but feel free to request additional information.

 

[Like2Strike]

I am hoping that I am answering your question acurately.
Others I am sure will either agree or fine tune my explanation.

First of all think of a standard 1/4-20 screw.
Therory says that it should take 20 turns or revolutions to move the screw 1". This means .050 per turn. Your thread and nut (internal thread) has to be indexed preferably in an advancement mode to take the play out of the screw. Much like using the handle dials to measure a cut when using a milling machine. A left hand or right hand thread is up to the application.

A tourque wrench device would help to make accurate measurements in fine tuning.

What I am getting at here is that you can pick a thread pitch which will suit your needs. You just need to be sure that the threads have been precision ground for any kind of measurements to come close to indexing.

Live Aloha Frank M.
Tradewind Resources

http://twrusa.tripod.com

 

[ChrisConley]

Thanks Frank,

You have helped me in understanding thread selection. The second part of my question related to handing. I was told that if I was to use RH threading of a certain pitch on one end of the shaft, and LH threading of a different pitch on the other then I could minimize the travel that occured during one rotation of the shaft (and as such increase the accuracy). What I need to know is if this is true, and if not what the proper methodology is.

 

[vonlueke]

The torque would be determined by thread friction, resistance (if any) of the attached part(s), and probably thread sizes. We might need more information to try to predict that.

Based on the your description so far, if I understand it, it so far seemed to me that finer adjustment would occur with the RH-RH combination. However, this is pure conjecture, because check out the second paragraph of your first post. Are you missing some words in there? We need the description of the connected structure to understand your problem, but part of the words went missing. Could you retype/explain the third sentence in paragraph 2?

 

[ChrisConley]

Sorry about the mix-up on that sentence. I don't know what happened to the rest of it. The orginal sentence was to read: one socket is attached to a fixed support, the other is attached to a beam on which a magnet rests. To further explain. A 2.5m beam (weight approximately 2000kg) is supported at each end by a concrete pillar. The beam is connected to each pillar by three struts. These struts are angled in their connection such that small adjustments can be made to the orientation of the beam by an operator. On top of the beam is a magnet with an approximate weight of 200 kg.

You mentioned that you believed that RH-RH would offer the finest adjustment, would that be with the same thread on each end, or different threads?

 

[winkler]

Some other considerations: There are commercially available optical adjustment "stages" which have fine motion and very little free play in their design.

If you can't use commercial hardware to move your magnet, you could consider joining together (or machining on a single shaft) two different thread pitches. With one nut fixed and the other nut free to translate, the translational motion per rotation is the difference in pitches. A 32 tpi nut and a 40 tpi nut combination gives you the equivalent of 160 tpi fine adjustment.
It may be much easier to take a piece of hex stock and tap one end at 32 and the other at 40 tpi and put in two different screws to achieve the desired translation by rotating your "coupler".

Make sure that you can clamp your magnet into position after adjustment so that "inadvertent" adjustment doesn't happen. Good luck.

Winkler

 

[Like2Strike]

Chris,

Before I became a moldmaker I was a machinist who built and designed fixtures for drilling operations. Since your device is to be fixed at one end, I believe a better solution to your design is to start with a large dia. hex or square bar. Thread the ID of that piece with a coarse thread.

Then take a hex (smaller dia.) and machine down the OD leaving a bit of the hex on the end(this is an area that you can put a wrench onto to adjust the fitment of your device). Then thread the remaining OD to fit the threaded ID of the larger hex/bar. Next bore and thread the ID with a fine thread.

Then take another smaller hex and once again leaving a bit of the hex, machine the matching fine thread onto the OD.

Now by assembling the three pieces together you can make your coarse adjustments by turning the middle hex, and turn the smallest hex for your fine adjustment. You will just have to do the math to figure out your travel per turn, which depends on the threads you have chosen.

You can also use a torque wrench to turn the hex's to make accurate adjustments.

Live Aloha Frank M.
Tradewind Resources

http://twrusa.tripod.com

 

[vonlueke]

Chris: Regarding your original design, if I understand it, I'd currently say you'll get the finest adjustment with a RH-RH combination having two slightly different thread pitches on each end of your threaded rod. The smaller the difference in pitches, the finer the adjustment. If socket at end 1 (fixed support) has RH pitch p1 and socket at end 2 (beam) has slightly finer RH pitch p2, then end 2 socket axial movement per turn of the threaded rod would be s = (p1-p2) mm/turn. Note that if p2 = p1, then s = 0 mm/turn; no advance.

As an example, say end 1 thread is M10 x 1.5 and end 2 thread is M10 x 1.25. Therefore p1 = 1.50 mm, p2 = 1.25 mm, and s = 1.50-1.25 = 0.25 mm/turn. Thus if you rotate threaded rod theta = 0.5 turn, then end 2 socket would advance in the threaded rod axial direction x = (theta)(s) = (0.5 turn)(0.25 mm/turn) = 0.125 mm.

Note: Above solution assumes end 2 socket cannot rotate.

Note: It really doesn't matter whether p1 > p2, or p1 < p2; this only changes the direction of advancement relative to shaft rotation direction. The amount of advancement per turn will always be the difference between p1 and p2, for RH-RH.

If you were to use a RH-LH combination, you'd get a coarser advancement rate of s = (p1+p2) mm/turn.

 

[ChrisConley]

Thank you all, my design has moved forward well.

The final part of my question was regarding torque. I'm not looking for the numeric value for the torque (although it would be nice). What I'm really looking for is a formula that would allow me to calculate the torque myself, that way I can track down whatever information I need.

 

[winkler]

My copy of Mechanical Engineering Design by Shigley says

Torque = Force*Pitch diameter/5.

That's a generality regardless of size of bolt employed and indifferent to coarse or fine pitch. If you're trying to achieve a certain height setting, I would get out the height gage or long calipers from the tool room and just measure displacement while adjusting. Lubricants like dry film molybdenum disulfide or messy ol' oil will help keep your adjustment torque related to force load only.

Winkler

 

[poetix99]

Perhaps its too late, but, as &quot;winkler&quot; has pointed out, there are positioning tables that are capable of moving large masses. These are available with &quot;zero&quot; backlash.

Another possibility is to use a hydraulic (or pneumatic) cylinder with linear positioning control capability. These are available as a package (Valtek actuators are an example).

 

[ChrisConley]

Can anyone verify winkler's formula for Torque? Seems a little too simple to me (although it is possible that I have been thinking about this problem to long and anything would seem too simple at this point)

I agree that a hydraulic cylinder would work, however the positioner is only intended to be used once, possibly twice.

 

[winkler]

Errata. I cited the wrong &quot;d&quot; in the equation.

Torque = Force*MAJOR diameter/5

Thanks,

Winkler

 

[CoryPad]

I can confirm that T = 0.2 * d * F is a simple equation for threaded fasteners. It really should only be used for determining the order of magnitude of the torque, not for a final value.

 

[ChrisConley]

Thanks again all.

In addition to horizontal positioning I'm also now looking at vertical positioning of the same apparatus mentioned above. The catch is that only one end of the shaft is threaded now. Here is another bad sketch:

________
I I
I I
I I
I I
I I
I I
<I I>
<I I>
<I I>
<I I>
________

The shaft I'm looking at has a diameter of 1&quot; and as far as I can find the tightest thread/inch that is available for this diameter is 20. (UNE 1-20). Unfortunately this gives a pitch of .05&quot; which doesn't give the accuracy that I need (on the order of 100 microns is what they are looking for).

My questions are:
1) Is 20 threads/inch the tightest threading available on a 1&quot; shaft, if not what is?

2) Can anyone think of a way to get tighter accuracy for vertical positioning out of this screw?

3) Failing the above what other options do I have for positioning?

 

[trihydrate]

Machinery's Handbook gives the specifications for microscope objective threads. They are for one diameter only, 0.800 inch with a pitch of 0.027778 inch (36 threads per inch).

You might consider using a stepper motor to rotate the screw. The rational here is that a steeper motor can be used to divide one revolution into 20,000 or more microsteps.

Good luck!

 

[vonlueke]

T = K*D*P, as defined in previous posts, where T = torque, K = torque coefficient, D = thread nominal major diameter (bolt nominal size), and P = axial load. Analyze your structure to obtain applied axial load P, then compute T = K*D*P. (As Winkler stated, this formula is virtually insensitive to thread pitch p.)

For normal thread tolerances, clean steel on steel, and something like zinc-plated threads, many people assume K = 0.20 (or 0.22) for &quot;dry&quot; threads, and K = 0.15 for greased threads.

However, you have no washer face friction, which normally accounts for roughly 47% of the above torque resistance, plus you have two loaded threads on your first (horizontal) threaded rod. So, assuming normal thread tolerances, I'd probably predict for clean, zinc-plated steel on steel, a torque coefficient of K = (2 threads)[(53%)0.22] = 0.24 for &quot;dry,&quot; or K = (2)(53%)(0.15) = 0.16 for greased. Longer than standard thread engagement lengths might increase these values slightly (?), though I'm currently not sure of this. Also, thread materials, surface finish, thread tolerance, etc., different from the above could greatly increase or slightly decrease K. And as CoryPad mentioned, the above K values are mere estimates.

Is axial load on your threaded rod always constant, or can it come on and off, vibrate, and/or change sign? If constant, is it tension or compression? If load is always constant, back lash is perhaps no issue, so normal tolerance threads (class 6H/6g, 2B/2A), instead of close tolerance threads (3B/3A), could be used, reducing friction.

Also, if load is compression and rod is long, and you are sizing rod diameter large only for buckling strength, you could perhaps machine your rod ends down to a smaller thread size (sized for stress equal to say 90% of compressive yield strength), leaving diameter of unthreaded portion sized large for an equivalent buckling strength. Good luck.

 

[RobWard]

You could also use gearboxes to give a finer adjustment of the threaded rods. But also remember, you're asking for real accuracy here.
Have you also factored in the deflection of the components used? The mass of the components could deflect the beam out of tolerence.
I'd recommend measuring the final position in some way and not relying on the maths i.e. &quot;I've turned this rod 5 times, therefore the magnet has moved this much&quot; which is just asking for trouble/

 

[ChrisConley]

I hope I haven't made it sound like I was trying to use math to solve all my positioning problems. The reason I'm looking for the very fine threading is that in order to be able to get the accuracy I need, I need to have a positioning device with the ability to hold that accuracy. The movement from the pitch during a reasonable fraction of rotation is the tightest accuracy that can be reasonably assumed.

Back to my original question. Does anyone know of a manufacturer that makes 1&quot; bolts with tighter threading than 20 tpi? Cost really isn't an issue so custom made products are acceptable.