Worm-Gear advice sought

Hi,

I need to make a set of cable winches that need to be light-weight, high-torque, and cheap. (Each winch will be lifting itself up on a paracord rope, so I need to minimise the weight.)

To get the required torque from the motor, I need a worm-gearbox with a high gear ratio. (somewhere in the region of 50:1 to 100:1)

The worm-wheel needs to be around 150mm to 200mm in diameter.

The advice I would like to elicit is regarding the choice of materials and the best thread-form for the worm and worm-wheel.

I have read that to maximise strength, that ACME threads are better than 60-degree v-threads. This is the reason that ACME threads are used for lead-screws.

An involute tooth form could easily be used to cut the worm wheel, but then where would I get the worm itself, making one myself would be difficult.

A 60-degree tooth form could easily be used to cut the worm wheel, and the wheel itself would then just be some M20 studding, but a 60-degree tooth form has greater radial force (than ACME) which will tend to force the worm and worm-wheel apart, and would also provide less torque than other thread forms. (Will any of this matter in practice?)

I am very tempted to use a length of ACME threaded rod as the worm gear, as it is really cheap. I was thinking of using a piece of 20mm single-start ACME threaded rod with a 4mm pitch (Only costs £15 for a 500mm length on eBay)

However, I am not sure if you can make a worm-wheel that will correctly mesh with ACME threaded rod to make a working worm-gearbox? Will the contact point / pressure-angles be all wrong? Intuitively I would guess it gets better the larger the diameter of the worm-wheel, as an infinite diameter worm-wheel would have the same tooth-form as an ACME half-nut. Does anyone definitely know the answer?

(I can make my own hob from a piece of the ACME threaded rod, to cut the worm-wheel with)

I am assuming that the worm will be made of steel,but the other question I have is what the best material choice would be for the worm-wheel.

A high-Tin Phosphor-Bronze would seem to be the traditional answer as it is strong, but has a low-friction coefficient vs steel. However, it is very expensive, and heavy, and hard to obtain in small-ish quantities.

An Aluminium-Bronze alloy is another possibility, but again, that very expensive and hard to obtain in small-ish quantities.

I am tempted to use a high-performance aircraft aluminium such as 7075 for the worm wheels as it is strong, light, and cheaper than bronze. However it has a higher coefficient of friction against steel, so maybe it will jam under a high-torque static-load? You tell me!

I am finding it hard to calculate the required tooth strength, and therefore the required materials needed. This strength calculation is also affected by the choice of tooth form, and how many teeth are engaged at once, etc.

I am assuming for this that I am making a single-enveloping worm & worm-gear.

I think that single-enveloping ones will be about as much as I can hope to achieve using my modest home milling machine and a rotary-table. Attempting to do a double-enveloping worm wheel would be an act of hubris on my part.

Best Wishes

Nick

Hi Nick,

TEA Machine Components Inc. sell transmission components (gearboxes, worms, gears, etc) and have agents in Australia, USA, Canada, Germany, China............

They sell all types of transmission equipment at a reasonable price. See if you can find an agent in UK.

https://www.tea.net.au/

Paul

How much torque? How much weight will they be lifting? How much power will they be transmitting? (ie what size motor will each be running on?) What is the duty/application of the lifting gear? (lifting anvils with it, or running up a bamboo window blind?)

Have you thought about using automotive windscreen wiper motors? Nice controllable 12volt motor hooked to a worm and wheel of about 50:1 ratio. Bit light duty of course.

Or for something a bit heavier duty, old cordless drills hooked up to a suitable DC power supply work well. They have a built in reduction gearbox using planetary gears or some such with a very good reduction ratio. When you think of the hammering they take driving in roofing screws etc, they are pretty heavy duty for their size, so the gearbox could even be hooked up to a bigger motor if needed.

I agree you might find it hard to cut a worm wheel to match a standard Acme thread. The more common method is to cut the wormwheel to a standard involute gear tooth profile, say 20dp. Then make an Acme, or very similar profile, worm of the matching pitch. For 20dp gears, the matching pitch for a worm is .157", so no standard TPI matches.This can be turned up in most lathes with proper chainge gear facilities by the use of compound gearing.

If you really want to get into it, GH Thomas's book Model Eng. Workshop Manual goes into it in quite some detail. Ivan Law's book on Gears and Gear Cutting a bit less so, and there is some on the topic in Martin Cleeves' "Screwcutting in the lathe" book. I'm in the middle of building my own dividing head from scrap metal so have been looking at these to refresh my memory on worm cutting.

As for materials, again it depends on the service/useage you intend to use them for. Some details there would help. Generally speaking, a cast iron or bronze wheel with a steel worm is common practice. But for light duty, or infrequent use as is often the case in the home workshop, plain steel for both might suffice. I have never come across an aluminium worm or wheel in industy, so not sure how durable even 7075 would be in this application. It depends to on lubrication, eg submerged oil bath vs open gears with a squirt of oil or grease on them.

Assuming a 2mm pitch worm with triangular thread and 100:1 ratio, the wheel will need 100 teeth and have a pitch circumference of 200 mm. It's pitch diameter would then be ~60mm (sorry, thinking alound here). With 100 teeth my spreadsheet to calculate the dimensions of gear cutters using the "button" method shows that the "buttons" need to be 27mm dia - so the flank profile isn't far from a rack in fact. It would be fairly easy to make a cutter for such a form.

Most worm reduction devices seem to have triangular thread forms for the worm and involute wheel teeth - certainly my dividing head has. Unless you are anticipating very large loads I'd have thought this would be fine if you use reasonable strength material for the wheel (e.g. steel). You could reduce the wheel weight by drilling away some of it, like clock wheels are crossed-out.

Giving that you are making a winch system, another possibility might be to use a differential pulley and avoid worms completely?

Edited By John Haine on 14/05/2016 06:40:58

Nick,

Just a thought, before you go too far along the 'high ratio' path:

It might be worth cascading two worm reductions to get the desired ratio.

MichaelG.

Have a read of Andrew Johnstons article on cutting worms & wheels in the mag and on here. Could be hobbed using an off the shelf ACME tap or make the hob

If you want to keep weight down then get some hollow cast bronze to make a "ring gear" and fit that to a 6082 aluminium inner wheel.

If the winch only needs to lift its own weight then strength won't be much and will really depend on weight of motor. Also amount of use will be a factor as will any other loads

Edited By JasonB on 14/05/2016 07:58:29

There are several options.

A worm wheel of sorts can be made by cutting a the teeth in a gear at an angle which matches the helix angle of the worm that will be used to drive it. Curiously certain standard mod gear circular pitches are a close match to imperial tpi. For instance I noticed recently that 1.25mod is less than a thou out from 6 1/2 tpi. There are others. I'm not aware of any DP's that work out well. Lapping could easily remove a few thou error, the worm can also be angled.

Worm wheels can be cut with a tap. It's best to gash the tooth pitch first at the right angle to suit the tap/worm helx angle. Currently the cheapest option on suitable nice new sharp taps is trapezoidal from the far east. They are long taps with a taper intended for cutting deralin etc nuts or cleaning up machined ones. There is even a bit on the end intended to size the tapping hole. It MIGHT be possible to use that to radius the edge of the embryonic worm wheel and get more contact area.

The flank angles are just like gears. Initially 14 1/2 degrees was chosen on the basis that it produces less force on a gears spindle than say 20. Later this was abandoned an 20 became the standard which also results in stronger teeth. In the case of a worm steeper angles will result in more outwards forces on it. Not something I would be worried about but it probably impacts efficiency. On ACME the effective flank angle would be 27 1/2 degrees and on trapezoidal 30. Negligible difference really.

There may be another ways of getting a high reduction ratios but never tried working through them. I believe they are far more efficient than worms.

Forum reckons that the post is too long if the image is shown. But a google for 100:1 planetary gear design will show an image from the royalmech site part way down the images page. It isn't 100:1 though. It might be possible to access the page during the week.

John

-

Before worrying about tooth loads a better specification is needed. For instance does the gearbox need to be self locking, ie, what happens when the motor is turned off? If it doesn't need to be self locking then I agree with another poster that an epicyclical gearbox is probably going to be more efficient. However, a high reduction worm drive will probably be self-locking. In essence the tangent of the lead angle of the worm needs to be less than the coefficient of friction between the worm and worm wheel.

A few random notes:

1. The worm is the easy part, simple screwcutting whatever the tooth form

2. As far as I'm aware a single enveloping worm wheel cannot be made using just an involute cutter - there needs to be a hobbing process involved.

3. The tooth form of a worm wheel to mate with an Acme worm is indeed involute, but the involute is not constant across the width of the tooth. I never managed to get to the bottom of how the actual tooth form varies.

4. A worm and a helical gear has a point contact, a worm and single enveloping worm wheel has a line contact, so pressures and bending forces are spread.

5. As far as I'm aware the worm wheel is the same for single and double enveloping; it's the worm that changes. Machining a double enveloping worm would indeed be a challenge - essentially you're screwcutting on a circular path rather than linear.

6. I'm pretty sure both my rotary table and dividing head use Acme form worms.

It is simple to free hob worm wheels using a tap, but more difficult to be sure of the number of teeth you get. That may not matter in this application. If the number of teeth is important then the worm wheel needs to be gashed and then hobbed:

To get the proper tooth form:

Andrew

Double enveloping looks interesting - At first glance a form cutter in place of the wheel blank on the miller as set-up in your picture might work? It would be an interesting exercise to try...

Mark

PS. On the original subject a Harmonic drive might be a very good solution with the exception of the resistance to back driving of a high ratio worm (I don't know for certain but recall that you can reverse a harmonic drive?).

Mark,

If you are interested in double enveloping; have a browse around this site. ... it's about sextants, rather than heavy-engineering, but well worth exploring.

MichaelG.

.

See p5 of this Baldor brochure.

Nick,

A three stage epicyclic gearbox will give you the ratio you need in a more compact (but not necessarily lighter) arrangement.

Link below to a rope winch I recently built using a 125:1 gearbox. All gears were simple straight cut spur gears, the most complex to make being the housing ring gear. This gearbox is massively over- engineered for the power transferred, so could be made much lighter if needed.

**LINK**

Andrew.

An epicyclic gearbox with enough gears is essentially the same as a worm and wheel type reduction, that is you cannot drive the motor backwards via the gearbox, but much more compact. Look at aero engine prop-pitch motors and cowl-gill motors for ideas. I have several of both types and they are immensely powerful. The smallest prop-pitch motors have a 3200:1 reduction and hold a 1100 HP aero engine prop absolutely static. The smallest cowl-gill motors are normally 720:1 epicyclic coaxial drives and also have oodles of torque. I am using two such motors on the new radio telescope for azimuth and elevation movement of a 4.6m dish antenna that weighs about 450kg! The smallest Curtiss-Wright PP motor is rated at 1000 ft-lb of torque. The biggest one is over 9000 ft-lb torque.

A small cowl-gill motor used extensively by the RAF was the Rotax and they are still around if you dig hard enough.

If you want pictures of what's inside one of these beautiful pieces of kit then just PM me with an email address and I will send pictures. All the reduction gears are the same diameter and DP only the internal ring-gear would present a problem to make in the home workshop.

Hi,

Thank you all for the flood of comments.

I should probably give a bit more background, and do a spot of maths.

NB: I don't actually have a lathe, I am trying to do all this on a milling machine + rotary table.

The winches are built into a custom-built Chandelier, up to a ceiling with no space above it to mount or fit a conventional winch. The Chandelier needs to be raised and lowered only rarely to change bulbs and for cleaning, so speed isn't important. The winch needs to be low-profile so it can be discretely fitted into the boss of the Chandelier. The Chandelier would be raised and lowered on 3 ropes, so it does not rotate as it does so.

The worm wheels would be horizontal, and the rope would turn 90-degrees over a small diameter pulley to go vertical. The three winch mechanisms all need to fit in a pancake-shaped volume, 600mm in diameter and 70mm thick.

The Chandelier will probably weigh around 30Kg, but I shall cautiously assume 50Kg in the calculations here. The rope used is paracord 550 Type III, which is rated to 250Kg. Three of these ropes will be used, on separate, but synchronised winches. In practice this will spread the loading across the 3 winches, but any single winch should be rated to lift the whole weight. The winches should be self-locking in case of power failure during the raising or lowering. Once at the top, solenoid-operated bolts will secure it safely in position.

NB: Multi-stage Epicyclic (planetary) gears are lovely, but I can't readily cut internal gears (unless you know a way?), they also might back-drive if the power went.

If the winch capstan has a diameter of 50mm, then the torque is 50Kg * 0.025m * G = 12.26Nm

A module 4 worm gear of 50 teeth would have a diameter of 200mm

If the worm wheel has a diameter of 200mm, then the tangential force would be 12.26Nm / 0.1m = 122.6N

A module 3 worm gear of 50 teeth would have a diameter of 150mm

If the worm wheel has a diameter of 150mm, then the tangential force would be 12.26Nm / 0.075m = 163.5N

A module 2 worm gear of 50 teeth would have a diameter of 100mm

If the worm wheel has a diameter of 100mm, then the tangential force would be 12.26Nm / 0.05m = 245.2N

With a single-enveloping worm wheel, this shear-force would be spread out over several teeth. This should tell us what gear materials are suitable, based on the yield-strength of the metal. (The details of this calculation are left as an exercise for the reader).

Assuming 50 teeth on the worm-wheel, this 50:1 gear ratio means the driving motor needs to apply 12.26Nm/50 = 0.245Nm of torque.

I was planning to use an MFA/ Como Drills 919D148:1 gear motor to drive the worm.

This gearmotor has a maximum torque of 6000g-cm, which is 0.588Nm

Whether this is sufficient will depend upon the worm-drive's efficiency, which depends upon many factors such as the material-interface's friction coefficient, and the tooth geometry. I know that dual-start worms would be more efficient, but as they only give half the gear-ratio, I shall stick to a single-start worm.

To calculate how long it will take to lower the Chandelier, we first take the RPM of the output of the gear-motor, which is 84 RPM, and divide it by 50 to get the winch capstan speed of 1.78 RPM. The diameter of the capstan is 50mm, so the circumference is PI*50mm = 0.157m. Thus 1.78 RPM * 0.157m = 0.279m per minute.

To lower the Chandelier by 3 metres will take 3m/0.279m per minute = 10.7 minutes.

This is slow, but not unreasonable for something that may only happen once every 6 to 12 months.

Best Wishes,

Nick Lee

Although you may be able to syncronise the winches, how will you ensure the cord spools onto teh winch drums evenly to keep all three effective drum diameters the same.

And what will wind up the power cable to the fitting and winches?

Might be worth looking for the motors that retract blinds and canopies, these are quite compact and geared down well.

Dear JasonB,

Micro-switches cut the power to each winch individually when they reach the top, so they all start off with the same tension at the start of every move. It doesn't need to be any more sophisticated than that to ensure enough synchronisation.

Power to the winches is externally applied using a connector on a pole and a long 12V cable. That avoids needing a mechanism to wind-up a winch power cable as it goes up and down.

I already have a box full of 919D148:1 gearmotors left over from an earlier project, so I wanted to use up some of them on this project to save money.

As a reminder, the original posting questions were:

1) what tooth profile to use?

2) what metal material to use?

NB: I really want to use the same tooth profile 'type' on the worm and worm-wheel, as otherwise the contact area between them will be all wrong, and the tooth-wear rate and the tooth-strength will be badly compromised.

Best Wishes,

Nicholas Lee

Edited By Nicholas Lee on 14/05/2016 17:56:22

I never realised how much you resembled Patrick Mcnee when you are wearing oiur traditional working man's protective headgear.

Neil

The bowler is steel-lined, naturally.

So if you can reach up with a pole to plug in the winch power supply why bother with motors, just use a pole like a window winder. Still need to wind up the power supply to the light fitting though.

At only 30kg I'd be tempted to hide a couple of tool balancers in the "pancake" which will make it weightless then just push it up with a pole and hook it back down again.

The other easy option is to get a chandelier winch and fit it in your pancake rather than in a ceiling void, they make them flat and round so should be a simple job

Edited By JasonB on 14/05/2016 18:38:45