4

I see that most worm gears are self-locking but I'm looking to make my own non-self-locking POM worm gear. It'll only be of module 1. Does anyone know what needs to be changed in order to make a self-locking worm gear non-locking? Is it just the angle of the threads?

EDIT

Many motorised locks tend to have a worm gear on the motor shaft though - like this one. The lock is designed to be manually turned as well. Does that mean the friction caused by manual turning will be high?

John M.
  • 1,033
  • 1
  • 9
  • 13
  • what's POM? Do you need a worm gear specifically or a high gear ration in a small package - there may be alternatives, IDK, maybe ask a separate question – mart Feb 03 '16 at 08:37
  • 1
    POM is essentially acetal. I believe it's commonly used in rc cars, electronic locks, etc. – John M. Feb 03 '16 at 09:08

2 Answers2

4

Whether or not a worm gear is self locking depends on the lead angle and the coefficient of friction between worm and wheel.

It will self lock when the coefficient of friction is greater than the arctan of the lead angle (from memory plus quick google so not guaranteed but if you are designing it yourself I'm assuming that you can take the time to look it up in Machinery's Handbook).

However this is not an absolute as varying conditions and vibration can change this in practice. Also for high ratios it is likely to be self locking whether you like it or not. Similarly it is not a sharp cutoff and backdriving a worm gear will result in very high friction so there are very limited circumstances where it is something that you would want. It is certainly not a practical way of getting an equivalent drive in forward and reverse.

Regarding the lock, I can't see exactly how it works but bear in mind that the whole point of a lock is to couple or decouple a mechanical linkage when you insert a key. In this case I suspect the key turns the cylinder and the motor turns the whole barrel. In this case the worm gear must be self locking or it wouldn't be a lock.

Chris Johns
  • 15,216
  • 3
  • 23
  • 42
  • It seems many motorised locks tend to have a worm gear on the motor shaft though - like this one. The lock is designed to be manually turned as well. Does that mean the friction caused by manual turning will be high? – John M. Feb 03 '16 at 06:14
  • @JohnMunroe it's more likely that the manual turning is disconnected using a friction clutch. – ratchet freak Feb 03 '16 at 11:48
  • @ratchetfreak You mean there's a clutch inside the motor assembly? Wouldn't that increase costs/complexity a fair bit given that the worm could simply be replaced with bevel gears? – John M. Feb 03 '16 at 12:48
  • 1
    The clutch will be in the gearing so the manual rotation of the lock won't reach the motor. The type of clutch I'm talking about can be created using a freewheeling gear that gets pushed against a disk by a metal clip. – ratchet freak Feb 03 '16 at 14:49
  • @ratchetfreak Have you got a link to the type of clutch you're talking about? Especially a low profile one that might fit in the subject assembly. That lock doesn't seem to require additional engagement for manual turning though. – Kar Feb 03 '16 at 16:07
  • @Kar I've seen one in a cheap battery-powered analogue clock for the knob to set the time. – ratchet freak Feb 03 '16 at 17:12
  • 1
    @Kar no not a watch, a wall clock/alarm. It's just rotate the knob to set the time. – ratchet freak Feb 03 '16 at 19:01
3

Like Chris Johns mentioned, locking is dependent on lead angle and a coefficient of friction; specifically a coefficient of static friction.

To get a worm gear design to back drive you can increase the lead angle or decrease the friction coefficient, both of which increases the ratio of torque to friction. It will not be efficient, and even though it is spinning it may not transfer much load before locking.

Many "locking" worm drives can back drive when they are already in motion when a force was applied. This is because the dynamic friction coefficient is less than the static friction coefficient. This is why many safety dependent systems also have brakes since the "locking" characteristic of the worm drive is not a guarantee.

If you want to run a worm drive the other direction with reasonable efficiency you may want to look at a ball wormdrive similar to a ball screw design. The ball bearings have rolling resistance which is much lower than sliding friction. Probably too expensive for most projects though.

ericnutsch
  • 8,146
  • 1
  • 11
  • 29