Backdriving efficiency and self-locking of worm gears

[hatchettr]

I have the backdriving efficiency of a two-stage worm gear reducer from the manufacturer as being 2%. I know that by definition a self-locking worm has a backdrive efficiency of zero. Is there any rule of thumb based on the value of backdriving efficiency to which you consider the gearset to be self-locking in practical use? (ie - would the above reducer be considered self-locking in practice?)

Thanks.

 

[insideman]

Even a self locking gear pair will creep when vibration occurs. This is a lesson each generation must learn anew!

 

[plasgears]

Winsmith Drives recommends that backdrive will not occur if the lead angle and helix angle are less than the arctan of the coeff of friction.

In metal worm gear systems that holds pretty well, but in plastic worm drives, the mesh runs-in nicely, and what did not backdrive will backdrive in time especially with vibratory loading.