This question is going to have an application in aviation (to avoid doubts and relocation to other forum) but let's concentrate on theoretical mechanics here.

If I want to allow a shaft to rotate in one direction and not the other, then a ratchet would be a solution.

But what I need is a base/platform with shaft M (master) and shaft S (slave) connected to each other the way that if I rotate shaft M (both clockwise or counter-clockwise) the shaft S will rotate with the same speed and direction. But if I try to rotate shaft S in any direction it should not be possible the same way as if I tried to rotate a ratchet in blocked direction.

What is the name of a device I described?

Shaft M --> ===[Device in]=== <-- Shaft S
_______________[question ]________ <-- platform
  • $\begingroup$ Do you need to avoid back-rotation while the M-shaft is driving (i.e. the S shaft is trying to fight the M-shaft), or just avoid driving the M-shaft when it's not active? $\endgroup$ Nov 15, 2016 at 14:42
  • $\begingroup$ Are you okay with activating the device in order to couple the master to slave? Or are you looking for a passive device? $\endgroup$
    – Chuck
    Nov 15, 2016 at 20:21
  • 1
    $\begingroup$ @CarlWitthoft - Any "back-rotation" where the slave shaft "fights" the master shaft is the load torque. The rotational inertia attached to the slave shaft will mean that the slave-shafted system "wants" to stay at zero rpm. That is, a driving torque from the master side could also be viewed as an attempt to back-rotate the master. $\endgroup$
    – Chuck
    Nov 15, 2016 at 20:24

1 Answer 1


One basic class of mechanism which can achieve this is a non-backdrivable worm drive, however this does come with a failry large reduction ration and so may not be suitable for your application.

It would also be achievable, in some circumstances with an automatically engaging clutch such as a centrifugal clutch or you could envisage a sytem where the master shaft has a cam lug which engages in a spiral groove to pull the shaft fowrard and engage with a dog clutch with a return spring.

Alternatively a clutch mechnism which is engaged by a governor a bit like an automatic gearbox, this could be achieved by mechanical, hydraullic or electrical means.

Of course a lot depends on the sorts of torques and angluar speeds you are looking to deliver as well as the response, stiffness and backlash that you require I've seen very simple inertial engagement mechanisms in toys etc which achieve this but wouldn't be up to precision or heavy duty applications.

  • $\begingroup$ I'd vote for the cam-lug approach or a full-on clutch. $\endgroup$ Nov 15, 2016 at 14:41
  • $\begingroup$ If OP is okay with activating a clutch, and the desired behavior when no rotation is transmitted is for the output shaft to lock up ("rotate shaft in any direction should not be possible the same way as if I tried to rotate a ratchet in blocked direction"), then maybe the easiest answer would be a brake band on the shaft. $\endgroup$
    – Chuck
    Nov 15, 2016 at 20:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.