Design of "jamming eccentric" fixings

So I have a practical back of the envelope level concept for a quick detachable dooby. I have an effective way to do the QD bit with a screw thread but its less than neat. More concept demonstrator than something I'd want to use.

I'd much prefer some sort of part turn, steel on steel, jamming eccentric device as used camloc chuck fittings, QD toolposts and all sorts of other things.

But I don't have any details on how to design such a thing so it doesn't come loose on its own and is able to generate enough holding force to stop the device moving. Heck I don't even know what the proper name for such things is so I can start looking on t'net.

I imagine the actual locking action is similar to that of a taper drill in merely requiring that the angle between the fixed and moving components is sufficiently small, but not flat on flat, for things to hold. Presumably there is some relationship between angle and area to define holding force. Ideally I'd like to get the surface forces down to a level that doesn't require hardened components.

From old catalogue pictures I know that sufficiently similar devices have been made to work well in the past so its clearly possible. Its just finding a design resource that has me stumped. I'd rather not follow the "monkey see, monkey do" approach of simply copying and re-dimensioning the camloc, QD tool post and other small systems I have to hand.

Thanks for any help.

Clive

Any help ? how to design a cam locking device

Thanks for that Limy, never thought to look on here.

Post #16 from Gordon Long quoting his book as saying "the cam should have about .001" rise per degree of rotation." looks to be an interesting starting point.

But most folk are implying that the cam needs to go over centre to lock implying something at least minimally springy to let it go over.

However the devices I'm referencing don't actually go over centre they jam up and stay jammed up just before centre. Just need to find out how its done. If worst comes to worst I could always draw up the device in a Dickson toolpost and feed in Gordons 0.001" rise per degree of rotation and see what that looks like. If the numbers fit it ought to be worth a try.

Clive

Clive603 said:

Thanks for that Limy, never thought to look on here.

Post #16 from Gordon Long quoting his book as saying "the cam should have about .001" rise per degree of rotation." looks to be an interesting starting point.

But most folk are implying that the cam needs to go over centre to lock implying something at least minimally springy to let it go over.

However the devices I'm referencing don't actually go over centre they jam up and stay jammed up just before centre. Just need to find out how its done. If worst comes to worst I could always draw up the device in a Dickson toolpost and feed in Gordons 0.001" rise per degree of rotation and see what that looks like. If the numbers fit it ought to be worth a try.

Clive

Click to expand...

Dickson toolposts and D camlock spindles and everything like that don't go over centre, but the one thing they do have in common is that the cam is only responsible for applying the load to hold the interface together. The rigidity of the interface is maintained by other features. Rigid interface = no vibration or movement between the mating components, so friction can hold the cam in place without fear of it rattling loose.

Here are a few pics from the cam design section of H.C. Town's "The Design and Construction of Machine Tools"...





Important bit is the spiral geometry of the cam.

BugRobotics said:

Important bit is the spiral geometry of the cam.

Click to expand...

Thanks for the pictures. Some decent data there, especially when combined with the 0.001 rise per degree cited by Gordons source.

As gregormarwick says its all a matter of generating enough friction to hold the lock steady. The actual device is positively located so thats no issue. Its just designing the lock so it has enough friction to stay put.

I'm certain I could just make something that would generally work but I really want a decent safety / reliability margin rather than something thats only just good enough if set just right. Been there with a piece of lab kit that worked and stayed put if the adjustment was set just so, maybe ± a thou or two. Setting it up every morning so it would be reliable through the day got old fast.

Clive