View Full Version : How strong are "split cotter" fasteners?

05-10-2013, 08:58 PM
As illustrated in a post by strider some time ago.


I saw a set of motorcycle triple trees using cotters to clamp the fork legs in position and wonder if they're truly strong enough for the job.

I know one clamps the heck out of my mill spindle.

05-10-2013, 10:20 PM
It's "ok" but there is no free lunch.. The clamp pressure is provided by the threaded fastener; the rest is surface area and coefficient of friction. A highly polished surface and "perfect" bore isn't a great combination. It is non-marring and certainly better than a couple of set screws. If you need to adjust and readjust a radial position, it's a pretty good solution.

I used a split-cotter on a bearing shaft with a 12 inch arm. It does a great job, BUT... it would work better on a lousy surface finish!

05-11-2013, 01:03 AM
It's using a wedging action to generate the clamping force, so it's exerting a significant outward load on the surrounding material. It seems like a slit on the end of the block with a clamping bolt would work just as well, and it would require fewer parts.

05-11-2013, 06:42 AM
Depends on the design proportions. If appropriate for the loads involved, they should be fine. Any that I've made have always locked up like the hinges of hell with minimum tightening.

Guy Lautard describes making them in one of his MBR books. In particular, he talks about the tradeoff between "clamping action" and "wedging action" depending on how the cotter intersects the piece being clamped and their relative diameters.

Lew Hartswick
05-11-2013, 10:10 AM
As a point or two of reference, the tailstock ram lock on the Clausing/Metosa lathes
and the quill lock on the Bridgeports seem to hold very well. :-)

Mark Rand
05-11-2013, 05:54 PM
I used a split-cotter on a bearing shaft with a 12 inch arm. It does a great job, BUT... it would work better on a lousy surface finish!

Think about Morse and Jacobs tapers and Jo blocks for a bit... A perfect fit with a smooth finish, will actually give a stronger connection than a rough finish because there's more surface area available to transmit the shear force and lower localised stresses as a result.

05-11-2013, 08:55 PM
Good theory ;)

Sure.... but the area of morse is 10's to 100's of times that of the typical split cotter, and the forces are much higher. Morse works because the female portion is stretched (ever so slightly, and some may say the male portion is very so slight compressed) - basically an interference fit. Split cotters doen't work that way. The force applied is from the threaded portion, and because of that they release instantly that force is removed (whch is generally good!).

If made carefully split-cotters hold great, but they (like all mechanical clamp devices) have limits. I have a 3/4 inch bearing shaft (hardened and polished) with the typical bored cotter. No matter how much I torque the 1/4 -20 screw SHCS (reasonable limits...) with a 12 inch moment arm I can make it move by strong arm pressure. It holds very well, but obviously I can exceed the limits.

As mentioned in a prior post, a split shaft with clamping screw probably works just as well or better. Spit cotters are however very repeatable - slight decrease in clamp presssure and they release.

Paul Alciatore
05-11-2013, 10:57 PM
One of the advantages of a split cotter is it does not apply any net rotational force on the shaft being clamped. This can be an advantage in certain circumstances. I made several blocks to clamp on lengths of rod for holding DIs in various positions. They are better than some other forms of clamps because the DI does not move as much when the clamping force is applied. The rods do not rotate or move longitudinally.

For the maximum in clamping force on a rod or shaft, I would try to employ a slot with two or more clamping screws to close that slot when tightened. This applies a uniform force all around the shaft (360 degrees) and if there are enough such screws, for the full length as well. I don't see how you can beat that unless you drill holes cross wise and insert (tapered) pins.