Split cotters

As promised, Martin : Here’s my version of the clamping geometry

… with the bar shown smaller diameter, for clarity.

Assume that the bar is ‘grounded’

• Tightening the clamp screw moves the two pads
• That in turn moves the whole block sideways
• Clamping then takes place at three lines [running into the page] with locations roughly indicated by the tips of the red arrows.
• Because the clamp radius and the bore in the block are [in reality very slightly] larger than the bar radius, there can be no other contact … it’s just those three lines.

With appropriate dimensions; these movements, and the clearances created, will be small : but they do say “a miss is as good as a mile”

MichaelG.

.

Edited By Michael Gilligan on 25/06/2021 21:38:42

There seems to be a bit of overthinking this wedging problem. If you accept that wedging can happen, the only place it can happen is where the clamping action is less than about 7 degrees. Eliminate that and solve the problem. It only needs a few microns of movement to cause wedging on a shallow angle and have no effect on the greater angles. The clearance on the sliding cotter/clamp bolt will allow this to happen.

Edited By david bennett 8 on 25/06/2021 23:39:30

Edited By david bennett 8 on 25/06/2021 23:41:43

.

So, please tell … how does that relate to the discussion that I am having with Martin ?

MichaelG.

.

P.S. __ personally, I would say there is just about the right amount of thinking going on.

Edited By Michael Gilligan on 25/06/2021 23:49:55

If you draw out what happens if the bar is ever so slightly smaller than the hole, as in reality it must be or it won't go in, you will find that the curved surface of the cotter first contacts the bar at the outer end of the curved profile, so it would appear that the angle at the inner end doesn't matter.

Edited By duncan webster on 26/06/2021 00:38:06

What makes you think it was addressed to you? It was a general comment on the thread.

I do not agree. The curved surface will contact both ends of the curved profile at the same time. Then the curve will force the sliding cotter around the bar and into a wedging action at the bottom.

.

Only the fact that it immediately followed my post.

Thanks for the clarification

MichaelG.

Edited By Michael Gilligan on 26/06/2021 06:24:01

Good Morning Guys,

I understand both points of view !

Only if one side of the cotter is longer than the other could there be any possibility of wedging, this could happen if the cotter were cut off centre or the cotter was a very loose fit.

Since a cut in the centre of the cotter would remove material and make both halves shorter by equal amounts the very fine taper on each would be removed.

Any burr on the cut edges will dig into the shaft being clamped and cause sticking and shaft damage so it must be removed.

I agree that the cotter provides a linear clamping action rather than a wedging one ! However as shown in the diagram that Michael posted the red arrows showing the force direction that will be directed at the back of the bore that the cotter slides in.

Wonder if there's a formulae to determine the spacing required between a posting and reply to ensure a navel contemplation as opposed to a posterior verbalisation without a hint of paranoia?

Regards Ian

Thank you for your valuable contribution, Ian

… it came like the proverbial breath of fresh air

[ or was it perhaps just wind ? ]

MichaelG.

Hope the weather is good in Wales this weekend, my next door neighbour has gone fishing with his mates.

Regards Ian.

What a performance to add a clip from a CAD drawing, but here it is. As you will see I've drawn the clearance exaggerated, but the principle holds, it contacts first at the outboard end

Edited By duncan webster on 26/06/2021 10:22:25

Hi Michael

I don't disagree with what you say and when the pad bolts are machined in situe when the main bore is bored then the arrangement will be pretty much as you describe. However even when machined as just stated there must be some clearance in the pad bolt bore itself and tightening the bolt may well move the line if contact quite possibly inwards.

To return to the situation of the OP that of the front bar of the Quorn it is impractical to machine in situe and a fixture is made to generate the pads seperately. In fact it cannot be guarenteed that the offset of the pad bore to the main bore in the fixture is exactly the same as that in the finished clamping system. In consequence the actual relationship of the curve on the pads may not be absolutely concentric with the main bar and it is anyones guess where the contact line will actually end up when fitted.

Taking the above statement to be the case. We should design for worst case which is when the line of contact is at the inner most end of the pad as this is the point of shallowest taper. Therefore if we use Duncan's formula of G = D/8 wherever the line of contact ends up there can be no wedging action and the cotter will always self release.

As I said I agree with your analysis for a perfect assembly, but even in that case parts wear and the wear on the pad will tend to eventually conform it to the radius of the main bar whereby your slightly larger radius on the pad disappears and contact is possible towards the inner ends of the pads. I'll ignore flexibility and material compression.

In practice locking is encountered at times. I have made a number of pad bolts and it can happen. My initial post was intended to show that offset was there just to allow clearence for the bolt and had nothing to do with the clamping action. This got us onto what does matter to clamping and to the subsequent debate.

So my conclusion is that whilst lockup will not normally happen if the system is made perfectly as you describe and is in an unworn condition, to ensure that it can never happen when that is not the case, having a Gap equal to one eigth of the main bar diameter is good practice, especially when the pads have not been machined in situe in the first instance.

regards Martin

Edited By Martin Kyte on 26/06/2021 11:15:07

I think you will find my post was timed BEFORE yours.

.

That’s rather odd, David … I do see your point

It appears to be timed before, but displayed after

… no idea how or why !!

MichaelG.

.

PS I have realised I cannot spell situ

:O)

Martin

.

No problem, Martin ... it's the thought that counts !

Thanks for this morning's note ... I do take your point about wear of the pads having the potential to change the geometry, and I must confess that I had only considered the 'as made' condition.

MichaelG.

Hi everyone,

As the person who started this thread I am been following the comment with interest and thanks again for all of your input.

I "think" we have at the answer to the question!

But here is another question, what about the material of the cotters?. I have always made them from brass to lock onto steel shafts, but Hemingways mention the cotters should be made of steel for the Quorn, as steel is more effective at locking. While this maybe true I will probably make mine out of brass for my Quorn as I have not had any issues with brass cotters and steel shafts on other equipment...

Regards

Derek

Hi Derek

I made mine from steel for what it's worth. The coefficients of friction are given as 0.51 brass to steel and 0.5 to 0.8 steel to steel. I mentioned earlier that I can get the front bar to lock by finger tightening the cap head screw (used for a temp try out) so with the ball handles it will just require nipping up. My preveous version had way too much movement on the ball handles with the slit casting design. Because of the limited space for the handles I feel that the cotters will eliminate the need for larfe swings of the handles. I lapped my front bar casting so the bore is only a few tenths bigger than the bar. I'm sure if you use brass it won't make much odds, you can always swap to steel later as you will have the fixture to machine another set if the fancy takes you.

Don't get too put off but all the theoretical chit chat that myself and others have indulged in, they aren't hard to make and do tend to work straight out of the box as it were.

regards Martin

Edited By Martin Kyte on 26/06/2021 22:20:27

Hi, I had a look at the arrangement on the table of an old pillar drill that I have. The pillar is a solid 3 inch diameter steel bar and the steel cotter with the hole for the bolt is 0.996 inches in diameter and 1.4 inches long and I assume the other one is the same, and as it will only come out by removing the table, I didn't measure it. The distance between the pillar and the cotter bolt is 3 / 16 inches, the bolt being half an inch in diameter and the hole for the cotter I removed is 1.004 inches, thus giving an 8 thou clearance between the two. There doesn't appear to be any noticeable wear on the cotters or the hole they are in and the table has no noticeable slop on the pillar, the table being a fairly heavy cast iron one. the distance between the two cotters when they are in contact with the pillar is approximately a quarter of an inch without being clamped up. The photo below bears evidence to Duncan Webster's CAD drawing that the outboard end makes contact first and there doesn't seem to be much hard contact to most of the area.

Regards Nick.