Making a knuckle joint.

[Robin GrahamParticipant @robingraham42208]

I have a project which involves making a knuckle joint along the lines shown here : https://en.wikipedia.org/wiki/Knuckle_joint_(mechanical)#/media /File:Knuckle_joint.jpg

Ideally I'd like to cut a radius in the female part of the joint in the horizontal plane to match the radius on the male part, rather than the vertical radius in the link.My first thought was that I could make some sort of fly-cutter to do it, but Doh! The axis of the cutter would have to be on the axis of the pin, so that ain't going to work.

Any ideas? It's not vital, I could do it as per the link, but it got me wondering – I am often amazed by the ingenious ways you guys come up with to solve seemingly (to me) impossible problems with home workshop kit.

Rob

[Robin GrahamParticipant @robingraham42208]

[Jeff DaymanParticipant @jeffdayman43397]

Not sure why you need to make the slot in the knuckle match the male rod end curve. If it's to keep fingers etc out, you could maybe cut the slot the usual way on a mill with a slitting saw or slot drill, then add two little patch pieces one each side to close the gaps. The patch pieces could be soldered or welded on into milled pockets on the sides of the female rod end, if done carefully they would be nearly invisible after cleanup with files etc.

If it's got to have the curve all the way through the slot, maybe you could make the female rod end and half the knuckle as one part and add the second eye of the knuckle as a separate piece. If done this way conventional milling and drilling could be used to make the female rod end and first eye, since the second eye would not be in the way. Maybe the second eye could be welded or soldered on, prior to drilling its' hole. The added-on eye hole position could then be picked up / located and drilled to match the other. Again, if welded or soldered carefully, after cleanup the joint would be almost invisible.

Personally I'd look for a less complex joint if possible, but the above ideas may help to figure out a method.

[Ady1Participant @ady1]

If it matches the male bit is it not then a hole all the way through??

So a boring bar job for a larger internal hole.

[Mick B1Participant @mickb1]

If you used a long endmill coming in from the opening in the forkend at angles, you could maybe cut a vee that would meet the flanks of the eye with a small clearance? Might need a counterbore or something to clean up the bottom where the 2 angled milling cuts meet.

Then no-one would ever know.

Except you, and He who sees all things…

[I.M. OUTAHEREParticipant @i-m-outahere]

What sort of size are you looking at ?

You could possibly treat the machining process as similar to cutting an internal groove for an "O"ring or circlip . The operation would be done on the lathe and set up in the four jaw chuck . Once the internal slot is cut the rest of the material to form the fork is removed with a slitting saw on the mill .

Ian.

[JasonBModerator @jasonb]

Again depends on size but a boring and facing head would do it.

[Neil WyattModerator @neilwyatt]

Is it just cosmetic?

Stick an endmill in the end and mill away most of the material that gets in the way, then file the lips in either side of the hole to match the male part

Or make a lost wax casting

Neil

[Perko7Participant @perko7]

Depending on what you are looking for, and if appearance is not a main issue, many places sell clevis pins and yokes of various sizes for marine or automation systems. In the UK look at Springfix Linkages or Online Bearings, both of whom appear to have a reasonable range of sizes, shapes and materials.

[Martin ConnellyParticipant @martinconnelly55370]

EDM would do it. Would need someone else to do it for you probably. Another option is to make a sandwich with a pre-machined central plate.

Martin C

[Mark RandParticipant @markrand96270]

One way would be to bore out the top arm of the female part to the same size as the radius you want in the centre part. Then make the pin with a smaller threaded end for attachment and a larger head to fit the larger hole that you bored in the top arm.

If that makes sense???

[Robin GrahamParticipant @robingraham42208]

Thanks for replies. It is I confess largely cosmetic; sometimes I over-engineer things just to get my head and hands round a new technique. Coming to this hobby relatively late in life and with no engineering background it's how I try to learn stuff and build a repertoire of skills.

In this case I was frustrated by my inability to think of a simple way of making the joint, and thought I might be overlooking something dead obvious – which happens to me too often! However, it seems not on this occasion. Thanks for all the suggestions though, which will be mulled over, and thanks Geoff Perkins for the link to Springfix which has gone into my bookmarks.

And Mark – yes, makes sense!

Regards, Robin

PS the joint will be fashioned from 3/4" square ali bar, one 'limb' about 600mm long, t'other maybe 60mm.

R.

[John ReeseParticipant @johnreese12848]

I think it could be done on a vertical rotary table. The eye of the yoke would have to be on the centerline of rotation. A long endmill could reach the bottom of the slot. By rotating the table the endmill could sweep the bottom of the slot. The resulting cut would be NEARLY flat bottomed. The eye could be contoured to match that curve. It would be a nasty job but would make the inside of the yoke match the outside of the eye.

[Robin GrahamParticipant @robingraham42208]

Posted by John Reese on 03/10/2017 02:49:26:

I think it could be done on a vertical rotary table. The eye of the yoke would have to be on the centerline of rotation. A long endmill could reach the bottom of the slot. By rotating the table the endmill could sweep the bottom of the slot. The resulting cut would be NEARLY flat bottomed. The eye could be contoured to match that curve. It would be a nasty job but would make the inside of the yoke match the outside of the eye.

Thanks for that idea John. For some reason I had a picture in my mind of the female part of the joint being in the long limb, so brain aborted possible solutions involving swinging in the lathe or using a rotary table at an embryonic stage. But of course I could make the female part in the short limb .

Once I realised that I came up with a solution similar to yours, but using a long ball-nose cutter which I think would give a consistent and well defined radius. But now I'm going to have to try and figure out exactly what you mean by NEARLY flat-bottomed using a square end cutter! I doubt that I'll actually make this joint for the project which brought it to mind, so it's a bit academic now, but it gives fodder for the 'leetle gray cells'.

Cheers, Rob.

[Tim StevensParticipant @timstevens64731]

The geometry of the slot machined as described above produces an outer surface (of the inside of the slot) which is curved a bit, because the sides of the slot are cut at the expected radius, but as the cutter cuts between them, it is not cutting at this radius but slightly outside it.

It is the same as cutting a straight slot with a slightly tilted cutter. Tilting in the direction of the slot, forwards or backwards, results in a half-ellipse being generated as the cutter rotates, at the bottom of the slot. Then, do the same round the inside of a drum and the outer surface is also curved rather than flat.

If you see what I mean.

Tim

[Robin GrahamParticipant @robingraham42208]

Posted by Tim Stevens on 04/10/2017 18:30:35:

The geometry of the slot machined as described above produces an outer surface (of the inside of the slot) which is curved a bit, because the sides of the slot are cut at the expected radius, but as the cutter cuts between them, it is not cutting at this radius but slightly outside it.

It is the same as cutting a straight slot with a slightly tilted cutter. Tilting in the direction of the slot, forwards or backwards, results in a half-ellipse being generated as the cutter rotates, at the bottom of the slot. Then, do the same round the inside of a drum and the outer surface is also curved rather than flat.

If you see what I mean.

Tim

Yes Tim, I do see what you mean – thanks. That confirms my gut feeling that the 'nearly flat' surface John referred to would be elliptical. I shall work this out in more detail though when I get a round (or possibly elliptical, parabolic or hyperbolic) tuit. I suspect conic sections come into it somewhere!

I dimly remember that John Stevenson posted some time back on calculating the profile cut by a canted endmill – I think a formula emerged, but I can't find the thread now.

Rob.

[Author]

Posts