Hirth coupling face angle needed.

Hello Im trying to locate the proper face angle and Im not having much luck.

Ive seen 6-7 degrees(84-83) from different places. Is the coupling that forgiving where +/- a whole degree doesn't matter? Split the difference and aim for the middle?

Im in way over my head making this coupling but I enjoy the challenge. I have no master or really a decent way to measure the individual half I just plan to mate them up and measure what I can once together. This is for personal use only.

Im sure there is 50 other better ways to do what Im doing but really want to go this route right now.

Thanks.

What is the application of the Hirth coupling? This coupling was popular for coupling crankshafts on twin cylinder 2 stroke motorcycle engines in the 1960s

Its just for indexing but need a rigid backlash free joint. I think curvic couplings are more popular but figured a hirth was easier to make.

A Hirth coupling is like a straight sided gear tooth form. mating surfaces of the teeth are way more than 6 degrees. A Curvic coupling has wide flat teeth with shallow angle sides, like a few degrees or so.

Your description sounds like Curvic, not Hirth

macchi7 said:

This coupling was popular for coupling crankshafts on twin cylinder 2 stroke motorcycle engines in the 1960s

Click to expand...

Could you mention a few ? I've taken apart my share of two strokes and never seen a Hirth inside. Porsche roller bearing cranks had them, and some radial engines, but the RD 350 crowd could not afford that. Even TZ's did not have hirth couplings.

turnworks said:

I think curvy couplings are more popular ...

Click to expand...

Umm, curvic. Unless you were making a funny

I think they are only more popular becaue they are easier to make if you have a spiral bevel generator.

I don't think the angle matters that much, as long as the two halves mate.

If you had a hobber (big if) ITW made a line of interchangeable tool holders that used their Spiroid version of a spiral bevel as a locking coupling, similar to the curvic and hirth but easier to make.

But if you are just doing it for fun, just pretend the thing is a flat face bevel gear.(0* pitch cone angle, I guess) That should give you a lot more information and you could do it on conventional mill. Jones & Ryffel, Simplified Gear Design ?

Maybe I have my terms wrong. I tried machining it with a flat face and the teeth were equal depth but the width of the teeth changed as they got closer to the center. Its a 4" diameter part with a 1/2" wide raised area around the outside of the face. I then machined a 6.5 angle on the raised area of the face so it was concave. Almost got the proper width across the face but think I went too much on the angle. Ill try 6 degrees in the morning and see what happens.

Correct I meant curvic. I edited my previous post.

I will admit I was very impressed playing with the coupling I made having never seen one before. Using a height gage I put the two halves together and was within .0005" flatness then did some random rotation of the top half and was within .0005" across the surface in both height and flatness. Was just a practice try out of aluminum but still neat to see.

First, for those who are as ignorant as I:

Hirth joint - Wikipedia

and

curvic coupling - Bing images

If this is for indexing I would assume the load is low and the need for angular accuracy is high. And you want something that is easy to make. I would think that a double, standard universal joint would be the obvious answer. By combining two universal joints you remove the angular errors that occur with a single one.

Universal joint - Wikipedia

And they are a lot simpler to make. The trade off is that more length is needed.

turnworks said:

Its just for indexing but need a rigid backlash free joint. I think curvic couplings are more popular but figured a hirth was easier to make.

Click to expand...

turnworks said:

Maybe I have my terms wrong. I tried machining it with a flat face and the teeth were equal depth but the width of the teeth changed as they got closer to the center.

Click to expand...

I was thinking of a pitch cone angle of 90*, which would be flat, should not have said "flat face" because the face would be concave /me bad.

Pitch cone angle on a bevel gear is similar to pitch diameter on a spur or helical. So if the pitch cone is 90*, it's flat but the addendum will be above that and the dedendum below. That would make what looks like two shallow concave toothed cones that mate with each other.

Straight bevel gear teeth are tapered in both directions - height and thickness. So if you think of a regular bevel gear then push the ouside edge of the teeth forward until a plane through the middle of the teeth is flat across, that's what I am talking about. Make another one just like that for a mate.

Pretty bad description, hunh ?

That's basically what a hirth coupling is like. Easy to make bevel gears on a bevel gear machine but you can also do it by hand. That's what you are trying, I think.

Jones & Ryffel has the cleanest description and formulas for doing that. If you have nc you can generate the tooth shape with a dinky little ball end mill, that's how DMG does some great big spiral bevels now. It's not a very good method but for a one-off, it's a lot better than buying a slew of special machines.

Gleason actually built a "#5 Curvic coupling generator" if you want to go into production

@ EPAIII, the coolest thing about both curvics and hirth couplings is, the more you index them the more accurate they get. Wear will even them out. That's the opposite of usual machine tool behaviour

This is pretty bad, sorry, but I hope it gets the idea across - very much like a bevel gear, use a pitch cone angle of 90* (got that wrong too, dang) and then use bevel gear calcs out of the book to get the proportions. If you run a plane thru the red lines, that's your flat face, with the big ends of the teeth going deeper into the spaces at the outer ends than in the center. Teeth tapered in both directions.

View attachment 276434

(You could probably use any dimensions you want, just do the math for inscribed arcs and so on, but I would use the proportions of a bevel gear because it's all been worked out over the past century to be practical sizes.)

Don't forget to leave clearance at the roots, but again, if you use addendum and dedendum from gear practice, that will be all worked out for you.