What's new
What's new

Spline Cutting Tips

superman22x

Plastic
Joined
Nov 5, 2017
Hello,

To start off... I'm still learning, and looking for some tips on cutting some splines. I am trying to make new CV axles for my LS3 944 out of a chunk of 4340HR I have. What I have seen on some YouTube videos is that people have used 90* double angle cutters for splines. Involutes would be ideal, but for splines, I think that splines are only involute due to ease of manufacturing more than anything. So the double angle cutter is used. But my thought is that rather than spend $100 on a decent double angle cutter that will not get used for anything else likely, is there any reason I can't just use a standard end mill on the 45* tangent of the shaft? See picture below, it would be the image on the left vs image on the right for the double angle cuter. The spline roots would be slightly different due to the angle of the flutes, but that is not important I think? If it is, something like a keyway cutter would work, that has basically 90* flutes to the rotation of the cutter. Am I missing something else?

Looking for input, maybe there is a better way to do this all together.

Capture.JPG

These are the splines I'm trying to cut for reference:
0218192001.jpg
 
Last edited:
First you have to find out what type of spline you need to cut. Most automotive use is a 90 degree serration, not an involute.

Ed.
 
All auto axles, CVs, etc. that I've seen made are involute.
Males are rolled not cut which makes a stronger part but may be what you refer to as easier. It is fast.
It is a process than can be a SOB to get right and into spec. and from what I've seen very few automotive engineers understand how to tweak it correctly.

A standard or even custom end mill may produce sharp corners which become stress points.
If you strip an axle spline at a buck plus things go bad very fast, the car becomes uncontrollable and you get to test the roof and cage strength.

That all said Ive done such as you want and lived to tell. Also put some cars dirty side up but that's part of building cars.
Anything that engages will work in a lower HP application, building such for say a Hellcat is maybe different.
Plans to heat treat after cutting? This a much better steel than most conventional axles are made of but heat and correct draw will be needed.
You likely don't need a 1/2 million mile life.

I would not toss to the curb the form normally used as simply easier to make. Many decades, now coming up on a century of what works and what fails.
If this toy will get pushed hard half of me says go for it, the other half says not a good idea.

It may look like just a spline but this is a safety critical joint.
Bob
 
If you’ve got a lathe, and some O1, you can make a button cutter to make something very close to an involute cutter for a spline.
Both made from O1, harden the button cutter to cut the involute cutter, then harden that.
Should get you a little closer to the correct form
 
First you have to find out what type of spline you need to cut. Most automotive use is a 90 degree serration, not an involute.

Ed.

Check out the picture I added to the original post, I believe these are involutes. I could be wrong.

All auto axles, CVs, etc. that I've seen made are involute.
Males are rolled not cut which makes a stronger part but may be what you refer to as easier. It is fast.
It is a process than can be a SOB to get right and into spec. and from what I've seen very few automotive engineers understand how to tweak it correctly.

A standard or even custom end mill may produce sharp corners which become stress points.
If you strip an axle spline at a buck plus things go bad very fast, the car becomes uncontrollable and you get to test the roof and cage strength.

That all said Ive done such as you want and lived to tell. Also put some cars dirty side up but that's part of building cars.
Anything that engages will work in a lower HP application, building such for say a Hellcat is maybe different.
Plans to heat treat after cutting? This a much better steel than most conventional axles are made of but heat and correct draw will be needed.
You likely don't need a 1/2 million mile life.

I would not toss to the curb the form normally used as simply easier to make. Many decades, now coming up on a century of what works and what fails.
If this toy will get pushed hard half of me says go for it, the other half says not a good idea.

It may look like just a spline but this is a safety critical joint.
Bob

Yes, I totally agree. A CV failure would be bad news. Although it typically is at launch more than anywhere else since your axle torque in 1st gear is way higher. An end mill with a slight radius (0.015") would give me the benefit of eliminating the stress riser in the corner at least. If I could simply cut these as involutes, I would be happy with that. But I think the gear cutters are generally too large. Maybe I'm wrong on that?
 
If you’ve got a lathe, and some O1, you can make a button cutter to make something very close to an involute cutter for a spline.
Both made from O1, harden the button cutter to cut the involute cutter, then harden that.
Should get you a little closer to the correct form

I do have a lathe. This might be the best way to go.

EDIT: Reading this again, I'm not totally following I was thinking just a single point cutter. But I'm not totally sure what you are describing now.
 
Carbide Bob is spot on. I do a lot of work for the GKN plant which makes CV joints for Ford, Chrysler,Honda, Toyota, Mazda, BMW, Mercedes and a few others. All are rolled involute and induction hardened. Internal are broached and induction hardened. Every single one is torque tested. We make spline ring gages and test equipment for the splines. There is way more to this than cutting something with an end mill that will fit into the female or transmission. Nothing safe about what you are doing. Make sure you have good brakes and can stand a loose half shaft coming at you through the firewall or floorboard. This is really a bad idea.
 
I do have a lathe. This might be the best way to go.

EDIT: Reading this again, I'm not totally following I was thinking just a single point cutter. But I'm not totally sure what you are describing now.

It's actually 2 button cutters, side-by-side, and the gap in the middle of them is what cuts the form cutter.
Whilst they are strictly speaking still circular rather than a true involute, the difference between circular and involute on such a short truncated form as a spline is minimal.

Theres a thread over on the 'other side' by John Stevenson which explains it all quite clearly, and shows the results of the cutters being made and used.

Link: Involute shaft
 
I've made half-shafts for a UOP Shadow Can-Am car, a Brabham BT-44 and a bunch of others .... used 300M and just hobbed the splines. Those are 30* involutes, standard hobs, most any gear shop will have them, pm Dan from Oakland here if you like, I know he does. Axles are pretty much a stressed component, if you lose one the car is going to make a very quick 90* turn, that could be not-fun. Cutting teeth is simple, not that expensive, will do a much better job than your homemade cutter (Altho that *is* possible if you absolutely have to, just don't try 3d printing, ugh.)

No, straight sides are not good (unless it's a straight-sided spline, which is a different animal), involutes self-center in the mating part and distribute the load evenly across all the teeth, you want the profiles to match please.

If you want to go for the ultimate, use Vascomax 250 or so ... that's what the mars rovers have. Mmmm :) I never could find anyone who wanted to pay that much (last I checked it was $18 a pound, most likely more now. But the heat treat is cheap :) ). Or "take the risk". Sissies :(

And if you like, gundrill out the middle. Now you got light, strong, and ductile. Also expen$$$$ive :)
 
It's actually 2 button cutters, side-by-side, and the gap in the middle of them is what cuts the form cutter.
Whilst they are strictly speaking still circular rather than a true involute, the difference between circular and involute on such a short truncated form as a spline is minimal.

Theres a thread over on the 'other side' by John Stevenson which explains it all quite clearly, and shows the results of the cutters being made and used.

Link: Involute shaft

I like this idea. Especially with my tiny splines, it should be VERY close.


I've made half-shafts for a UOP Shadow Can-Am car, a Brabham BT-44 and a bunch of others .... used 300M and just hobbed the splines. Those are 30* involutes, standard hobs, most any gear shop will have them, pm Dan from Oakland here if you like, I know he does. Axles are pretty much a stressed component, if you lose one the car is going to make a very quick 90* turn, that could be not-fun. Cutting teeth is simple, not that expensive, will do a much better job than your homemade cutter (Altho that *is* possible if you absolutely have to, just don't try 3d printing, ugh.)

No, straight sides are not good (unless it's a straight-sided spline, which is a different animal), involutes self-center in the mating part and distribute the load evenly across all the teeth, you want the profiles to match please.

If you want to go for the ultimate, use Vascomax 250 or so ... that's what the mars rovers have. Mmmm :) I never could find anyone who wanted to pay that much (last I checked it was $18 a pound, most likely more now. But the heat treat is cheap :) ). Or "take the risk". Sissies :(

And if you like, gundrill out the middle. Now you got light, strong, and ductile. Also expen$$$$ive :)

Sending them out would probably be a decent idea. But I would like to be able to make them myself if I could...
 
Another shot of the splines. The more I look at them, I'm not even sure they are involute. I would have thought they would be, but they sure look like 90* serrations. They are the same spline VW has been using since the 60s I believe.

20190219_214905.jpg
 
Another shot of the splines. The more I look at them, I'm not even sure they are involute.
Yeah, well I am. 30*, flat root. You probably want to go fillet root if you care about strength.

The difficult part will be to figure out what kind. They didn't have metric involute splines for a long time, they were in fact DP involutes translated to millimeters. It was ugly.

Nowadays I think they finally have actual metric module involute splines, which is even worse. I wish to hell someone would take the effing metric system and drown it :(

(You didn't mention it was off some furriner thing initially. You probably should do that. It affects the answers.)
 
Yeah, well I am. 30*, flat root. You probably want to go fillet root if you care about strength.

The difficult part will be to figure out what kind. They didn't have metric involute splines for a long time, they were in fact DP involutes translated to millimeters. It was ugly.

Nowadays I think they finally have actual metric module involute splines, which is even worse. I wish to hell someone would take the effing metric system and drown it :(

(You didn't mention it was off some furriner thing initially. You probably should do that. It affects the answers.)

Sorry, I mentioned it, but just by vehicle model. Porsche 944, with an LS3 engine.
So the DP on these could very well be some sort of decimal value then I guess... The shafts are 33 spline and the major diameter is 27mm. This actually puts me at a module of 0.8 if I calculated that right. So maybe I can simply find a module 0.8 gear cutter and do these things right. Further, a module 0.8 is a 31.75DP. So maybe a 32DP would be close enough to get the job done. Closer than the button cutter method, or as close?
 
When I am not sure of type of spline I take my 6" flexible rule and stick the corner into a spline and look at it with a back light, if it matches its a 90 degree serration if not go for a 30 PA spline. Ash Gear has a lot of info on this stuff. The automotive stuff people bring me are almost always 90 degree.

Ed.
 
Ash gear does have the cutter to do an involute, unfortunately the price is way higher than I can do for this one off type project. I can't seem to find any 30 PA spline cutters smaller than a module of M1, nor can I find one with a DP of 32. There are hobs available. When I look with my ruler, it looks straight to me. But Emanuel is pretty confident they are PA30 involutes.

I thought serrations were often used on Axles due to involutes binding, and they are not able to slide under torque.
 
Ash gear does have the cutter to do an involute, unfortunately the price is way higher than I can do for this one off type project.

I will not abide this for myself and I forbid their use here, but this isn't my project. If money is your primary concern call them back and ask for an import mill. I have no doubt that you will be able to find one for ~ $150 + or -.
 
Nowadays I think they finally have actual metric module involute splines, which is even worse. I wish to hell someone would take the effing metric system and drown it :(

Great idea provided you are happy living in the 19th Century.

The rest of us on the planet manage just fine with the metric system, right up until we find some bastardised version of some fitting done by someone from the dark ages.

Right now I have a couple on my work bench - 0.855" OD 26 TPI and 14.5mm 0.75mm pitch ID thread. I am NOT happy.....

PDW
 
Great idea provided you are happy living in the 19th Century.

The rest of us on the planet manage just fine with the metric system, right up until we find some bastardised version of some fitting done by someone from the dark ages.
That's because you don't have to deal with teeth. Metric module is absolute trash. Unbelievable garbage.

btw, I'd be plenty happy living in the 19th century. Can you arrange that ? :D
 
That's because you don't have to deal with teeth. Metric module is absolute trash. Unbelievable garbage.

btw, I'd be plenty happy living in the 19th century. Can you arrange that ? :D

Thought you were moving to British Columbia. Isn't that close enough to the 19th C for you?

Personally I'll give up 21C medicine & dentistry over *your* dead body.....

PDW
 
One more question on all this. If I get the right cutter for the job, would you spline the shaft after heat treating? I am planning to through-harden to somewhere in the mid-upper C30s. I have seem some recommend heat treating before, some say after? Either way, I will be making a finishing pass on the rest of the shaft, but the splines I would like to be able to test fit before removing from the fixture.
 








 
Back
Top