Making splined shafts/axles, heat treat before machining?

[xr4x4ti]

Hello,

I am in need of making few splined shafts. I plan on making them out of 4140 or 4340 and heat treating them to 40-45 Rc. The longest one will be about 25in long, 1.125in OD and with a 1 module 27 tooth spline on both ends.

I have made splines before and I have carbide cutters. In the past I have always cut the splines on an existing part that was already hardened and the carbide inserts did the job no problem. My question is, should I rough out the shafts, leaving some stock, and then get them heat treated and cut the splines in the hardened state? I am concerned about the shafts warping after heat treat and having to be straightened. If I clean up the bearing surfaces by hard turning and cut the splines after heat treat, that will not be an issue.

Thoughts?

Related, do any suppliers provide 4140/4340 already heat treated to 40-45? 28-35 is quite common but a little soft for my application.

Thanks in advance,
Tim

 

[eKretz]

Yes, finish after heat treatment. I have never seen 4140 or its ilk supplied in bar form in higher hardness than the normal Q&T of around 28-32 HRc.

 

[EmGo]

I am concerned about the shafts warping after heat treat and having to be straightened.

Yup, they will, no matter what you do. But no biggy, not that hard to straighten.

They do warp less if the heat treater quenches them end-on instead of crosswise. Some places are smart about that, some not so much. You can ask, and if they get annoyed it's still better than having to take out a half inch bend

 

[guythatbrews]

Yup, they will, no matter what you do. But no biggy, not that hard to straighten.

They do warp less if the heat treater quenches them end-on instead of crosswise. Some places are smart about that, some not so much. You can ask, and if they get annoyed it's still better than having to take out a half inch bend

Used to do a lot of this. 9310 and 8620 case hardened and straightened after heat treat. Roughly same size as your shafts. Shafts drilled and tapped and sent to heat treat with eyebolts installed. 62 HRc so cutting after HT not a good option. And we had an Eitel straightening press so pretty easy to straighten.

Our heat treater knew the time of day and did a good job. Not all do!

If I was making just a few shafts 40-45 HRc and hadn't straightened a lot of stuff I might I finish after. Especially if low runout is a high priority. Peening is one way to straighten but isn't the prettiest if shafts are very crooked.

If you decide to finish before, assuming the shaft has centers both ends, leave a 1/4" wide land with a nice finish on the inboard side of splines. Easier to indicate on than coarse feed lines. Do the macro straightening at point of greatest deflection. Fine tune each end at the indicating journal. Make an educated guess about allowance for HT growth and apply it to your pre-HT sizes.

 

[eKretz]

Find a good heat treat shop and they will straighten parts if necessary as part of the job. I can straighten parts pretty well and do a good job of it, but I'd rather let the vendor do it if they can be trusted - they do it all the time on a day to day basis. And yeah, vertical quench on a long part can be good, but only if the part is actually kept fairly vertical on the way in! And you need good even heat for that to work without warping too.

 

[guythatbrews]

Find a good heat treat shop and they will straighten parts if necessary as part of the job. I can straighten parts pretty well and do a good job of it, but I'd rather let the vendor do it if they can be trusted - they do it all the time on a day to day basis.

Good idea! It is best to straighten ASAP after heat treat. None in my area will straighten.

 

[CarbideBob]

On a 4140 1.25 inch 10 inches long we expect .006/.008 warp from heat treat. Some will come out under .001.
One can rough in when soft and then finish .010 when hard. This is easier on the tools.
Many are fans of straightening by bending or peening.
Not me. This puts you in print today for shipping but I have to wonder how many have a rechecked such a part after 2-5 years on the shelf?
Seems it may want to wander back a bit.

Have made many millions of car axles but these never through hardened. Induction on the splines and bearing surface and this hard to control.
Every steel heat batch wants a different setting and you scrap the first one to ten to dial it in.
Worse yet you sit down and wait while metlab cuts and analyzes the structure.
Bob

 

[xr4x4ti]

Thanks for all of the great feedback. It looks like I was not out in left field when it comes to heat treating the shafts first. I think that is going to be my plan.

I am using carbide inserts from Advent Tool.

They can cut the 45 Rc material no problem.

Thanks,
Tim

 

[EmGo]

Have made many millions of car axles but these never through hardened. Induction on the splines and bearing surface and this hard to control.

There's quantity and there's quality Would not dare do this on the UOP Shadow, or the Brabham. I wanted to use vascomax and drill down the center but the chickenhearted owners insisted on 300M. And no lightening hole

 

[eKretz]

On a 4140 1.25 inch 10 inches long we expect .006/.008 warp from heat treat. Some will come out under .001.
One can rough in when soft and then finish .010 when hard. This is easier on the tools.
Many are fans of straightening by bending or peening.
Not me. This puts you in print today for shipping but I have to wonder how many have a rechecked such a part after 2-5 years on the shelf?
Seems it may want to wander back a bit.

Have made many millions of car axles but these never through hardened. Induction on the splines and bearing surface and this hard to control.
Every steel heat batch wants a different setting and you scrap the first one to ten to dial it in.
Worse yet you sit down and wait while metlab cuts and analyzes the structure.
Bob

You do it when you have to. There's no room for sussing out things that will work better or won't when you're doing onesie twosie parts. Nor is there room for scrapping up to ten pieces to make one or two. When they're all different, you don't get that luxury. So straightening it is unless the customer wants to pay a lot more, which is a pretty rare occurrence. That's a difference between production and job shop work.

 

[EmGo]

One can rough in when soft and then finish .010 when hard. This is easier on the tools.

What are you doing with cut splines, anyway ? Roll, baby, roll !

 

[memphisjed]

Non machinist question: why are shafts even hardened? Not the splines, the shaft even at bearings?

 

[eKretz]

Non machinist question: why are shafts even hardened? Not the splines, the shaft even at bearings?

Several reasons. One is increased bending resistance, (NOT stiffness, or flex - that doesn't change much with hardness) another is wear resistance. Lots of places press bearings on or off rather than use heat expansion. After a few times of doing this, a soft steel journal will show some wear and may even raise a gall.

 

[UptownSport]

Anyone have the method ?
Lost?
I have the 1915 book on the factory, but that's two years too early
"
Again, the camshaft has to have heat treatment in order to make
the surface hard; the cam shafts always came out of the heat-treat oven
somewhat warped, and even back in 1918, we employed 37 men just to
straighten the shafts. Several of our men experimented for about a year
and finally worked out a new form of oven in which the shafts could not
warp. In 1921, with the production much larger than in 1918, we employed
only eight men in the whole operation."

 

[eKretz]

I would guess the method is more even heating of the parts and a vertical quench. If you get it right there's usually very little warpage that way. The part being the same temperature throughout is also important to keep the parts as straight as possible. If it's hotter on one side than the other, even a vertical quench can result in warped parts. Many modern heat treating lines in mass production rotate the parts as they go through the oven for this reason.

 

[evil monkey]

Glad I checked out this post. I have never had to have parts heat treated so this was an informative discussion

 

[dgfoster]

Monarch had a setup for hardening lead screws hung vertically over a long vertical oil quench tank. The screw was held with an eye bolt on one end and lowers ate about 3 or 4 inches per minute through an induction heating ring located an inch or two above the tank lOne of their old films(not video!) showed the process. Seems like the ideal production method.

Denis

I would guess the method is more even heating of the parts and a vertical quench. If you get it right there's usually very little warpage that way. The part being the same temperature throughout is also important to keep the parts as straight as possible. If it's hotter on one side than the other, even a vertical quench can result in warped parts. Many modern heat treating lines in mass production rotate the parts as they go through the oven for this reason.

 

[sfriedberg]

A very old-school device was the "muffle furnace". Basically a ceramic or iron tube that shielded the work from direct exposure to the heat, and helped ensure a uniform temperature throughough the work.

 

[xr4x4ti]

Great conversation!

I am only making 1 or 2 of each shaft. These are CV axle shafts for custom AWD car build.

I am leaning very strongly towards the the heat treat and then machine. I am going to reach out to my local heat treater today and see how long of a bar they can do and if they can hang them vertically. But, my thought would be to by a 6 or 12ft piece and have the whole bar heat treated to 45 Rc. That way I would have plenty of stock on hand to make axles. After machining, they are ready to go.

Thoughts?
Tim

 

[eKretz]

Monarch had a setup for hardening lead screws hung vertically over a long vertical oil quench tank. The screw was held with an eye bolt on one end and lowers ate about 3 or 4 inches per minute through an induction heating ring located an inch or two above the tank lOne of their old films(not video!) showed the process. Seems like the ideal production method.

Denis

Yeah, definitely. Those induction rings have to be sprcifically sized for each individual part for efficient heating though. And that would also still benefit from rotation; I wonder if they rotated the lead screws during the process. Likely a high production only process like you said.