4140 shaft distortion during heat treat

[scottn]

I have a fluid power transmission shaft made from 4140. The shaft has a 3.75" diameter center section with helical spline and steps down on each end to roughly 2" diameter with bearing supports and straight spline outputs on either end. We currently induction harden only the sealing and bearing diameters which has proven insufficient to achieve the required service life so we've chosen to carburize the entire shaft.
I need to minimize distortion during heat treat so I specified 4140 annealed material with a rough turn op, then thermal stress relieve, finish turn and gear cutting, heat treat (vertically oriented), and grind to final diameters on sealing and bearing surfaces.
My heat treater says that won't solve the problem but doesn't offer any assistance. Any process help would be appreciated. I'm not limited by the choice of material at this point either, but I must cut the splines prior to heat treat. We have not seen any problems related to heat treating the splines.

 

[metengr]

Scottn;
You mention that you don't get the expected service life by induction hardening the 4140 fluid power drive shaft. Are you failing the shafts or have in service wear problems with the journals? Do you specify a heat treatment for the shaft material prior to local induction hardening?

I would think for this application that you would want to have a quench and temper heat treatment specified for this fluid power transmission shaft. In fact based on the stated shaft diameters, I would look at using an AISI Type 4340 alloy steel with an oil quench and temper heat treatment to provide a hardness range of 28 to 32 HRC (Rockwell C scale). The quench and temper heat treatment will result in optimum drive shaft strength, good toughness and fatigue resistance.

I don't believe just carburizing the entire drive shaft is going to help.

 

[unclesyd]

You need to look at the preheat treated varieties of this material. You can purchase heat treated 4140 specifically for shafts or you might want to try a proprietary steel like Maxel 3 1/2, or any of the preheat treated shaft materials. We have used Maxel 3 1/2 (preheat treated) for many years on shafts above two inches. We also have had good luck with Mirraloy TG&P.

http://www.associatedsteel.com

http://www.crucibleservice.com/eselector/prodbyapp/carbon/max3half.html

 

[TVP]

I agree that carburizing does not seem like the best way forward, especially not with 4140. If you decide to pursue carburizing, then you should be looking at carburizing grades like 4320, 8620, 9310, etc. Ascometal (part of Lucchini) has recently developed a new low-distortion carburizing steel called Jomasco 23 MnCrMo5.

Can you provide some additional details on the hardening that has been performed so far? What is the hardened depth? Is wear a problem, or just strength? Is failure through the splines?

 

[scottn]

We curently use a 4140 HT at 28-32 Rc, finish machine and cut the gears, then induction harden. In the application there are two failure modes. The first is helical sliding spline wear. We run a mating helical spline "sleeve" made from 80-55-06 ductile that fully encircles the shaft spline and the abrasiveness of the ductile wears out the 4140 at about an 8:1 ratio of direct measured wear. So my objective is to increase hardness of the 4140 until I even out the wear ratio. We have the capability to cut the external spline in material up to about Rc 35-38 and I believe I will need to be at least in the mid 40's to even out the wear if I stay with 4140.
The second problem I have is fatigue cracking at the output ends which typically starts at the spline relief diameter. This shaft currently has a reciprocating cycle (180 degree rotation forward loaded, backward unloaded) life of 1 million cycles transmitting approximately 30,000-40,000 in-lb of torque evenly divided between both shaft ends in an environment that produces frequent shock loading. The design target is 3 million cycles under the same conditions.
So my concerns are improving the wear characteristics on the helical spline and improving fatigue resistance at each end. Due to the gear cutting limitations, I must treat the shaft after machining and before grinding and still maintain reasonable dimensional stability.
I have considered cryogenic treatment, but I am under a tight budget and cannot add "significant" expense to this component.

 

[israelkk]

Why not go with selective nitriding of the spline which is done in much lower temperatures then induction hadenIng and quenching and carburizing

For the fatigue improvement a shot peening of the area where the cracks start should improve fatique life.

 

[NickE]

In my experience tripling the fatigue life without significantly increasing cost are mutually exclusive goals.

That said, there are two ways to increse fatigue life.

1) Increase the Tensile strength of the material. I dont think that you want to increase just the case hardness. I would be concerned about the overall strength of the material. You mention shock loading, the severity of the shocks will place an upper limit on the (core HRc) hardness that your application can withstand. Your HRc 32 corresponds to a Tensile Strength of ~140ksi, I dont have S-N curves handy for your steel, however I'll offer a guess that you'll want to keep peak stresses below 50-70ksi to get the 3M cycles.

2) Surface preparation. By removing crack initiation points and creating a compressive residual stress concentration in the surface you can dramatically increase the fatigue life. (In my work I see a 20x increase in life cycles to failure through surface preparation alone.) As israelkk says shot peening will help increase the fatigue life.

HTH (hope this helps)

nick

 

[TVP]

I agree with NickE-- investigate shot peening, as it is a low cost way to improve fatigue strength. israelkk's suggestion of low temperature nitriding is also a good one. The QPQ salt bath nitriding process from Kolene would be worth consideration.

While the additional details on the process were helpful, you still didn't provide information regarding the surface hardness or depth of hardening. 4140 is capable of surface hardening considerably in excess of 40 HRC when using the induction process. My initial guess is that induction hardening and low temperature tempering should be used, with a surface hardness requirement of 45-48 HRC minimum. The depth of hardening should be on the order of a couple of mm, say 2-3 mm, before returning to the core hardness.

NickE's suggestion of higher core strength/hardness is worth noting-- you may want to increase to something like 32-34 HRC minimum. Feel free to post additional details, as this is an interesting topic. See the following thread for more discussion on a similar application:

thread330-83411

 

[scottn]

Right now we're getting surface hardness of 52-56 Rc with no post-tempering and a case depth of .025"-.050" however we are not induction hardening the output splines or helical spline. The hardened zone washes out roughly where the cracks are forming- on the inboard side of the splines at each end. Up till now, we didn't have such a high cycle requirement so these issues weren't considered in the design. I started with our current shaft design to observe the behavior after 1M cycles. Perhaps we need to extend the induction hardening through the output splines and also look at the same process on the center helical spline. However, we would be induction hardening a full 80% of the shaft and at that point I started looking for an alternative to treat the entire shaft (i.e a carburize or Q&T) and minimize distortion. Another key element in this is the post treatment grinding. The sealing surfaces require a 4-8 Rms finish so I'm attempting to stay away from nitriding altogether because I'd likely grind away most of the process in the final operations.
I'm very appreciative of all you who have and are contributing to this discussion. Will supply more info as needed. Thanks!

 

[Flesh]

For dimensional stability we will austemper our 4140 shaft components. For you this will provide high strength with better overall toughness (compared to Q&T) and a ductile core. You could then follow with the induction harden process on bearing surfaces.

We also bump up our fatigue life by shot peening critical areas.

 

[TVP]

scottn,

Thanks for the additional information. I think that your hardened depth of 0.6-1.3 mm is too small-- recommend increasing to something more like 2-3 mm. And I would definitely investigate increasing the hardness (both surface and sub-surface) of the splines. Austempering is one technique for achieving higher hardness while minimizing distortion, and may be attractive if you want to start with normalized or as-rolled/cold-drawn bar for easier machining.

 

[alexit]

I would recommend core treatment (air quench) then machine followed by surface treatment (nitriding.) You can also look to ion implantation for very hard surface after nitriding, but this is expensive.

If you can machine at 35-38Rc get the core that high first, nitriding isn't going to distort that much (and leaving the white layer gives greatly incresed wear resistance.)

Alex