Splined axle design references

[bmcbride]

Hi All
I am wanting to find some definitive (published)information about "optimal design for high torsional loaded splinned transmission shafts) AKA Race car axles.
There is lots of comments on different forums about reducing the axle to below the splines root diameter etc, but little in the way of hard data such as effective stress concentration factors etc.

The other issue, is how to work in things like the fracture toughness of the ?? steel and hardness selected. I can work out steady state surface stresses etc easy enough, and the common throw away statement of a soft core is a tough core - whilst correct, when I Q&T 4340 I am going to get a uniform 40+ Rc hardness.

The hard part is working out the transient stress from a standing start (high rpm, clutch dump).

If anyone knows a robust reference on this topic, I would love to hear of it.

Thanks
Bill

 

[mfgenggear]

OP
Transmission shafts are based on the amount of torque and RPM. A gear box or transmission have to follow the same equations.
And is a very intricate and time consuming.
Best bet is to hire a consultant to help in thus matter. Race cars has heavy shock loads, and high torque values. And high RPM.
An average Trans or rear end [differential will fail. High chrome molly steels. With case harden surfaces. Its all in SAE. ISO , and AGMA specification. To save time a gear and shaft
Gear software saves time . The constraints are the size in side the gear box , transmission and center distance . Most automatics have epicyclic gears. High precision, ground or lapped. High precision agma. That means the shafts have to be case harden and ground. To remove all vibration. And diameters to with stand the torque.

 

[mfgenggear]

Ps I suspect race car manufacturers are are not giving up their trade secrets. That would be a hard nut to Crack.
But spline configurations are standard and can take enormous amount of torque and RPM. Such as turbo prop gear boxes, and shafts. 1500 HP.
The deal is it has to be sized correctly.
The root diameter of the spline. With the correct diametral pitch of module. And pressure angle. It must be resistant to tooth contact stress, tooth bending failure, and excessive runout. The shafts must be verified with free body diagram, for wobble, bending,
Per von misses. Have to have the right hardness, toughness, and diameter.

 

[bmcbride]

Thanks mfgenggear
Story goes I designed a fully floating cambered rear end conversion for a live rear axle car. Never broke a standard configuration FoMoCo axle which as far as I know is a 1040 steel with an induction hardened case - etching and hardness testing tend to confirm this. Matching and machining the spline etc is easy as it is just a serration rather than an involute.
My axles were a though hardened 4340 with a short length necked down to terminate the spline. 43 Rc was measured from a sample I cut off, polished and etched.
Only 1 side broke, the other side looks perfect, though I need to crack test it at some point.

I am limited in size unless I want to change out the diff centre - which I would rather not.

So a 43Rc 4340 steel 'should' runs circles around an induction hardened 1040? So perhaps there was a quench crack, perhaps there was a something when straightening the axles post heat treat - they were fully machined pre heat treat.

There was perhaps an inch of spline between the diff side gears and the necked diam, and that bit of spline has twisted 30 degrees before eventual failure. Cracks radiating from the centre to the root of many of the spline cuts.

Increasing the fillet radius, and reducing the amount of free spline to the necked area, and lengthening the necked length - possibly to the entire length of the axle - all seems possible if not sensible. But it would be "nice" to have something definitive to base this off.

 

[mfgenggear]

BM
Pull the data sheets on both steels.
And the SN charts as well.
Run the numbers von misses for torque and shear. It should be less then the yield of
The data sheet.
It is very well possible the shaft was heat treated and straightened incorrectly.

My suggestions.
Shafts are difficult to maintain as in straightness. During heat treat
Rack the splines vertical find a heat treated thar has a large enough oven.

Marquench or martemper. Oil or salt.
Higher temperature prevents parts tranforming from austenitic to martensitic grain structure.
To fast, Which lesson distortion. Slow acting oil.

Absolutely follow the AMS 2759/2 heat treat procedures. Even if it cost more.
Have the heat treat supplier NDT before and after

Request a fixed process.

Now the core can achieve 50 HRc avg
Look at Dat sheet and AMS spec for exact hardness.
The core should be approximately 36-43 HRc
The spline should be induction harden to
The max hardness. Minus 5 points.

Do not allow the heat treated to clean the shafts with abrasive silicon beads.
Must do it your self to control spline size.

Use spline gages. Go and no go, $1500

 

[mfgenggear]

Ps also if the shafts are not cold stabilize and double or triple tempered with in an hour I believe the shafts will Crack. It is imperative to cold stabilize ASAP. Or temper or both.
Straightening after quench and temper can Crack shafts if in correctly done. Shot peening to straightened works well, but the operator
Must be experienced.

 

[GregLocock]

There are various SAE papers around on spline shaft design but given the ridiculous price they are these days it is probably cheaper to buy a textbook like "Machine Design" by Deutschmann Michels & Wilson, which has about 70 relevant pages. I haven't designed a halfshaft using this book, but I have used it for other things successfully.

https://www.amazon.com/Machine-Design-Practice-Aaron-Deutschman/dp/0023290005

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[mfgenggear]

I posted a copy of the SAE splines at the pulley and gear forum, years ago

https://www.eng-tips.com/viewthread.cfm?qid=383119