Axle shaft FEA

[James86004]

I am doing an FEA on the infamous 10 spline axle shafts. I was wondering what torque I should apply to it? If I take the 130 ft-lbs max of my engine, multiply it by the 40:1 gear ratio for 1st gear low range, assume an efficiency of 80%, and assume all the torque goes to 1 wheel, I get scary stress concentrations in the roots of the splines. Like 1 million psi, which is 10 to 20 times more than the material can probably handle. The point is right at the edge where the half shaft is sticking out of the spider gear.

In reality it yields around that point, changing shape, and most of the time the power is going to more than one wheel, but still, no wonder these things break!

 

[revor]

My guess is that it's going to be breaking somewhere around 1200 ft lbs. The old 10 spline axle material was pretty good material, "something Like" 4340

A late model front breaks at about 1500 on the CV end, the rears twist alot 720 degrees) so the ultimate strength is hard to establish.

On the fronts that I make the break point is about 6500 ft lbs the rears are about the same.

 

[Alaska Mike]

Didn't Jim already do the testing? I think he was considering making 10 spline axles, but it wasn't worth it. They're very good for what they are.

 

[Mercedesrover]

I had a chemical analysis of factory 10-spline series axles, the results of which are back at work. I can post them tomorrow if you want. And as Keith said, yes, they are very close to 4340 alloy.

And I think I was the first to say it, and I'll say it again: 10-spline Land Rover axles are pretty darn good! The problem isn't composition but it's how this composition is affected by doing the job their asked to do. I would go as far as saying that if most people would change their axles every 15k miles, they would never have a problem with them.

Why these axles fail is the fact that they get work-hardened over the miles and become brittle. An axle shaft of larger diameter can be engineered to twist less (but still a fair amount) but because of their diameter can be stronger overall. I looked into having 10-spline axles made but for the money they really wouldn't be a huge improvement over the stock axles.

Now, are stock 10-splines good enough for a heavy Dormoble? Probably not. But for the average 88" running 31" tires? Sure. Just change them once a year.

 

[revor]

What he said...

 

[RonL]

Kevin,

Are you saying that your axle shaft are 5x stronger(1200lb vs. 6500lb.) then stock 24 spline?
That is incredible!!!!

 

[revor]

Kevin,

Are you saying that your axle shaft are 5x stronger(1200lb vs. 6500lb.) then stock 24 spline?
That is incredible!!!!

It's all in the heat treat, design and material. Oh and there were no accountants involved.

(no offense to the many accountant I know are out there )

 

[UnfrozenCaveman]

Hey, I resemble that remark!

 

[James86004]

Here is a screenshot of the analysis.

I never broke a halfshaft in my 88, nor did the guy I bought it from, who was the original owner.

The one we broke a couple of weeks ago in the Dormobile was brand new. I have heard from sources that may not be reputable that the new halfshafts you can buy today are not as good quality as the originals.

The amount of twist on a shaft is proportional to 1/(diameter^4), so a diameter increase from 1.1 to 1.25 inches decreases the amount of twist by 60%.

The material in the 10 spline shafts may be good, but the shape of the splines causes large stress concentrations. More modern splines use rounded edges and other tricks to lower the stress concentrations. This makes them more expensive to manufacture, at least in small batches in the 1960s.

 

[Mercedesrover]

I went and dug up my old analysis that I had done back in January of '07. I used an original Rover 10-spline Series shaft.

Hardness was 380 Vickers or 36 Rockwell "C" and they are thru-hardened.
Closest alloy is 4140

 

[revor]

The stress risers caused by the square pattern 10 spline... lots of near 90 degree corners..

Modern, involute pattern is soo much better and soo much cheaper to make.

But even today LR insists on using basically that same spline on the input gear of the LT230. The load is not nearly what it is on an axleshaft but still...

That said I'm happy they haven't changed it to something different as I can put a 2006 Transfer case in a 1988 RRC

 

[ntsqd]

Oh! 4140.......
Correct me if I'm wrong, but as I recall 4140 is rather sensitive to rapid or large section changes.

4340M or 300M would be what I'd first look at using. If that wasn't the budget, then 1045 or 1144

 

[greenmeanie]

E-150 although it is hard to find. Induction harden the spline to a minimum Rc48 rather than thru harden. Copper flash for giggles. Ultimately there are well understood and cheaper ways to achieve axle strength than tweaking a 10 spline design.

You can have a lot of fun with ANSI B92.1 for imperial or ANSI B92.2M for metric splines. Spline technology is not modern. Selecting the ptich, fit and the root shape makes a huge difference.

In my experience of older axles I tend to notice that it is less to do with the spline shape and more to do with spline wear allowing shock loading. James, what were your boundary conditions for your analysis. It would be interesting to simulate the shock loading case to compare.

It is one reason I like Jim's axles is that he eliminates the outer spline interface by using integral flanges. When replacing shafts it pays to check the fit with the old flange as this is often worn too.

 

[James86004]

I think Rover's spline technology was stuck in the pre-war era. Whether that was WWI or WWII I can't say.

This was a quick and dirty FEA. The BCs assumed the splines were a perfect fit and the load was static. I am still getting used to this Ansys Workbench, which does things differently than regular Ansys. It is way quicker to use, but you can sure make mistakes quicker, too.

I like Jim's integral flanges, too. Anywhere you can eliminate slop is a good thing.

 

[revor]

I think Rover's spline technology was stuck in the pre-war era. Whether that was WWI or WWII I can't say.


I like Jim's integral flanges, too. Anywhere you can eliminate slop is a good thing.

The War of 1812 comes to mind...

Flanged axles? What a novel thought...