Differential Output Shaft - What material?

[tigdlo]

Maybe the problem here is too much tyre surface contact & grip?

This is racing, no such thing as too much grip or too much power/torque.

So I’ll ask him to drop the diameter slightly between the end of the spline

This needs designing properly IMO. Simscale is a free online FEA program if you are into that kind of thing.

 

[Ubique]

This is racing, no such thing as too much grip or too much power/torque.

If it causes the point of failure to be pushed further down the drivetrain where it becomes catastrophic, I'd argue there is such a thing, I'm of a different mindset so I'll leave you to your 'death or glory'

 

[Maker]

The angled break is exactly how you'd expect to see torsional failure in a brittle material. I can't tell if the mark at 12 O'Clock in the second picture is where it gouged against the part it broke free from, but it looks a little bit like a material defect, perhaps a little inclusion in the billet. There's stripes on the left hand side that look a bit like "beach marks" too, suggesting it was a crack which propagated over a few loading cycles, but I'm not sure about that, it seems unlikely. It would be interesting to get some dye-pen spray on the other spline and see if there's any cracks.

EN36 is probably your best bet from the standard alloy steels.

 

[tigdlo]

If it causes the point of failure to be pushed further down the drivetrain where it becomes catastrophic, I'd argue there is such a thing, I'm of a different mindset so I'll leave you to your 'death or glory'

I get your point but start line forces are much greater than the rest of the race due to releasing all the stored energy of the engine+all the torque it can produce, if that is too much for the weakest link of the driveline then launch at a lower engine speed, no need to handicap yourself for the rest of the race. In anycase it's a 30 year old race part that is worn, it was going to break eventually, should have been changed out before this as there is clear wear, in hindsight ofcourse....

 

[Keithmac]

As @pressbrake1 says 300m was common for uprate shafts,

I bought a 300m AWD output shafft for my GTO as they were a common failure when launching (1/4 mileing).

 

[PhillipM]

In anycase it's a 30 year old race part that is worn, it was going to break eventually, should have been changed out before this as there is clear wear, in hindsight ofcourse....

This is why everyone is recommending S155/300M - yes it's not cheap but because it's a vacuum remelted alloy it has far less impurities and therefore much higher fatigue strength even though the ultimate strength and elongations seem initially similar.

There's a lot of cyclic loading on a driveshaft.

 

[EddyP]

Thanks for all the comments guys.

It's not as simple as just swap the diff out for something else, the packaging is extremely tight, there's certain ratios I need to have available as we change them for certain tracks, it needs to work with independant suspension and ideally it wants to be as light as possible.
The logical upgrade would be to a Sierra diff, but it's pretty heavy and I can't get the ratios I need so it would give us real problems elsewhere with the performance.
The diff itself is used in all sorts of ford escorts with loads more power etc. the difference is that they have half shafts that are 400mm long or so betweent he diff and the hubs, so the half shaft can twist a bit and take some of the forces on the start line, our car is unusual as it has a de dion suspension as a result it has these short output shafts on the diff which don't have the length in them to twist in the same way.
It's possible 6700 rpm is too much, to be fari to our driver she stalled it on the grid in the previous race so gave it a bit more beans to prevent that again, she's still getting to grips with the car and a pretty agressive plate clutch that doesn't allow any slippage.

As for the type of diff, it's got a torque biasing diff because it never allows full lock up, it's very progressive so won't put all of the power through a single shaft, the photo of the dragster twisting is down to engine torque rather than wheels slipping.

 

[jsf55]

Won't tell you how too race or how too leave the line as i'm way down in the fun runners, but have you tried a slow mo video of it leaving the line too see if you have a little bit of slippage too take the initial jar out of the shaft?

 

[8ob]

As for the type of diff, it's got a torque biasing diff because it never allows full lock up, it's very progressive so won't put all of the power through a single shaft, the photo of the dragster twisting is down to engine torque rather than wheels slipping.

That pic was only to demonstrate how full revs and dumping the clutch leaves one rear wheel light, a mates Camaro with a silly engine always broke the one driveshaft for the same reason as do a lot of drag cars . It would explain why you have one broken shaft and one that looks as good as new.

Bob

 

[EddyP]

Some of the cars with a live axle always break the same side, the one that's a bit shorter, guessing the longer one allows for a little more twist before failing.

 

[EddyP]

Few pics to give you an idea of what we’re talking about.

 

[Blobber]

Nothing helpfull to add on the technical side but the car looks like great fun!

 

[pressbrake1]

This is why everyone is recommending S155/300M - yes it's not cheap but because it's a vacuum remelted alloy it has far less impurities and therefore much higher fatigue strength even though the ultimate strength and elongations seem initially similar.

There's a lot of cyclic loading on a driveshaft.

People mix up yeald and uts
I tried en24 for cylinder studs , useless as it fatigued but humble en8 was perfect though on paper not as good but less prone to cracking

 

[PhillipM]

It's not even the yield that's really the issue, it's the fatigue limits. S155/300M will happily cope with 30% more torque for the same amount of fatigue cycles before failure than the 4340 it's based on. If it's the same torque for both it'll do about 50% more lifespan.

 

[Barking Mat]

I'll try and ask the rallycross company next door tomorrow, he knows his stuff.

 

[pressbrake1]

It's not even the yield that's really the issue, it's the fatigue limits. S155/300M will happily cope with 30% more torque for the same amount of fatigue cycles before failure than the 4340 it's based on. If it's the same torque for both it'll do about 50% more lifespan.

Yeald starts the fatigue

 

[PhillipM]

You can be well under the yield limits and still fatigue steel.

 

[pressbrake1]

You can be well under the yield limits and still fatigue steel.

I’m not explaining myself properly and don’t have enough command of writen English to do so .

 

[EddyP]

Just to follow this up, got the new shafts back today, took some advice from here and had the diameter reduced slightly between the bearing journal and the spline.
They’ve been done from EN36T in the end.