Thread: Isuzu MSA Gearbox to LT230

Originally Posted by bee utey

Unfortunately that shaft design shrieks of a lack of understanding of mechanical engineering. What you have is a gearbox mainshaft that is supported by two bearings inside the rear end of the gearbox and the end of the input shaft. With gear loading, especially in first, the shaft will flex ever so slightly. Now you've added an extension rigidly bolted to the rear and firmly supported by the transfer input gear and it can't flex like the original any more. Not only that but you can't rely on the absolute accuracy of the bolted joint to keep the shaft precisely straight. The point of maximum stress in the adaptor shaft would be exactly where it broke. Now you could make a new shaft that's thicker at the point of failure, as well as cutting the spline into shaft material at the OD of the spline, but it's still fighting the extra support in the transfer case. If you were really unlucky you'd break the end off the mainshaft where the output flange is attached.

What to do? In the short term, don't use first gear under load, use low range and third - fifth instead. There'll be much less flexing load on the adaptor shaft. Later you'll need a shaft with a flexible join to the output flange, not sure what the final answer will be but an engineering shop might provide a durable solution. A splined flange as shown earlier would be a definite improvement but keeping wear under control will be difficult. An actual universal/CV/rotoflex joint might work too, not sure if you'd have the room.

Agree that everything can flex, but the Isuzu mainshaft is a huge bit of kit and the box is quite short.
Given the lack of alignment dowels, and the fabricated construction of the adapter housing (potentially with no post-weld machining) I'd think it much more likely that there is a static alignment issue with that.
The previous reports of housing failures don't create much confidence in that area either.

Steve

Steve, great minds think alike (I'm bound to annoy you at some stage so that analogy is not going to last ). While you were tapping on the keyboard I was out measuring the splines and agree with your comments.

The 10 spline measures approx 28mm dia while the Isuzu MSA5 is approx 37mm. Ignoring the loss of material where the spines are cut, and if the math is right, you get about 615 square mm and 1075 square mm respectively ie the MSA shaft is nearly twice the cross section.

And people, don't shoot me down when I mention Outcast I'm just as concerned about their product as you are about Sheldon's.. I gave them a ring today to see how their rigs are going. I've mentioned elsewhere they made about 6-7 units, one being mine. This is just some knowledge on practical application so we have some idea whether the Outcast/Sheldon design has some merit or not, if in theory construction is properly executed.

Outcast say they have had no shaft breakages to date and no case distortion.

What is the highest horsepower application? A twin turbo 4bd1t on 37 inch tyres 800nm at somewhere or other (didn't quite hear) and 305 kilo somethings at the fly wheel - quicker than an LS1 Range Rover off the mark. Owner regularly flogs the vehicle. I leave it for the torque heads to figure out what the specs mean and if impressive or not

What is the most traveled vehicle? Defender type, main test vehicle, 40,000kms to date with the shaft removed two or three times for inspection.

How long to get a shaft? 2-3 weeks. Heat treated, then cryo treated. And cost? Not really worked out to the dollar but less than 1/3 total kit cost.

Have you considered any other shaft design, giving Ancient Marinates example of a floating shaft? No, didn't consider any other design, but could be interesting in terms of easier replacement.

Then other bits of general conversation, some about not being able to compete price wise with the likes of Marks Adapters etc.

Originally Posted by workingonit

The 10 spline measures approx 28mm dia ...

I am sure the 10 spline is the same as the rover 10 spline axles:

Originally Posted by MaxiDrive

10 SPLINE: Used for axle shaft to diff side gear in ROVER Type diffs from the first L/R in 1948 to app 1992. Also used in L/R series 1 & 11 vehicles at drive flange and early R/R & Disco CV joints.
Square form (40 o included angle) Major dia 28.2 (1.11") Root dia 25.3 (.997") Zp ( polar section) .212
App torque capability for axle of 1500 Mpa UTS Hytuf 5635 Nm (4160 ft lbs)

Lets make it 100% clear that this is fine for the application. As stated above, 10-spline axles were used extensively at the wheel end from 1948-1992. The wheels receive up to 15 times the torque seen at the back of the gearbox. So although they were somewhat marginal for driving the wheels they are fine at the back of the gearbox.

The issue here is either the heat treatment or misalignment. The isuzu gearbox mainshaft is well supported so I doubt runout is an issue like bee utey suggests. I know MDE (and presumably also HTE) only trusted one company to heat treat their axles (in Melbourne I recall), so used to send batches down every 3-6 months.

AM's 2-piece design though is a better one, as long as the inner splines can be lubricated. I wouldn't use grease.

Originally Posted by isuzurover

<snip>... I know MDE (and presumably also HTE) only trusted one company to heat treat their axles (in Melbourne I recall), so used to send batches down every 3-6 months.
...<snip>.

I'm sure that Mal told me Sydney when he was making up my special axles (ENV to Stage one CV). More than that, when they receive the halfshafts back they still have to straighten each one.

This straightening is a much more involved process when its a short shaft that ends with a large flange at the end, like the MSA to LT230 adapter shaft.

I spoke to MR Auto today, and among other things talked about the spline shredding in non lubed input shafts and gears.

If I understood correctly, the problem was primarily in association with the manual boxes (LT77? R380?). A failed spline meant replacing the entire main gearbox shaft, as it was one piece from inside the gear box and out to the transfer case. This takes us back to Lots-a's idea of a one piece system - which may have some merit, given the Landy failure was actually lube related and not alignment related. The obvious downside - if you do have a failure it's expensive.

The auto boxes tended to protect the input shaft/input gear setup because they had 'quill stub shafts' (MR Auto wording if I remember the term correctly).

I think I understand quill shafts as they relate to milling machines, allowing the chuck to go up and down while being driven. Is Ancient Mariners idea of a floating shaft much the same idea, just doesn't have the same range of movement as in a milling machine? Or is an automotive 'quill shaft' some other beast?

I went to my museum of Land Rover parts (junk on the shed floor) and wiggled the output shaft on an auto box - at least 4-5 mm of end movement, side to side wobble. No backlash when rotated either direction. Couldn't pull it forward, nor push it back. So here is an output shaft that seems well integrated, but still has enough end movement to allow for some minor misalignment to the transfer case. How does this work?!

I consulted the exploded diagrams for the auto in the BMW workshop manual. Very small images with no parts list, so I'm not sure how the output shaft works. Can someone enlighten me? Yeh, ok, I should take the box apart and find out for myself someone is thinking...

...no, because I'm busy taking the transfer case and Outcast adapter off the MSA gearbox to make some observations .. The input shaft of the MSA (on which the clutch system rides) has the same 'forgiving' wobble exhibited by the ZF auto output shaft. Not the same story when I examine the MSA output shaft - seems solid as a rock. Must get a workshop manual for this box.

So the idea of something that is easy on the MSA output shaft is a good idea - but what is it? - is it necessary given the MSA shaft is nearly twice the cross section of the 10 spline?

My impression is that something made by Land Rover is going to break before anything on the Isuzu side does.

Provided Ancients idea is stronger than LRover but just slightly less than Isuzu then it is superior to the one piece shaft due to ease of replacement.

I've no experience with divorced systems (Suzuki and Nissan used it as factory fitted??). Does sound like more work initially; and alignment, hows that work?

The ZF output shaft is supported by the LT230 so can wobble when not installed. The Isuzu input shaft is supported by the spigot bearing when in service. Both cases have a single bearing in the box where the shaft sticks out and the movement is a function of the required bearing clearance. The output shaft of an Isuzu box won't have any discernable movement because of the rear housing having an extra bearing. Hence the need for a flexible joint to allow for slight misalignment in the LT230 adaptor.

A ZF "quill" shaft is attached to the back of the trans by a centre bolt and I suppose is replaceable but I never looked. The broken 1986 ZF and LT230 I acquired for free last year for the transfer box had a very worn quill shaft and corresponding input gear spline. Ashcrofts sold me a new gear and a special oil redirection plate to overcome further wear problems. This was a known problem with early LT230s and the oil plate is a later factory part.

http://www.ashcroft-transmissions.co...eed-plate.html

Thanks bee utey. I understand how a shaft only supported at one end by a ball bearing can wobble at the unsupported end, due to the 'bearing clearance' ie slack in the race.

But if the supported end is attached to another shaft by a high tolerance spline set then the role of the bearing for support of the first shaft becomes redundant (at least when stationary). How can the wobble at the unsupported end continue? My only conclusion is that the continued wobble at the unsupported end is due again to 'bearing clearance', but the bearings in question are now those supporting the other shaft and the first shaft? Guess I'm going to have to take the auto box output shaft off the box to really see what's going on.

The only reason I'm banging on about this is to see if there is yet another system available to us, not yet realised.

With those pics Damien posted, I can see how the shaft would do what it did...


JC