Thread: Life of Puma Rear Diff's

From what I have seen of building dana axles in the USA, the tubes are made an shrink fit into the cast housing, then they are plug welded through the hole provided for that purpose.

If all this is done to good standard practice, the tubes won't pull out, If the wall thickness of the tube is sufficient then they wont fail in bending.

Then if the axle will subject to severe abuse (beyond most on this board), the cast housing may become the weak link and a brace and heavy cover can restore the required strength.

Ok, I must be missing something bigtime here.

I know a block (eg. a cube) of steel will carry the same load in compression or tension, however, a rod or tube of a given diameter loaded in tension will carry far greater loads than the same tube in compression, which will buckle at a substantially reduced load, won't it ?

Really rough from a mechanics POV.

A tube in tension will take a load up to its UTS, elastically deform (stretch), if you take the load past that point it will permanently yield/plastically deform, take it further and it will fail.

The same tube in compression will buckle long before its UTS is reached.
The longer the tube, the lower the compressive load it can take.

You have to increase the diameter/cross sectional area of the tube (increase its stiffness in bending) to reduce/eliminate the buckling.

Am I on the right track so far ?

All I know is that when working on race cars that went from pullrod to pushrod suspension with the same leverage ratios the pushrods had to be nearly three times the diameter for the same load or they'd buckle, and sometimes they still did just when hitting a ripple strip, whereas the pullrod cars could clout them all day long.





The old time axle trusses were merely a rod (3/8" or 1/2" ? I can't remember) that was attached from one side, under the diff banjo where it was also attached and connected to the axle tube on the other side.
If you tried the same arrangement over the top of the housing it'd be like pushing string, wouldn't it ?

It'd be a different kettle of fish if you used a single substantial plate as a single web from one side of the axle housing to the other, but how thick a web do you need to stop any buckling ?


[edit] oops, John has answered it before I even posted this.

Originally Posted by Bush65

From what I have seen of building dana axles in the USA, the tubes are made an shrink fit into the cast housing, then they are plug welded through the hole provided for that purpose.

If all this is done to good standard practice, the tubes won't pull out, If the wall thickness of the tube is sufficient then they wont fail in bending.

Then if the axle will subject to severe abuse (beyond most on this board), the cast housing may become the weak link and a brace and heavy cover can restore the required strength.

Correct, I have seen a D44 that had an 'upgrade' to thicker wall tubes for hard offroad use, only to witness it break the cast centre into pieces either side where the tubes go in because there was no flex in the tubes anymore.
An egg on face moment for all involved I reckon
Forces have to dissipate somehow/ somewhere, You need to allow them to dissipate along a length of tube, rather than trying to get a piece of brittle cast iron to do that....

JC

Originally Posted by Bush65

From what I have seen of building dana axles in the USA, the tubes are made an shrink fit into the cast housing, then they are plug welded through the hole provided for that purpose.

If all this is done to good standard practice, the tubes won't pull out, If the wall thickness of the tube is sufficient then they wont fail in bending.

Then if the axle will subject to severe abuse (beyond most on this board), the cast housing may become the weak link and a brace and heavy cover can restore the required strength.

Thanks John!

You are right - the tubes are a press fit then are plug welded.

As I mentioned, the only failures I have seen is on vehicles with MD lockers fitted, where a crack has propagated from the actuator mounting block (as per pic on previous link - previous page). I assume this is either due to the hole/block/welding(HAZ) creating a stress riser or weak point. This has only been on vehicles which hae done extensive travel heavily loaded on corrugated roads.

Originally Posted by wrinklearthur

I must have went to a different school! As design plays a far greater part in steel work than simply adding more material to the cross section.


I disagree, we are not talking about fabricated building girders here.
The problem here is, where the connection of the tube to the Salisbury's centre housing is, it fails under tension. When a strengthening plate is used under the centre housing and welded to the axle tube just beside the insides of the spring mounts, you form a triangulation on both sides of the diff centre housing. The result is the flat plate is now in tension and the axle tubes are in compression.
.

John explained it better than I can in his posts above.

Originally Posted by PAT303

Dave,is there a reason you don't make 3.54 HD CW&P for the P38?. Pat

yes 2 reasons,

1) the offset means you have a thin ring gear, we could I guess get this fatter by using the rover locker which has an additional 5mm offset (distance between ring gear mounting face and pinion centre line)

2) even with different teeth you still have the main problem of the pinion being too short ie the pinion bearings are too close together, these need to come apart to add strength

Dave

I think im about the dumbest guy in here. But it cant be a clear case of Tension or compression. Ben I think Johns post proves that steel is stronger in tension all be it not much but it is true.

Regarding the Sals housing, you cant just think of it as the wheels pushing the ends of the axle tubes up with the pivot point being the tube to cast center section connection....the coils that are inboard of the wheel are pushing down on the top of the axle tube, so it is now bending. If the truss is on top from inside of coil to inside of coil, would it not be in some tention as it resists the downward force of the coil.

every thing is a compromise, clearance, room, other existing links etc... the racers are lucky, the build the housing, truss it and then add links etc....most run coil overs so this is alot easier to accomidate, especially if they are mounted to a trailing arm.

see attached pic for a real housing

Originally Posted by uninformed

... But it cant be a clear case of Tension or compression. Ben I think Johns post proves that steel is stronger in tension all be it not much but it is true....

No Serg - steel performs the same in tension or compression. And John's post shows that if buckling can be avoided (as would be the case in any proper design) then that holds true for a steel diff bracing structure.

The AS though which John has cited gives a 15% higher safety margin for structures in compression.

For the average backyard fabricator who doesn't have access to FEA, that means that for all practical purposes their (usually overengineered) brace will be just as strong if they weld it to the top or bottom.

In most cases as we have seen in the pics the strength of a brace is limited by the design and other constraints, rather than whetehr it is welded top or bottom of the diff.

so are you saying that a truss can be designed to work in compression and avoid bucklling and be the same weight as one in tension?