Salisbury (Dana Powr-Lok) cross-pin redesign

[nadopine]

I have an old Dana 30 Powr-Lok limited-slip differential for my club race car. The stock cross pins that the 4 spider gears rotate on and engage the case ramps have a very high failure rate. They fail in the same area each time. It's been a few years since I was in materials lab, but it looks like a shear failure.

These cross pins are machined from castings as far as I can tell. There are sharp corners (i.e. no radii) and poor machining quality (i.e. rough tool marks). I would like to replicate these cross pins and continue to use this affordable Salisbury differential.

Does anyone know of any references to calculate the stress on these shafts?

Any material recommendations? My impression is that camshaft materials might work well.

Regards,

David

 

[NormPeterson]

I see this being more of an impact loading and fatigue issue. Just getting rid of the stress raisers should provide noticeable improvement.

For what it's worth, ASME piping stress analysis uses intensification factors in excess of 2.0 for fillet welds and threaded connections (the closest matches of discontinuity geometries with tabulated SIF's to your description). Fatigue life is in (roughly) an inverse 5th power relation to computed stress, including the stress intensification and impact factor effects . . .

Are you running a solid-center competition clutch disc?


Norm

 

[Philrock]

You might consider making a pin by machining down a socket head cap screw.

 

[nadopine]

Norm,

Impact loads are definitely a factor.

I'm running a solid hub clutch disc.

Philrock,

The shape doesn't lend itself to machining it from a SHCS.

See this link for pictures:

http://wwwrsphysse.anu.edu.au/~amh110/Diff_LSD/diff_lsd_volvo_dana.htm

Thanks for the input.

David

 

[GregLocock]

Would it be possible to make the two spider shafts as one unit, thereby eliminating the stress raisers?

The presence/absence of the hole is a bit suspicious.

As to materials I'd use 'axle steel' but given the likely cost of machining would be more likely to go for a decent titanium alloy, if calculations lead me to think it would be ok as a bearing.








Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[MikeHalloran]

Okay, you need a _big_ SHCS, or an axle shaft, and you need to remove a lot of it... and induction hardening would be a good idea.

Have you got enough friends to support a small production run?



Mike Halloran
Pembroke Pines, FL, USA

 

[MikeHalloran]

If I'm understanding the photos correctly, the torque transmitted puts the shaft in bending, as with any diff pin, and the force applied to the clutches applies two bending moments in a plane normal to the torque- induced bending. The notch needed to allow the two pin shafts to cross each other is of course disadvantageous for resisting the latter moment, and the hole could hardly be in a worse place.

I agree with Greg; a single piece resembling a u-joint cross would be much stronger... but then the mechanism wouldn't work. The notched areas of the shafts are forced away from each other as the shafts slide slightly on the opposing carrier bevels and push the side gears apart to compress the clutches.

If I were tooling up to produce replacement pins, I'd make the center a little thicker, face a corresponding amount from the small end of the side gears, eliminate the central hole, increase the radius of the transition between the center hub and the cylindrical part where the pin gears ride, use axle steel, and induction harden the part.




Mike Halloran
Pembroke Pines, FL, USA

 

[GregLocock]

I wondered where the mysterious ramp was that generated the preload in the clutch discs -obviously that's what the angled faces in the middle do.

That's pretty ugly, I wonder if a curved ramp could be used that gave a better path for the loads and fewer stress raisers.

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[MikeHalloran]

No, the ramp is not in the middle; it's at the ends of the pins. Where you'd expect to find pin holes in the carrier, there's half a round hole in one half of the carrier (which I think is big enough to never touch the rounded half of the pin ends), and a v-groove in the opposing half of the carrier that bears on the angled faces on the ends of the pins.

( At least I think that's how it works. I had to study the photos for most of an hour to get that far. )

Agree, it's ugly, and the center is full of stress raisers.





Mike Halloran
Pembroke Pines, FL, USA

 

[GregLocock]

Oh, so the squarish centre section is just as it seems, a flattened bit so the two shafts can cross over each other?

What do your think the hole is for - an attempt at controlling the stress path, lubrication, or what?



Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[MikeHalloran]

Can't be for lubrication.

Now that you mention it... maybe the hole is there so the pin is guaranteed to crack through the center and the diff will keep working for a while instead of exploding right away.

Makes you wonder if any of them really made it through the warranty period.





Mike Halloran
Pembroke Pines, FL, USA

 

[Tmoose]

I'd take some close up shots of the fracture surfaces and post them on this board.

I'd take 2 new ones and magnaflux (wet method, usually called magnaglow) them. They would have to be magnetized in a few directions.
Them I would gently grind radiuses and organic shapes in all areas involved in the cracking.
Then I'd mask the bearing surfaces and get the areas that crack shot-peened with steel shot to Almen A 10 to 15 or so.

 

[RossABQ]

I would guess the easiest, cheapest solution is to step up to a Dana 44, a Ford 8.8, etc. I can't believe you could make a piece that is "guaranteed" to live under these loads in a 30 that costs less than a complete 44 built up with aftermarket parts.

 

[nadopine]

Mike,

I'm willing to look into making a small production run as these cross shafts are NLA from Dana and most differential shops.

Tmoose,

I'll post some close ups shortly.

RossABQ,

A Dana 44 is completely overkill and too heavy for my needs. I have a Volvo race car (ironic I know) and the Dana 30 is the stock differential. I'm only making 150 ft-lb or so at the crank so I don't need a Corvette/Viper rear end. Thanks for the input.

David

 

[GregLocock]

Is there any spare room around the centre so that you could increase the profile where they cross? If you dimension them up I could FEA them for you.

Do they always crack in the same direction?





Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[MikeHalloran]

Some details that I would change:

The flats on the pin ends are crowned, so they rub mostly on the center of the v-grooves, and probably so the pin won't fracture the carrier after it cracks. I'd still crown the flats, but move the center of the crown inward so as to reduce the bending moment induced on the pin.

The 'crank cheeks' near the center appear to have been axially broached. Said hypothetical broach teeth are v-shaped and sharply pointed. The crack goes right through the intersection of that v's root and the generous radius of the (as- forged?) relief in the center. I think I see a little tit on the fracture surface at that exact location. I'd put a radius in there at the root of the v, and maybe do the whole area with a ball mill, to reduce the stress raiser. Think CNC mill, not broach.

There appears to be a lot of clearance between the 'crank cheeks' and the pin gears. I'd move everything out as close to the pin gears as I could get it, leaving more metal. I might extend the flats all the way to the center. They probably have to do with lubrication of the pin gears, and the extra metal in the flat area inboard of the pin gears does not appear to have been doing anything.



Mike Halloran
Pembroke Pines, FL, USA