Diff Lubrication

[Wesley]

So this marks the third year we're running a Salisbury style differential in the car, and we've had longevity issues with clutches. Admittedly, some of those were due to things like forgetting oil, but even when it was done properly, the clutches haven't been lasting more than 5 to 10 endurances.We were wondering about the diff oil being effective at rotational speeds, so we build a little mock up for kicks. The results were interesting.

At ~20mph:
http://photos-d.ll.facebook.com/photos-ll-snc1/v1324/181/98/9619773/n9619773_38493915_3727.jpg
http://photos-h.ll.facebook.com/photos-ll-snc1/v1324/181/98/9619773/n9619773_38493919_4667.jpg

It's really interesting to watch, because above a critical speed, the oil forms a ring on the inside of the case, pretty much avoiding the internals. We only had one solution, and that was to fill the diff up until it touched even at speed, but we're worried about our seal design containing that much fluid, especially since they were just designed for splash, not submerging.

Video (http://www.youtube.com/watch?v=wCqfSBcm6l4)

[Wesley]

So this marks the third year we're running a Salisbury style differential in the car, and we've had longevity issues with clutches. Admittedly, some of those were due to things like forgetting oil, but even when it was done properly, the clutches haven't been lasting more than 5 to 10 endurances.We were wondering about the diff oil being effective at rotational speeds, so we build a little mock up for kicks. The results were interesting.

At ~20mph:
http://photos-d.ll.facebook.com/photos-ll-snc1/v1324/181/98/9619773/n9619773_38493915_3727.jpg
http://photos-h.ll.facebook.com/photos-ll-snc1/v1324/181/98/9619773/n9619773_38493919_4667.jpg

It's really interesting to watch, because above a critical speed, the oil forms a ring on the inside of the case, pretty much avoiding the internals. We only had one solution, and that was to fill the diff up until it touched even at speed, but we're worried about our seal design containing that much fluid, especially since they were just designed for splash, not submerging.

Video (http://www.youtube.com/watch?v=wCqfSBcm6l4)

[exFSAE]

That's very interesting! Where do the clutches live? In what would be that white area? Or do they go out further into the "ring of oil" ?

Though in general I've been under the impression Salisbury clutch plates do tend to wear out.

Edit - Kinda wonder if there's enough room in there to put some sort of fins or baffles? Something to keep the oil from doing that.

[Wesley]

Yeah, the clutches are right in the dead space, placed concentric on the axis of rotation. We thought about fins and baffles - but since the oil isn't moving in relation to the case, it wouldn't move in relation to the fins either - the velocity profile would take longer to become uniform in the oil ring, but other than that, the same thing would happen.

[jaca]

woah, that is interesting, I never considered the forces acting on the oil

fortunately the gears in the torsen protrude from the case a bit, so hopefully we weren't running it dry... then again those things are bullet proof in FSAE application

the only simple solution I can think of is running the diff full of oil, maybe with a smaller case volume

have you tried running the diff fully wet? Intuitively at least, if you have exactly the right amount of oil, the centrifugal force should keep it pinned to the outside of the casing while still giving good internal coverage with very little pressure on the seals. That's in a straight line though, cornering might be another matter.

[vreihen]

I'd be curious to see what fibrous grease did in that test rig in the same environment. My guess would be that it would sling to the outside and not return to the center once the rotation stopped. Not that it would do any good on Salisbury clutches, but there have been debates in other circles about the merits of packing a Torsen-type differential with grease versus gear oil.....

[fixitmattman]

The problem is that since your housing is also rotating with the diff you've effectively built a centrifuge.

I'm not sure if you're familiar with oil flow or oil slingers in a manual transmission. One common approach when hollow layshafts are used is to insers a spiral shaped piece of sheet metal into the hollow section of the shaft to act as an auger to feed oil from the open end. If you take that idea and put some ribs/ridges on the inside of of the housing it "should" produce a similar effect and generate enough pumping action to throw the oil against the side housing it's directed at and cause significant splashing. I'm sure if you used a chevron pattern you could get it to pump the oil in the middle if that's where you need it.

[Superfast Matt McCoy]

Are you using a synthetic oil?

[Conor]

Have you tried mixing in some Lucas? We used it in our diff with success. Our issue was more related to sealing, but it worked great when we ran it. I think your validation method should be taken with some caution as there will be a great deal of internals for the lube to stick too as well. Just some suggestions.

[Yellow Ranger]

We are sponsored by Lucas and use their synthetic 75-90, although that was Castrol synthetic in the experiment. This was a very quick and dirty test- as you can see by the acrylic 'diff case' and the cross drilled PVC 'diff carrier'. We wanted to see how quickly the diff would have to spin before the oil would stick to the case. Of course, now we could make a more realistic model to see how full we need to fill the diff so that the clutches are lubricated.

Has anyone had issues with leaking differentials from filling them too full? Like jaca said, I wonder how they'd hold up under corning...

[Conor]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Yellow Ranger:
We are sponsored by Lucas and use their synthetic 75-90, although that was Castrol synthetic in the experiment. This was a very quick and dirty test- as you can see by the acrylic 'diff case' and the cross drilled PVC 'diff carrier'. We wanted to see how quickly the diff would have to spin before the oil would stick to the case. Of course, now we could make a more realistic model to see how full we need to fill the diff so that the clutches are lubricated.

Has anyone had issues with leaking differentials from filling them too full? Like jaca said, I wonder how they'd hold up under corning... </div></BLOCKQUOTE>

Certainly have. Make sure you have some good seals. We used rotary shaft seals that held up decently, but tended to puke a little from plunging during heavy cornering. I would suggest the type that has a circular spring.