Ive got the floating seal installed. I ended up converting to grease for the toric ring lubricant as the manual describes. The seal is springy concentric and has less than 1mm stack height difference around the circumference. Because there's grease on the o-ring its very easy to see it roll as I deflect the seal face. I don't see any drop of the toric seal as I deflect only a distinct roll in and roll out as it springs back. It appears that in this assembly the issue of slipping vs rolling is not occurring, at least at this stage. The range I'm deflecting it seems beyond any drum deflection and wobble it would encounter in service. There may have been some recession or slipping initially as I seated it but now I only see a roll and unroll with no t ring migration. In fact the only way I can get the t ring to recess is to push in one spot with the blunt tool and shove it off the backside of the seals lip/ramp, this is somewhat analogous to prying the seal out of the T-ring. Not having fit any other seals i cant say for sure but it seems this assembly is not suffering from an issue of sliding into the housing and not returning. It brings up some questions though:
1) Could I get this result only to have it migrate later? Part of the reason I used grease is to encourage the seal to migrate to wherever it was inclined to go without any housing drag or friction. I wanted to make it fail and I couldn't and it seems entirely captured between ramps from rest height to deep compression.
2) If I understand the gear box correctly there will be oil pooled and splattered behind the ring in service and the t ring will roll in and out with drum and bearing deflections. Isn't it expected that the t ring will be entirely wetted on the inside and rolling back and forth in this oil with some inevitable creep of oil to the outside? Maybe the dry/alcohol fit is only an installation issue but not a service issue?
3) I went to bring the two parts together and was surprised how far apart the bearings were with the seal supporting the drum weight. My initial thought was- why did you buy non dealer parts. Reflecting a bit, well you have to have some seal load and buying expensive dealer parts is an assurance but no gurantee my stack height is correct. After some searching I came across this:
Tips on how to replace your excavator final drive seal. These seals are variously known as floating seals, lifetime seals or duo cone seals.
finaldrive.eu
Basically it says the combined part separation and spring force of the seal would likely prevent starting the bearing nut and drawing the parts together if this repair were done on the machine without the assistance of gravity and weight of the assembly compressing the seal.
I am working without aid of a hoist so the first thing I did was ditch the sprocket and still could barely carry it to the van. Now with the lid off and gears out I can easily manage it on a table by hand but only have around 50 lbs to compress the seal and had estimated another 100 lbs would be required to bring the bearings together. So perhaps this separation is normal? I haven't calculated the seal contact force but could estimate it with a bathroom scale, I'm guessing 10-20 psi at the seal face?
Considering the need to exclude dirt and resist pressure differences this may be totally reasonable or even low. Is the gear case really unvented or have I just not found it?
4) Another thing I found useful from the article above was guidance to fit and check the bearings before the seal install so you can feel or measure the bearing condition. Even with the 35 yo seal the drag made it hard to assess the bearings before disassembly so Ill back up a step and check bearings even though there original in excellent condition.
I'm guessing 3 and 4 is common sense to people experienced in this operation but 3 unhinged me for a while. The weight, volumes of oil etc makes me really want to avoid redoing this, its bad enough that the other side may be coming soon. I will do the brake seals on the other side before I start using it again. The segmented breaks in the t ring seemed odd to me but 92U pointed out that's typical. Seeing the rolling action I can now see why they fail that way. I guess if your unlucky the break is at the bottom and instantly drains the case otherwise you might get a partial loss and more time to notice the leak before you destroy things, that may be what happened here.