Thursday, October 23, 2014

Peanut Butter (no Jelly) Time

No, I won't post that dancing banana video from a few years back, but I will reflect on a valuable use for peanut butter that I didn't expect.

That's Nasty
old pic from 2011 but
check the nasty back-end
When driving along, the air directly behind the vehicle creates a vacuum.  Car designers have worked all kinds of miracles in that area over the years, but with the old loaf-of-bread shaped VW bus, the vacuum is quite significant.  I've heard of water-cooled transplant failures caused because the builder put a radiator where the engine hatch is.  The vacuum made the radiator useless, unless the builder had the fans suck air in from behind, sort of leveraging the vacuum.  Even then, those installs look pretty atrocious and would pull all kinds of crap into the radiator fins, but I digress.  The vacuum grabs up water, road grime, dripping oil... and splats it on the rear bumper, engine hatch and rear hatch.  Over time, it gets pretty grimy.

In my bus' case, it was worse.  I mentioned the transaxle gear oil leak in previous posts.  That was bad.  On top of that, for a few years I was running the original tail lights and the old housings failed.  I was broke, so I held the tail lights on with duct tape.  Yes, the bus looked horribly ghetto.  I scored a pair of new (but not China-crap) housings from the BusDepot when they were on sale a few years ago, and replaced them.  I did not, however, clean up the old duct tape sticky residue left behind by the tape.

For the Greatest Shine You Ever Tasted (from the SNL Shimmer skit)
While cleaning out the garage, Boo came out from playing in the garden and started asking about the nasty back end of the bus.  We realized that once the newly painted bumper is put on, it will look even worse.  She ducked inside, and I went back to the garage.  A couple minutes later, she had returned with a fist of paper towels and a teaspoon of peanut butter.  "Hungry?" I asked.  "No, haven't you ever done this before?" she replied, and then proceeded to put a dab of peanut butter on the nasty duct tape.  "uh... what... are.. you..," was as far as I got as a response before I saw one of the 4 tacky squares which had been left behind by the duct tape had disappeared.  In a few minutes, all 4 of the big black marks were gone.  I was converted.  I left the garage in its not-yet-organized state and grabbed some peanut butter.  I polished the whole rear end with that magical stuff.  It acted like a cleanser-wax, cleaning the oil and grime while leaving a shine (presumably from the oils in the peanut butter).  Of course, there are probably lots of little peanut bits in there which would trash an otherwise nice paint job.  For original paint, complete with rust, patina and grease-finger prints, it is magic.

For kicks, I peanut oiled all of the little rust spots I had.  The petina-style rust didn't seem very affected.  Rather than a dull spot, it shined like the paint, but otherwise, it was about the same.  The rust which was a little deeper into the steel, though, did react differently.  The rust closest to the paint was unaffected, but as I moved more than a couple MM away from the paint edge, the rust turned silver.  Trippy.  I'll re-check those spots in a couple of days to see if that lasted or was just because the peanut butter picked up some of the body color and it simply attached to the rust a little bit.  Regardless, the experimenting was fun.

Like so may other times, I didn't take many pictures beforehand, but I do have one here of the back after it was cleaned.  Wow.  And there's a shine to it now.  Yes, the paint is still atrocious, but that's part of the ongoing work-in-progress.  The journey is the destination.

Of course, it rained all night the next night.  It will be interesting to see how long a peanut oil wax job will last in a NorthWest autumn (read: nearly constant rain or mist in the air).  Today, it really looks great.

Monday, October 20, 2014

Ceiling Sag No More

Quick post today.

Well, That's Ugly
I recently realized that I've owned this bus for more than 11 years.  Wow.  Where did the time go?  A few years after I bought it, I decided that the 1972 Westy interior was a safety hazard for my (then quite young) boys.  I played this scenario out in my head where we got in a wreck and the boys, while unharmed, couldn't reach the sliding door handle behind the sink/icebox unit.  This lead to removing the sink/icebox, and, eventually, removing the entire early interior, replacing it with one from 1979.  Now, the 1979 interior assumed that the ceiling over the rear section of the bus was flat, whereas the 1972 ceiling was rounded.  Because of the difference, I couldn't install the '79 headbanger closet.  The 1972 storage shelf assumed that the closet was on the right side, so I couldn't reuse that either.  So, the net result was no shelf, and the ceiling was left bare... and unsupported.  Fast forward 5-8 years and the Baltic Birch sheet no longer clings to the upper steel sheet like it used to.  Quite the contrary, it droops pretty badly.

Now, That's Better
The solution was really quite simple: sheet metal screws.  I had put one in the center at the very rear a while ago, but that just held off the inevitable.  I used that as a starting point, running a sheet metal screw up into the ceiling, through the ceiling steel sheet a foot apart, front to back.  I set a screw on either side about a foot to each side as well.  While this isn't a final state either, it no longer sags.

Ideas on Final State
I've thought about what the interior of the bus should look like, and I've trolled around the internet a little bit looking for ideas.  I think it comes down to 2 options: buy automotive headliner material in bulk and install it with epoxy/glue -or- do something custom.  I think by now, you know I'll do something custom, its just a question of what.  To give the headliner a fair shake, though, it is the right material for the job, it's anti-microbial, non-flammable (even flame retardant) and comes in a variety of colors and textures.  My challenge is how it adheres to the Baltic Birch.  If it requires solely epoxy, gravity will be its undoing.  Sending staples through it to help hold it in place will ruin it.  Hmm.. more thought and research needed.

That's all for now.  As always, thanks for following along-

Wednesday, October 15, 2014

Rear Bumper Clean-up

On the drive to the coast for the family gathering for Independence Day, I realized that something in the drive-train was leaking oil.  Drips would hit the air turbulence under the bus, atomize, and the resulting mist would get caught in the vacuum behind the moving bus.  Within the vacuum, it would splatter against the rear of the bus; especially the rear bumper.  Over the last few posts, I explained how I solved the leaking oil (it was a bad input shaft seal, damaged by a faulty oil slinger, which was either damaged because I didn't have a pilot bearing in my flywheel or because AA transaxle used an old bell-housing).  Now that the oil leak is solved, I can focus on the last thing I wanted done before the Fall (see Setting and Re-setting Expectations): cleaned up and painted rear bumper.  Today's post covers the start of that journey.

The Missing and the Modesty Skirt
scrub clean first
Like the bolts holding the front bumper on, the square-lock bolts holding the rear bumper are in pretty bad shape.  Rusty.  Painted over with what looks like interior paint.  Missing.  I'll need at least 3 replacements when I'm all done.  At some point during the engine swap project, I removed the splash pans from either end of the rear bumper.  I'm sure that was a good idea at the time, and it gave me the opportunity to troll through my garage looking for them in my boxes of parts.  Like a trip down memory lane, it took longer than it should have, but I found them.  I set them aside and removed the modesty skirt from the bottom of the bumper.  I don't know why the bus has one of these; and I don't see any on the handful of buses I regularly see.  Maybe 1972 was an especially shy year for cars.  Anyway, I hammered it semi-straight-ish.  Then, I taped the front, bottom and sides and put it with the splash pans.  On a sunny day, I spray painted the pans and the unmasked front of the modesty skirt with rubberized undercoating.  Once dry, I removed the tape from the modesty skirt so rest of it could be painted white like the main bumper.

Sanding and Banging
Unlike the front bumper which had 3 distinct pieces, the rear bumper is a long stretch of curved steel.  The curves run top to bottom as well as curl in on the ends.  This made banging it straight more challenging.  Again, like with the front bumper, I aimed for 'good enough for camping kwality' in my bodywork.  Unlike the front bumper, the rear had a couple of very old bumper stickers on it and when they were removed, they took a couple coats of crappy paint with them.  This led to some unwelcome contours I tried to solve with 150-grit sand paper.  I learned along the way that the sand paper could help me identify minor dings by bringing up a lower level of paint around an area.  It was with this technique that I discovered that the bolt holes were centered on dimples the size of US quarters.  Some more banging with the hammer got them flat.

Paint, Wet Sand
undercoated
Once I declared "good enough" on the cyclical sand, hammer, sand process, I pulled out the white Rustoleum I used on the front bumper.  It's been about a year since I did that front bumper, so its time to review how it is holding up.  To be fair, the bus hasn't seen many miles (probably less than a thousand) since it was done, but it has been outside in my driveway where the kids are playing when it wasn't driving to music festivals.  The bumper looks good, and I mean really good.  No rust staining, no blemishes, no cracks or chips.  I applied straight Rustoleum with a foam brush inside and out and set it to dry.  The next morning, it was totally dry.  Total drying time was over 12 hours, but some ranters on the interweb claim it takes forever for this paint to dry.  False.  Just apply it properly: spread a thin coat always keeping a wet edge by brushing into the already painted surface.

Spackle?
After the paint set up, I could see large sections where the paint underneath had been pulled away by bumper stickers.  In a real body shop, they'd probably strip the whole bumper down.  I thought about that.  In a cheap-o body shop, they'd fill with Bondo.  I thought I could experiment with spackle, since that was what I had, and see how it behaves over time.  Worst case, it looks like crap and I strip the bumper down later.  Big deal.  First, I wet-sanded the paint.  I needed to do it anyway, and I figured that would give some teeth for the spackle.  So, with a basic putty knife and a small tub of spackle, I smeared paste over the bigger spots, and some small ones.  I slipped inside and watched some football for a couple hours while it dried and then cuffed the spackle down with 150-grit sandpaper.

More Paint
Once smooth, I wiped the bumper down with a damp paper towel and applied a second coat of white Rustoleum.  Same method, similar results.  This time, though, you could tell where the spackle spots were because the white paint wasn't shiny in those spots.  At this point, I figured I probably should have done something else, but I just plowed ahead: wet sand with 320 grit, another coat of paint, more 320 grit wet sanding.  This time, you really can't tell where the spackle is, but I wasn't 100% sold.  SO, I decided to wait a week and look again next weekend to see if another coat of paint would be needed, would help, or if I'd just be throwing product at a bad idea.

Sunday, October 12, 2014

Back Together, Test Drive

In my last posts, I covered the hazardous re-assembly of the engine-to-transaxle and the install of the combined unit at 15*.  Today, completes that journey.  In order to do the re-assembly as quickly as possible, I tried doing it a little differently.  I usually do system by system.  This time, I did it based on where I was and did everything I could before I moved.  I'm not sure I'll do that again, but it was interesting.  I cleaned every part with brake cleaner before installing it.  It made a world of difference, and I'll do that every time now.  Handling a clean part is just so much better than the oily, greasy alternative.

Lower Right Rear
Starting below the right side of the engine, there are a few things to do, starting with the CV joint.  I place a jack under the lower shock-absorber mount and lift until the tire just barely leaves the ground.  Then, with 6 new Allen-head bolts (one striped last time), attach the CV joint to the side of the transaxle and drop the tire back onto the ground.  At the front of the transaxle on the right side, is the ground strap.  It requires a 13mm socket.  Install the starter.  It requires a 17mm socket (need to verify) for the 2 bolts plus a 13mm for the B+ cable from the battery.  Plug in the trigger signal wire to the starter.  Connect the B+ wire to the Atlernator and plug in the sensor wires.  Double check all the other visible sensor wires (like on the cooling system) are plugged in.  Last, zip-tie the cooling pipes up out of the way from the axle.

Rear
Put in the rear support bar.  Don't forget to put the rear bumper mounts back in when you do the bar; they use the same bolts.  Install the support tower and the rear engine-frame mount.  If you get this far and the mount isn't lining up, check the support bar install.  If the mount and tower are on different angles, you probably mated the engine to the transaxle at something other than 15*.  Go back one posting :)  Once the rear mount is in, hook up the exhaust.  The muffler pipe has a hook that fits onto the rubber mount which is hanging off the right side of the rear support bar.  Once hanging, it should fit with enough room to maneuver the rest of the pipe without it hitting the ground nor the bus and no jacking required.

The exhaust-to-turbo mating has a gasket and 3 13mm nut/bolts.  One mount is a stud coming out of the turbo.  Because of the design, at least one of the 13mm needs to be tightened with a wrench instead of a socket.  Next comes the intake system.  Start with the non-charged air pipe from the air cleaner to the turbo.  This is held on with 10mm bolts, and most of the work is from above rather than through the rear hatch.  Next, the charged-air plumbing with the intercooler is most easily installed as a unit.  Slide the turbo-end on first, then set the intercooler and complete the system to the intake last.  The pipe clamps are a combination of slotted-screwdriver and 10mm socket.  The intercooler is held in place with bailing wire (yes, bailing wire) looped between the mounting hole on the intercooler and the open loop in the engine mount attached to the engine.  This allows for the vibration of the entire system to resonate as one.  Route the vacuum to the turbo and snap it into the channels on the plastic tubes.

Lower Left Rear
Shifting back under the bus, repeat the CV joint process for the left side.  Verify what's been done thus far from the new perspective: the turbo plumbing, the air system, the vacuum.  Hook up the bowden tube and the clutch cable to the clutch activation arm.  If you run out of adjustment threads, you need some head-washers (I used 4) between the bowden tube and the mount it slides into.  That mount is held on with 2 13mm bolts.

Lower Middle
Mount the nose of the transaxle.  Plug in the reverse wires for the reverse light switch.  Don't brain yourself on the radiator.  Thread the short hollow tube to connect the gear selector at the front of the transaxle to the long gear shift tube coming from the front of the bus.  Zip-tie the tube into place.  I'm sure that last line will upset some purists (as if the whole project hasn't already), but the original bolt/nut solution falls out; the zip-tie doesn't.

From Above
There's little left now.  Mount the coolant overflow bottle, verifing the top coolant line runs where it's suppoed to.  Plug the sensor into the bottle.  Note the coolant level, and come back and check again if its not spot-on.  Verify the vacuum bulb is set and tight.  Put the fuel filter into it's mount and tighten it down, making sure that the lines run properly.  I have a cheap $1 clear filter in front of the spendy stock filter to extend its life, but it also shows me that fuel is getting that far.  I check that for a fuel level at this point.

All Over
Last, the battery is hooked back up.  Before that, all of the electrical connections, including all sensor plugs should be double checked.  This is where the "do everything you can see" model falls apart.  By doing it one system at a time, you know everything was done.  By working zonally, you don't really know for sure without checking system by system.  At that point, you may as well have just done it that way.

Testing
This is the part everyone wants to jump to: grabbing the keys and giving a test fire.  The first time I tried this time around, I got nothing: I'd failed to hook up the battery.  Once resolved, he started right up.  I applied the clutch and shifted through the gears.  Nice and smooth.  Seatbelt on, and on the road to test.  There was less noise than I remembered.  The gears shifted easily up through 3rd (never got to  4th), and the bus felt peppy.  Overall, it was a great, albeit short, drive.  I declare the bus road-worthy again.  By the way, there wasn't a single drop of oil on the transaxle nor engine after the test drive and cool-down.  Success!

Thanks for following along.  Next time, I look at the ceiling of the bus and/or knocking the dents out of and painting the rear bumper.

Tuesday, October 7, 2014

This End Up

When I first constructed the engine cradle, it wasn't really designed to be used a bunch of times.  I never thought that far ahead.  Today's post covers what happened when a hot and bothered shade-tree mechanic experiences a failure with a critical tool... after first losing his patience.  I got so caught up in solving the problems that I failed to take any pictures.  So sorry!

Engine Husbandry
A couple of years ago, I mated the transaxle to the engine under cover of the bus in my real estate agent's drive way (See Transaxle Transition).  I figured, "if I could do it in MS' drive way, I can certainly do it in mine."  Sure.  Like before, I removed the mating studs first.  Unlike before, I had the pilot bearing in the flywheel, so the mating was much harder.  With a small wheeled 2-ton jack holding the transaxle up, I was not able to maneuver the input shaft through the pressure-plate, clutch and flywheel.  Instead, I did what my old wrestling coach once described to me.  I put the transaxle on my belly/chest and heaved it on, twisting it until the splines of the clutch fit into the input shaft, and then slid it home.  I slid myself out from under the transaxle and slid the jack into my place to hold it together.  I don't want to do that again.  One more reason to cut a removal valence into the rear apron.

Studly
With the transaxle and engine mostly mated, I rotated the transaxle such that the stud holes lined up, and slid the bolts in.  At the time, I'd forgotten that the adapter plate had 3 settings: 0*, 15* and 50*.  I nutted the studs down, and rather than take my winnings and going home, I doubled down and tried to get the bell-housing mounts set before calling it a day.  The driver side is easier to get to, so I started there, and after 10 minutes of up/down with the jack, I was able to slide it through the hole, the mickey-mouse mounting ear and against the nut.  Then the trouble started.  I couldn't get it all the way through to the nut, nor could I get the other bolt started at all.  Jack up, jack down.  Shake engine.  Repeat.  30 minutes.  40 minutes Nothing worked.  On one of my attempts, the engine shifted on the cradle.  Instead of leaning over to the left, it was leaning over to the right, and partway off the cradle.  The cradle was also showing signs of weakness, with it separating in a few places.

Is It Bleeding?
At this point, I kind of freaked out, but it got worse.  I reached in through my home-made top-side engine hatch to pull on the driver side mounting bolt.  Rather than grab the bolt, my head hit the hinge.  Hard.  Sudden pain and blood ensued.  Instinctively, I slapped my hand against the gash.  Ordinarily, that would have been good; pressure on the wound and all that.  My hand, though, was wrapped in a grease/oil covered glove, so the cut now had oil and grease in it.  Fortunately, Boo was gardening near by, heard my curses and came over to see what was going on.  She snapped into nurse-mode, cleaning it out with various potions and butterfly taping the gash closed.  Once calmed back down, I went back out to survey the wreckage.  The engine was shifted clockwise by a couple hours, and from underneath, it was clear the transaxle wan't in-line with the centerline of the bus anymore.  I resolved to simply get the one bolt out and call it a day.  20 minutes later, the whole operation was sitting on the ATV jack and cradle, though woefully mis-aligned and partly rolled over.  I cracked a beer and left it for the night.

Ship-Righting
Getting the engine/transaxle upright was a series of "how about this?" ideas.  I had the cradle and jack somewhat square under the engine, but simply tugging on the engine didn't really budge it.  I tried nutting-down the nose of the transaxle, but that didn't work.  I placed a steel bar across the top of the engine hatch and used load-lock webbed hooks to attach the engine to the bar.  I then slowly lowered the jack.  The engine shifted.  I moved the jack, supporting it in its new orientation.  I used the other jack under the cooling fins on the right side of the transaxle to help it rotate.  I changed the ATV jack so it was perpendicular to the engine.  This proved to be most effective: leveraging the steel bar/hook and the secondary jack, I settled the engine onto the ATV jack.  Then, while pushing the base of the oil pan with my foot, I pushed down on the turbo.  The engine rotated while the ATV jack moved to the right.
After a series of similar moves, the mickey-mouse ears seemed aligned with the holes in the mount.

Engine, Meet Hatch Lid
Once the engine was upright, I slid under the bus and got the passenger-side bolt in.  Note for my future self: always start on this (passenger, right) side.  If the engine flops over again, better that it flop onto the turbo-side: you can put a jack there.  There's no good spot to set a jack on the right side.  After about 15 minutes, I had the bolts in, and tightened down.  Probably a personal record.  Once in, I thought the engine looked a little funny.  The intake felt high and the oil pan appeared to be pointed at the ground instead of slightly to the right.  My fears were confirmed when I set the engine hatch on top of the intake.  Drat.  now we know that the 0* adapter plate configuration would not have fit.

Re-Studly
After all that, I wasn't going to start all over again.  Instead, I thought I'd try rotating the engine 15* by removing the transaxle mounting studs.  Since I'd already installed the starter, that had to come out first.  Then, I re-supported the engine with the cradle/ATV jack, set the steel bar with the hooks and removed the 4 studs.  I repeated the pushing on the oil pan, intake and turbo like I did before.  The engine shifted somewhat quickly into an almost-there position.  I had to do the final little bit while laying alongside the left side of the engine.  This was arguably pretty hazardous.  I slipped the lower left stud through the transaxle and pulled down on the turbo while wiggling the stud.  The engine needed to shift about 1/4", but it sure felt dicey.  Once the first one was in, I knew it was aligned, so I slid the upper left stud in and dashed around the right side and did the starter-stud and lower right studs.  Once nutted and torqued, I dropped the ATV jack, and re-set it with the cradle.

It was a crazy couple of days getting the transaxle and engine roughed back in.  I still had the transaxle nose-mount, CV joints, rear engine mount and all of the accessories to do.  I'll post on that later.  Fortunately, there was virtually no more drama after I got the central mount done.
Thanks for following along,

Wednesday, October 1, 2014

Snap Ring Pliers review

I don't usually review tools.  Usually, I research a bunch and then pick whatever the interweb tells me to buy.  I wasn't able to get a good direction from the web for Snap Ring Pliers, so I bought a couple pairs and experimented.  Today's brief post summarizes those experiments.

Why Bother?
In my previous post, I talked about solving a leaking oil seal.  Some of my research on that issue pointed me to believe that the input shaft was loose and needed to be removed and re-installed.  To do that, a circlip needs to be backed off so a thick lock-ring can be pulled back (see picture to the right. The input shaft rises from the center of the picture to the left).  Then, the input shaft can be threaded off the reverse gear.  First step in that process requires a pair of Snap Ring Pliers.

Size
Some things about the old Volksies are very well documented on the web.  Arguably, some are over documented.  If you want to know what kind of oil to run, for example, you can go through literally hundreds of opinions.  Tires are equally well-opinion'd.  Getting something as simple as the size of the tip needed to remove the circlip on the input shaft, however, can sometimes prove impossible.  Hours of searching netted no useful information.  The size is 0.07 (US), by the way.  This is the largest size available in the standard multi-tip Snap Ring Pliers sets, and individual pliers with fixed-size tips can be found.

NAPA
boink! tip fail.
Off to the friendly local auto parts store (FLAPS), I went to find a pair that would work.  Since I'm only really doing this once, I didn't want a spendy pair, just some that would do the job.  So, I tried the tip-adjustable set available at NAPA.  Using the largest tip (0.07), I was unable to open the circlip wide enough to get the circlip out of its seating channel without the tip failing.  The tips are held on with small Phillips head bolts pressing a removable plate against the plier arm.  The removable plate is slightly bent, presumably to best fit the tip on, but the engineering is flawed such that the final bit of torque needed to hold the tip is robbed from you by the curve of the plate.  Neat.  I tried to undermine that torque-robbery by jamming a finish nail under the plate on the opposite side of the bolt, increasing the torque on the tip.  That didn't work either.

mod no worky
Net-result: not good for this job.  Maybe, if the circlip you need to remove requires less than 10 foot-pounds of torque it would suffice, but for real automotive situations, its junk.

Husky
After setting the NAPA pliers aside, I hit Home Depot looking for a pair of Husky one-size-only 0.07 Snap Ring Pliers.  While their web site showed that they were in stock at the store (for $13), neither I nor the clerk could find them.  I'd love to review them, but instead I offer just a head-shaking at Home Depot.  Either offer the product or don't; don't post it on your web site if you don't have it.  Boo.

Channel Lock
Channel Lock works!
Home Depot did have Channel Lock 927 Snap Ring Pliers (for, like $23).  These are multi-tip pliers like the NAPA ones, but are clearly better made.  They are much heavier.  They have a flip-switch to set inner or outer ring direction.  Oh, they're made in the US, so somewhere my countryman benefited from my buying them, so that's nice.  But do they work for this?  Why, yes, they do.  Unlike the NAPA pair, the ring easily opened up enough to slip out of the channel and out of the way.  Like the NAPA, Channel Lock 927's are sold with a collection of smaller tips, and are delivered with the largest (0.07) already installed.  This made the test easy, removing the possibility of me putting the tips in wrong.  I didn't bother removing and re-installing to see how bad it was, but the engineering is very different.  Unlike the NAPA which has a plate screwed on, leveraging the friction created to hold the tip in place, the Channel Lock has a hole in the plier arm to slide the tip into.  The securing bolt simply prevents the tip from falling out, so the arms need to be able to support far less torque.  The arms are much thicker too, implying that they can handle far more pressure than the NAPA tool could.

In the end, I didn't need to remove and re-install my input shaft.  After I installed the oil slinger and input shaft seal (see: Transaxle Re-Assembled), I re-tested the wiggle in the input shaft.  It barely moved.  I concluded that the input shaft was appropriately loose, and some additional digging into theSamba verified that conclusion.  I'm keeping the Channel Lock's with the car tools, and returning the NAPA's.

That's it for today.  I'm working on a post summarizing the engine-trans re-mating as well as the adventure of raising the engine-tranny unit back into the bus.

Monday, September 29, 2014

Transaxle Re-assembled

Last weekend, I got the transaxle put back together and back into the bus.  Fast sentence to write.  Lots of effort to do.  Today covers the first half of that 'effort to do': getting the transaxle ready.

Slinging Oil
slinger by hole in bell housing
where it's supposed to be attached
The bellhousing for the 002 and 091 transaxles are supposed to have this ringed collar around the input shaft to help guide the hypoid oil away from the input shaft seal and back towards the gears.  There are a few names for this bit, but the boards call it an "Oil Slinger".  According to the Bentley, this was supposed to be a non-repair item that was pressed into the bellhousing with a 4-ton press.  If it separates, you get a new bellhousing or re-press it in.  Well, the VW world has discovered that's not exactly right anymore.  Instead, you create a rough edge on the inner edge and along the rear (engine side) of the bellhousing.  I used a chisel and a file for this.  Then, rough up the lip and outside of the slinger near the lip.  Peanut butter the edge with JB Weld and press it into place by hand.  I treated the excess like caulking, and ran a gloved finger along the edges, pressing it in and making the edge clean.  Last, hold it firm with a simple weight and gravity.  Once its cured (18 hours), its ready.

Install Input Shaft Seal
roughed up edges
This was actually quite simple.  Lightly brush the inner and outer edge with hypoid oil.  Set into place by hand, and press it enough by hand so it sits still.  Find a short section of 3/4" PVC pipe (I used a 3/4" - 1/2" T from Home Depot).  With a rubber mallet, smack the seal until it seats flush with the face of the bellhousing.

Into One Piece
When the transaxle was built by AA Transaxle in Seattle, the bellhousing was sealed to the transaxle body with what looked like clear silicone caulking.  Not a fan.  I scraped all that out and instead ran hi-temp form-a-gasket along all of the mating surfaces.  Once it becomes tacky, set the 2 sides together and set the bolts in.  Rubber mallet the bellhousing into place and finger-tighten the bolts.  Following the jump-the-center torquing process, set them to 18ft-pounds (if memory serves. check the Bentley).

Topped Off and Ready
slinger with JB Weld
Before I put the bellhousing on, I made sure the oil drain plug could be loosened.  I had vaguely remembered that it was frozen in-place.  I was right.  After some hard torquing, I got the nut out.  I cleaned off the tiny metal filings off the attached magnetized rod, cleaned the threads and re-seated it.
Once the transaxle and bellhousing were one piece again, you need to give the form-a-gasket time to cure.  I gave it 24 hours and then opened up the fill hole.  It took all the gear oil I had on hand, and was still a touch low.  I'll need a fresh bottle to top it off before I take much of a test drive.  Remember, the tranny is full when oil spills out of the fill-hole when the tranny is on level ground.

That's it for today.  The tranny looks good, the input shaft has almost 0 wiggle to it and the seal / form-a-gasket are holding oil.  Hazah!