Project Silvia – The official guide to not screwing up your wheel bearings like I did…
by Dave Coleman
The first one died after an ordinary day at the Streets of Willow Springs. Rolling to the trackside gas pumps to satiate the car's bottomless thirst for petroleum, there was a nasty growling noise. Turning right made it go away. Turning left made it get worse.
There really isn't much more diagnosis needed to identify a bad wheel bearing, but these bearings were not only new (less than 10,000 of the car's 300,000 miles had been with these bearings) but they were the bigger S14 bearings that should easily be able to handle anything this little S13 could throw at them.
The bearings weren't worn enough to feel loose when rocking the wheel, or to cause any change in brake feel (loose bearings can let the rotor rock relative to the caliper, pushing the pads back into the caliper and adding slop to the brake pedal), so I pulled off the right front wheel and caliper in an effort to spin the bearing without any brake pad drag. Spinning the rotor made a distinct growling noise, and the motion actually felt rough if you held your hand on hub while turning. It doesn't get any more clear.
A few months later, Mazda Raceway Laguna Seca, and the familiar sound reappeared, this time from the left side. The bearing issue could no longer be shrugged off as a fluke. S13 bearings are relatively cheap, being standard bearing cartradges that can be removed form the hub, but replacing an S14 bearing means buying a brand new hub. At over $150 a pop, periodic bearing replacement is not an option.
Follow the camera, then, on a journey of discovery. I finally figured out what was causing the problem, and made the new bearings even more indestructible in the process.
The first sign of trouble (after the noise) was when I pulled the hub off and found this. You aren't supposed to be able to see the inside of the bearing when you pull the hub…
And the stub axle isn't supposed to look like this. The inner bearing race had fuzed to the stub.
Removing the race without damaging the stub was no small feat. After cutting a groove into the race with a cutoff wheel, soaking everything in penetrating oil, and heating the race with a propane torch, I hammered a screwdriver into the groove, at an angle, and got it to rotate on the stub axle. After rocking it back and forth a few times, I managed to walk it off the back of the knuckle just enough that I could get a screwdriver in behind it. From there, two screwdrivers wedged in opposite sides could work together to slide the race off the stub. Throughout this whole process, it's incredibly important not to put a nick in the stub axle. Any nick will significantly weaken it, and you really don't want this to snap off…
Popping all the balls out of the hub and cleaning away the grease revealed damage to the outer bearing race. Matching marks were on the inner races. It's worth noting that there isn't any unexpected discoloration any part of the bearings. Blue balls would suggest overheating, but the galling shown here, especially in evenly-spaced intervals, suggests damage done when the bearings were stationary. The bearings failed because the car was parked for two years… That shouldn't really happen…
Now to prep the new bearings for a longer life. The first trick is to get the sealed bearings apart. If you just stick a screwdriver into the crack between the inner and outer bearing races and give it a gentle tap, the race should pop out.
The grease Nissan uses isn't terribly heat resistant. My bearings didn't fail from heat damage, but the condition of the grease in the used bearings suggests that lubrication-related failure wasn't far away.
So, to get everything apart and get the waxy stock grease out of there, use that same screwdriver to gently pop the ball bearings out one by one. Wipe each one clean and set it in a clean bowl or something you aren't going to drop. If you lose anything during this process, you're buying another $150 hub.
Repeat the process on the other bearing. When you pull the plastic bearing cage out (just rotate it 90 degrees and slide it out), be careful not to break any of its fingers. Its a surprisingly delicate piece.
Now, the magic goo. NEO sythetic wheel bearing grease comes highly recommended by several race teams I've spoken with, including the guys at Stoptech, who really like dumping a lot of heat into their hubs. It's expensive, but cheaper than another hub.
After reassembly, the bearing races will pop back in place with another gentle tap from the hammer.
Ok, so what happened? Why did the bearings fail just from being parked for two years? I finally figured this out when re-installing my second replacement bearing. I've been using an S13 service manual, but these bearings are from an S14. S13 wheel bearings are torqued to 108-150 ft-lbs, while S14 bearings are supposed to be torqued 152-210 ft-lbs. I had been torquing my S14 bearings to S13 specs.
Uner-torquing a wheel bearing changes the load path through the bearing, concentrating the weight of the car on a smaller piece of the interface between the balls and the bearing race. This probably made a series of little dents in the race, which then caused a shock load every time a ball rolled past and fell into that dent. That shock would over-load some other spot on the bearing, causing another dent, and so on and so on.
That's the lesson for the day, kids. RTFM.
|Go back to the main story->|