Bad wheel bearing

Unscheduled maintenance on a second-generation Prius?? Unthinkable!

With about 190,000 miles on mine, though, one needs to accept the idea that anything could happen. In this case, it was a front wheel bearing that decided it had taken enough pounding for one lifetime. This leadup text is largely lifted from some posts I made to the Prius_Technical_Stuff yahoogroup while tracking the progress of the issue over the course of a week or so.

Timken bearing image  (swiped from I was on the way to visit relatives over the holidays, and realized about a third of the way into the trip that the right front wheel bearing was definitely on the way out. More groaning when I turned toward the left, and as it developed further when I stopped to check it I could feel a tiny little bit of play when shaking that wheel in and out at the top. Definitely toast, nothing really for it except replacement. This bearing probably didn't owe me a dime at this point because I had frankly put it through hell, but on the first leg of a long roadtrip was sort of an inconvenient time for it to start failing for real.

Oddly, the standard bearing check of jacking the wheel up and trying a top-to-bottom shake *didn't* let me feel any play, which is one reason I started the trip on that bearing in the first place -- I had suspicions maybe a month before and did that test and it felt fine, so I assumed the bearing was still good. Now that I was more sure about its condition, I still found it odd that *only* with the weight of that corner of the car on the bearing and using the momentum of the rest of the body as I shook things in and out could I feel any sort of thunking. So it certainly wasn't totally shredding itself inside yet.

Given that popular wisdom seemed to hold that one could ride fairly far on a bad wheel bearing, I managed to nurse that sucker for another 1000-plus miles after starting to really hear/feel it through the floorboards and steering wheel, doing periodic side-play checks to make sure it wasn't about to fall off, and got where I needed to be. Fortunately over fairly flat and straight interstates. The long wandering road-trip involving mountains I was hoping to turn this run into was kinda off the table by now, but at least I was still rolling instead of stranded in south bumfuk somewhere. I even managed to record what I was hearing in a short video (MP4, 1 Mb, ~7 sec) -- the steady tone that's about a G2 on the musical scale is the sound in question, and I could make it come and go a little by gently swaying the steering back and forth in the lane.

As the stories of Prius owners trying to replace their own front bearings are in the nightmare class and various special tools and possibly presses are needed, I decided to let the local "stealership" do this one. They got off on the wrong foot with me right up front by suggesting that my suspicion of a bad bearing might be wrong and they'd need me in for a "diagnosis" visit first, and then another visit to actually do the work after ordering and receiving the parts as wheel bearings generally aren't stocked at dealer parts departments. After expressing that I didn't really enjoy being talked to like an idiot and maybe they should find me somebody who isn't afraid of knowledgeable clients, we started the conversation over and they finally clarified the logistics of what they wanted me to do, and then scheduled the diag appointment a couple of days out.

I did a little reading through the service-manual procedure for replacing bearings and figured that with the right tools it should be a fairly straightforward fix, if they don't screw up the delicate dust cup inside the steering knuckle or mess up the suspension alignment. The car had *never* been aligned the whole time I've had it other than a minor rear wheel tweak using Bob's shims years ago, so it would be interesting if they followed the recommended procedure of a full alignment after having some the suspension apart.

Before heading for the dealership I popped the wheel and rotor off and had a look at the area, just to make sure there wasn't any collateral damage or signs of heating or any of that. Still no discernible hub play with the weight off the wheel! The CV boots were in good shape -- that was one thing I wanted to make sure of, and having a fresh look at the rest of the area would limit the scope of what the dealership could legitimately try and tell me was "bad". I could definitely see evidence of corrosion between the aluminum knuckle and the hub piece, which according to the stories tends to glue them together almost as one piece, so I expected Toyota would have a bit of similar trouble separating the two. The brake shroud in between might be a loss as well, being caught in the middle of the assembly and fairly rusty. But in general I've kept things fairly clean under there so maybe that would make their lives easier. Worst case, they'd get me a new knuckle *and* hub and then they would only have to worry about the balljoint and suspension attachment bolts.

After spending a summer dealing with a bunch of construction planning on the house, including research into what kinds of fasteners to use with different materials, I have to seriously wonder -- when aluminum parts used in auto manufacture are in contact with steel parts, why aren't said steel parts galvanized?? Having that would allow steel be in much closer proximity to aluminum in wet environments and not instantly turn into a self-destroying battery. I.e. the automakers need to look up "galvanic scale" and design their surface treatments accordingly.

I would have only this one bearing replaced, playing the odds that one failure is often an anomalous fluke and the rest of them were still good for the longer haul. Despite tempting thoughts of changing both fronts, collective experience seems to indicate it doesn't make sense to replace any more bearings than the failed one even at fairly high mileage. Not to mention additional cost. So off I went to the dealer, and pulled in next to the "fishbowl" where all the service-writers hang out. One of the techs passing by stopped with a pretty serious double-take and started eyeballing the car, so I gave him a website flyer. The service writer seemed curious so I gave him one too, with encouragement to distribute it around.

The paperwork mill involved the usual warnings about the various outstanding recalls like the steering shaft and hybrid water pump and floor mats, with me going "decline, decline, decline" all the way through. This car never even got the ECU reflash for the supposed "stalling" problem way back when, and I reassured them that I was well aware of all these conditions and in the interest of science [not to mention keeping their mitts off my car as much as possible] was deliberately not following up on any of them. They finally got me checked in, and whisked the car away into the holding-pen lot out back. Except that they'd also taken my backpack with it, and I had to have them walk me out there to fetch it.

About 20 minutes later the SW came to find me and said that yes, my diagnosis was right on the money, and better yet they actually *had* a new hub in stock and if I was willing to hang out at the place for another 2 or 3 hours they could have me fixed up and back on the road that very afternoon. Great, I said, let's do it! I didn't need to be anywhere specific that day and I'd brought my own entertainment along with me. They said they'd put "their best guy" on it, but I bet they tell that to all the customers.

Gaggle of techs around the car Soon they brought the car back, and into a bay conveniently near the fishbowl where I could actually see some of what was going on. Toyota places generally don't let customers in the shop areas where work is being done -- too many lawyers -- but this could afford at least a partial view of the progress. And the SW said he could escort me out there briefly for a picture or two later on without getting in trouble.

As the car went up on the lift a couple of other techs wandering by stopped to look at the stickers on the back and scan the [now somewhat out of date but still hangin' in] mods-list sheets in the windows, and soon there was a whole gaggle of them clustered around chatting and pointing. I'm guessing that they don't get too many visitors of quite this geeky a nature very often down here.

Suspension apart on other side Figuring things were probably in good hands, I found myself a little table in their waiting room and sat down at the laptop to work on more of my house story. After a short while I wandered back out to the fishbowl to have a look, and it was immediately clear that the whole steering knuckle was already off the far side and being worked on at the bench.

Applying the BFH to get the bearing out As I watched a little more, another tech brought over the Big Effing Hammer and handed it to mine, who then made like the Mighty Thor in the process of trying to bust the infamous dissimilar-metal corrosion between the aluminum knuckle and steel hub housing. I couldn't quite see past the little half-wall but it was pretty obvious that the thing was giving him a real run for the money. It's a very snug fit to begin with and as soon as any electrolytic bonding happens between the surfaces, it takes a whole lot of force to crack the thing out of there. Again, this is a well-known issue frequently mentioned on the forums and another reason I didn't really want to attack this myself.

Dremeling out corrosion in knuckle At this point I poked the service-writer and suggested that now would probably be a good time for a quick visit out there to grab a couple of pictures, so he finished up with a customer and we popped over to the bay. The tech was in the process of dremeling the remaining corrosion and roughness out of the mounting hole in the knuckle. I get the idea this is pretty routine for wheel-bearing jobs.

The new hub is sitting on the far corner of the bench here.

Naked wheelwell Brake caliper hanging on hose
I tried to grab a shot of the naked wheel-well, but I was trying to hurry as they didn't want to take a lot of time with me out there so it came out kinda crappy. I got the impression they were bending their own rules a little for me and didn't want to push it.

Later, though, I noticed from my viewing post in the fishbowl that the guy was letting the whole brake caliper hang on its hose, which is considered poor practice. In fact he let it dangle several times as he was trying to realign something else on the hub which was giving him a bit of a fight going back on. Really, how hard is it to tie the caliper up to the spring with a little twine, like the manual says? Well, the brake hose seems to have survived okay. He said later that rust inside the rotor "hat" was interfering with hub clearance and needed to be scraped out -- which I found odd because I've had it off and on any number of times with the interior rust undisturbed, and never ran into such issues. It's not like the new hub was any bigger.

First look at old bearing The tech had put the pieces of the old hub in the new one's box and given it to me during our visit, so I took it out for a quick look right there on the floor of the fishbowl. The two cone pieces had come apart in the process of the pounding, so I wouldn't even have to cut the thing open. It was already rather clear which side was the ailing one, with all the brown discoloration on the balls and cone and the other side's parts still shiny-silver and smooth in a healthy bed of white grease.

This raised a little confusion for someone who hadn't ever examined a wheel bearing and how it attaches before. Was it really that easy to separate the halves? Wasn't this supposed to be a sealed unit? I couldn't see anything other than a smooth interference press-fit that was supposed to keep this together, which had simply slipped apart collateral with the forcible extraction of the whole unit, and wasn't quite understanding how it could possibly withstand the forces from the road.

Hub bearing retainer hack
The little mystery was solved by finding a relevant section in a service manual, and confirmed with a brief chat with a support guy over at Timken. He was cordial and they've got a lot of good info on their website. [What I didn't think to inquire about was the relationship between Timken and Koyo other than both appearing in the leadoff picture here]. This is probably a generic diagram, as the specific bearing configuration in the Prius is clearly a little different.

Nonetheless, the "Special Service Tool" here is just a big-ass bolt and nut that substitutes for the missing axle shaft coming from the CV joint, and prevents the bearing from separating under load. When assembled into the car, the splined axle shaft has a wider shoulder up against the inside cone of the bearing [where the bolt head is here] and the big "jesus nut" on the end of the axle holds the whole thing together. This, said the Timken fellow, is why the axle nut torque is so important -- the bearing has to have the right preload to maintain zero play, but not get so squashed down that it starts binding.

So the whole key to retention is the axle nut, which gets staked in place after torquing to make sure it can't back off. This still amounts to 150 foot-pounds, which is like twice the torque applied to lug nugs. That's quite a lot, and why special big-ass wrenches are needed.

New bearing in place About 3 hours and $500 later I was out the door of the dealership. Most of the cost was labor and "shop charges"; the hub part itself only listed for about $110 even at the dealer's rack rate. On the bright side, it was nice to no longer hear that "hooning" as I went down the road! A day or so later I pulled the wheel and rotor off again to inspect the new unit and make sure the brake was back on to my satisfaction. In a quick wrapup conversation with the tech after he'd put the car back together he mentioned that my brake pads were getting a bit low; I reassured him that they were the originals and I knew exactly how worn they were and I would probably ceremonially change them after topping 200k if for no other reason than to avoid them delaminating from sheer age.

Oh, and they didn't bother with any alignment; just put the suspension back together exactly as it was before and assumed it was good. I couldn't tell any difference while driving. The backing plate behind the rotor was a little bent up and needed a tiny bit of straightening.

Old bearing semi-exploded view So now it was autopsy time, made far easier by the fact that the bearing was already apart. To check for wear on the various components I extracted one ball from each side, surprisingly easy to do as they ride in a flexible plastic cage and pop right out when gently pried with a screwdriver.

Wanna see my balls? And now you get to examine my balls. Well, er, you know. I could see no difference at all between my samples from the inside race vs. the outside one, so these are clearly the hardest metal in the assembly and hadn't experienced any wear.

Ooh, shiny.

The reflections are amusingly Escher-esque, with my camera lens and hands in prominence and the house, car, and driveway surrounding -- but on a much smaller scale. And the reflections of the reflections of the *other* ball would theoretically have an infinite progression.

Light fatigue spalling on cone Spalling in the race
The cones and races were a different story, however. After cleaning off the cone [and putting a nasty slice into my thumb when that sharp flange slipped, grrr] the spalling was obvious. In the race I wiped down into the cage where I'd popped the ball out and uncovered more roughness there too. Textbook metal fatigue. But what caused it??

I would encourage a critical look at the failure-analysis documents found at Timken's web site in the light duty automotive area. My guess would start with water ingress leading to corrosion, but it could also have been accompanied by some abrasive contamination from all the dust encountered in my various out-west roadtrips including pounding down lots of gravel roads in Death Valley. This may not have particularly good implications for any of my other bearings, but I can fervently hope their seals are all still unbreached.

_H*   121220