4 Race old Timken decided it didn't want to play anymore. (~4hrs track time)



The back story is everything felt kosher on day 1 (practice, quali, race 1 then 1 hour enduro) Second day of the weekend now and towards the end of our morning warm up session I noticed what felt like a wheel balance issue or wheel bearing starting to go bad. "No way" I thought, I had just put new fronts on the second day of last weekend. Came in checked tire pressures, wheel torque, then jacked the car up and gave the wheel a wiggle, nothing. Hmm well maybe it's not the bearing then. Looked at the tire and saw a fair amount of rubber marbles and a few guys had said that sometimes they give you a good wheel vib so wrote it off to that.

Went out for the race and right out of the gate I'm thinking it felt worse! I start to give the tires a good side to side scrub and make it 3 corners before my driver front tire pops off and I slowly slide into the gravel, thank christ it happened on the out lap.

So post race analysis is why the **** did this thing fail?

Cheap china metal? Hub torque too lo-hi? Wheel lug torque too lo-hi? Machined rear sections? Too much grip/wheel offset? ARP studs forced in wrong?

Looking at my old stock hub (new Mazdaspeed replacement looks same btw) the back side is straight cast vs the Timken machined, this section of the Timken is also 1mm thinner at the inside area of the studs vs the oem units, looks like a great stress riser to me.




So game plan is I currently have the cheapest pieces of **** on rockauto on the way, these were to be emergency spares incase a wheel bearing failed, curious to see how they look when they arrive. Going with the Raybestos hubs next as they are cast with the thicker shoulder sections, hopefully never have to deal with this kind of failure again.

 

I've never seen a Timken machined like that. That's turrible.

 

I did that to an OEM unit. A lot more hours though. Brought it in as soon as it started vibrating (way more than the typical tire booger stuff) and it also gave hellacious pad knockback (had to pump to get a pedal). I was cracked 75% around in the same pattern as yours.

There's some discussion about the stud holes being undersized on certain replacement hubs. When you press in the ARPs, you overstress and get some microcracks started. Maybe? Seems to me that bbundy posted about it.

 

I went all fancy and bought NTNs. The hub stud holes were undersized, even though those are supposed to be among the more legitimate.

I talked with an engineer at ARP on the phone, and he strongly recommended I ream the holes to .005 under the major diameter of the stud knurl. I bought the appropriate reamer and did it since I was afraid to press the studs into the small holes.

That being said, the backside machining on this hub is really unusual -- seems like a more likely cause than the undersized stud holes, since a lot of people have faced that stud issue.

 

Found it:

When was the last time you changed your rear hubs? - Page 10 - Miata Turbo Forum -Boost cars, acquire cats.

Pertains to rear hubs though. I've cracked those too.

Agree, I've never seen machining like that.

 

The stud hole sizes in aftermarket hubs seem to be a crap shoot. I replaced mine a couple of years ago, and installed ARP studs during the process. I had to bore out the rears, and the fronts were oversized, so I ended up returning them and going with OEM. Wasn't the usual 1.8/1.6 size differences, just wonky sizes from the mfgrs. I guess they figure if the stud is the right size/thread, the spline size/hole doesn't matter.

 

I feel if it was a ARP stud/hole size issue the cracks would propagate through them vs how it broke. And yea hornetball there was no knock back on the previous session when i noticed the slight vibs but I didn't recognize it fast enough on the outlap because it definitely had it then, learning experience I hope everyone here can take without having to experience.

I've sent Timken an email regarding the failure, we'll see what they send back.

-Dean



An arm chair analysis by a friend who has some experience with material failure stuff

tldr; machining bad

 

The material around the studs is still intact. Bad machining for sure.

 

Do you have a photo of the front side of the flange where the studs pass through? And also the back side of the rotors where the studs pass through?
Its possible with the back side of the hub flange machined the total flange thickness was considerably thinner. If this is the case, the larger knurled diameter of the studs would extend further from the face of the flange and possibly the rotors were only supported by the 4 points of the studs (where their diameter was larger) and were NOT clamped to the front face of the hub. As you did your first few sessions, the larger diameter of the stud base may have started to press into the back side of the rotor stud holes causing the wheel, to effectively become loose - vibration, increased stress loading on the hub, etc. When you torqued the wheels following the vibration were they already tight or did you have to add another 1/4 or 1/2 turn?

Another theory... It is preferable for a machined surface such as the back side of the hub to have a large radius from the back side of the flange to the cylindrical surface behind the bearing. A tight radius can cause stress concentrations which will lead to earlier fatigue failure. Using this theory however the total load cycles is way too low for something like a wheel bearing. (based on 2 hours of running time, less than 200 miles total distance). If this was the case, then similar failures would be occurring on regular road cars at say 20,000 miles. Sounds like a class action law suit in the making...

 

This is terrifying.

I hope these get finalized soonish
https://www.miataturbo.net/race-prep...2/#post1344681

 

Just a side story. I got the timkins from autozone with the 2 year warranty. every 3-4 races the hub bearings would fail and i would need to replace them. Last time i went autozone was out of stock. so they handed me the house brand. I've gone 8 races so far and they are still good!

 

Did you have any car-to-car contact after this hub was installed? Sorry if that seems like a dumb/obvious question, but this is SM we're talking about...

I tried the $30 heat-treated Centric hubs for a while and I won't use them again, even as spares. I didn't have any failures, but they developed significant play REALLY fast. I went through 4 or 5 in 1 year. Switched to Timkens and haven't replaced a wheel bearing in about 2 years so far. Based on internet pictures alone, the Raybestos and Moog non-ABS bearings look like the current best options. I think the Moog is just a rebranded Timken 513152. I'll be able to confirm that tomorrow after UPS shows up. I'm curious to see if it has this same machining on the back of the stud hole.

 

No wheel contact other than curbing, our LF takes a decent beating at the local track.

A spare for me if it lasts a day is good enough

 

Hub flange sits square and flush on the rotor face, just checked so not a concern but great thought. The lugs were torque checked and as far as I can remember none were loose.

Edit: WJB $20 wheel bearings just arrived, non machined just fyi

 

For those who don't stray outside the Race Prep section- update on the Moog here:

https://www.miataturbo.net/suspensio...2/#post1358131

 

@OGRacing did you repack them or run them as is?

 

did it both ways. first few i would re pack them with high speed synthetic grease. the last few where track repairs and i just stabbed them on.

 

So you've done 8 races on AutoZone Duralast hubs without repacking them?

woot. I can probably buy those for nothing. research time.

 

Pictures of my broken hub from TWS. Cracked 75% around. I'd rather be lucky . . . .

I cracked this hub early in my ownership of the Red car. According to the PO, the hubs are original to the car (1995 model year). They exhibit the same kind of machining as OP's hubs. Interesting. What's also interesting is that the bearings still turn smooth and tight. I've heard stories about the original bearings from years past being much better than what we're getting now. Maybe there's something to that.







Also, for Aidan, the grease seals on this hub had visible markings:

 

Update: I emailed Timken including pictures of the failure and after a couple emails I received this reply.

Wow….a failure like this is extremely rare. I would expect to see this from a heavy impact of some sort.

The challenge I have in this situation is that Rock Auto is not an authorized Timken Distributor.

I cannot with certainty tell you that what you purchased is authentic Timken product as we do not know who Rock Auto is purchasing from (nor will they tell us).

I am awaiting instructions from our product team as whether they want these hubs back for failure analysis and determine if they are in fact Timken.

That being said, I will send you a couple replacement hubs.

Can you send me your mailing address and I will get these to you asap.

Thanks

Patrick

--------------------------------------------------------------------------------------
The photos have been reviewed by our Technical Trainer and based on what he sees, his view is that the hub failed as a result of the vehicle being used on a racing circuit. This is a common failure in this environment.

Before I can proceed, I will need additional photos of the vehicle and a detailed explanation of how the vehicle was being operated when the failure occurred.
----------------------------------------------------------------------------------------
I replied honestly about the use
----------------------------------------------------------------------------------------
Nice car!

Our official policy is no warranty as the product is designed for normal use, not racing of any kind. It’s not relevant if the car is street legal.

I’m told this a very common failure point for Miata’s when raced.
I’ll still send you a couple hubs, only because I like to support racing where I can.
Thanks
Patrick

So I'm content with the result, hopefully they don't send the machined bearings though