80 replies to this topic

[Bench Racer]

X-ray vision
J~

Dave, I'd never accuse you of be a smart ass observer sitting on the side lines with your serious lifetime knowledge in automotive/racing vehicles doing nothing but running his mouth like some folks on this site. Saul, Keith, Brandon and others that contributed by doing something/investigating thank you. Now that I said that I feel much better.    With all that in the rearview mirror I finished a another round of, is there inner race inside end gap or no gap. My first measurements which were made indicated there was a .005/.006 gap. Johnny, I did something and proved myself wrong. Been there before. I'm convinced with my latest investigating there is zero clearance between the inside ends of the inner races. The failed hub from Jared Gerber #9 SM where the balls, the inner/outer races and the cage visually looked (other than spalling/craters on the inside ends of the inner races, doesn't look like rotational) ok was IMHJ an abnormality. The whole point being guys, I enjoy doing something to eliminate my curiosity.
 
Ron, a slouch drives my car every time it's at the track, hence the front hubs last a loooooog time.

[SaulSpeedwell]

Here is something objective.  Note that Mazda specifies an endplay spec. 

 

 

1993-95 929 and 1993-97 Miata

The wheel bearings on these vehicles are not adjustable. To check if the bearing requires service, remove the wheel and tire assembly, brake caliper and disc brake rotor. Install a dial indicator with the indicator foot resting on the wheel hub. Try to move the hub in and out. If there is more than 0.002 in. (0.05mm) bearing play, check the wheel hub nut torque or replace the hub and bearing assembly.

 

http://www.mazdabg.c...el_Bearings.htm

[davew]

Now I have to carry a dial indicator to the track!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

 

Dave

[mellen]

Now I have to carry a dial indicator to the track!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

 

Dave

doesn't everybody??

[Parity]

Now I have to carry a dial indicator to the track!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

 

Dave

Yeah, but is it calibrated? Make sure you bring the cert:)

[davew]

Now you're telling me that I need to be calibrated and certified.

 

All I want to do is drive a chick car around in circles

[Parity]

Here is something objective.  Note that Mazda specifies an endplay spec. 

 

 

1993-95 929 and 1993-97 Miata

The wheel bearings on these vehicles are not adjustable. To check if the bearing requires service, remove the wheel and tire assembly, brake caliper and disc brake rotor. Install a dial indicator with the indicator foot resting on the wheel hub. Try to move the hub in and out. If there is more than 0.002 in. (0.05mm) bearing play, check the wheel hub nut torque or replace the hub and bearing assembly.

 

http://www.mazdabg.c...el_Bearings.htm

I wish Mazda, or any of the bearing manufacturers, gave us a little more detail on play tolerance. It can be tough to differentiate end play from overall slop. If the measurement is actually checking end play then you should get the same reading if inspected at the tire bead since the bearing is moving parallel to the axis.  If you put (2) indicators 180 degrees appart and they move in the same direction then you're measuring end play... but if one moves in the opposite direction then you're measuring other clearances and .002" of slop would be about .024" at the tread surface assuming  a 24" diameter tire. ANyone feeling that kind of movement would figure the bearing is shot.

 

Edit - My manual states "wheel bearing play" not end play but it does say to "push and pull in the axial direction" which is end play.

[Bench Racer]

Now I have to carry a dial indicator to the track!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

 

Dave

 

Now you're telling me that I need to be calibrated and certified.

 

All I want to do is drive a chick car around in circles 

Dave, not an issue, when I load in the morning I'll bring a pair of magnet dial indicators.    (actually from my home made bump steer gauge)  They'll be as good as the operator. And I suspect the operator in your paddock can handle the job.

[Brandon]

I'm still awaiting the return of my MSP hub from ProParts.  What I was originally told ("no problem found") has been amended to "possible plastic race issue" and the issue was being correct along with the hub being cleaned/re-packed and sent out 'soon'.

 

Once I receive it I'll give it a once-over and report back anything different (unlikely) from the one I still have installed on the car (RF).

 

What are people's thoughts about painting the inner race faces (the sides that form the gap in the 'middle' of the hub) with some sort of color-change paint?

Something that would indicate a maximum temperature reached inside the hub - would that be helpful in our diagnosis at all?

[Mike L]

Brandon, I had gotten my Mazda HD hub back last week and they had confirmed the hub was fine.
YMMV, but when I put it on and tightened to 159 ft/lbs, there was still noticeable play. Spoke to Mazda and they said they would take note and bring it up with their manager.

Needless to say that hub is sitting in my bin as a "backup" hub.
Won't be buying the Mazda HD hubs anymore. Staying with Planet Miata hubs or buying NTNs. Can't trust the Timkens anymore because the last few I received didn't have any of the NTN stampings and it looked noticeably thinner where the studs were.

[FTodaro]

Interesting comments''

 

I have also noticed some of of the different hub offering have different hub castings,I think i saw and compared one of Novak's Sk hubs to my NTN and it looked like the area around the races was beefier but the flange where he studs attach was thinner. I decided to pass on that in light of failures being reported there.

 

I would encourage anyone to run any test they dream up to compare or diagnose issues, since we don;t know what we are looking for, no need to limit what we look at.

 

I will repeat, unlike many, i have continued to use hubs with a light noticeable play in the hub if the grease appears to look good and not contaminated.

 

I think we also need to look to see if the plastic retainer rings are leading to a failure or sorts.

 

i will examine the different race retainers i have and report back if i can find or measurable any variances.

 

I have accumulated some original OEM hubs, they had a different grease seal on them. At some point it would be nice to compare them to the new offerings.

 

I know a local forensic metallurgist in town. I may approach him and see if he can do comprehensive comparison  of the two.

 

If I can't get him to do the for free " Jim " any budget with SCCA for investigate safety failures?

 

I would suggest that we accumulate the different hubs on the market and include some old original OEM hubs. approach this expert, get his opinion on what we should examine and compare, present the list to the group, figure out the cost of the project and then figure out how we will proceed.

 

would be nice to get some SCCA/NASA help on this issue.

 

I also think we need to try and find  out more about the history of who is manufacturing the hub offerings, this may be difficult to obtain, voluntarily. May have to get a litigiousness lawyer to file a discover action where we have the right to subpoena documents, under the theory of product defect. (this would not be plan A) The legal action would  be to get information only, the manufacturers would be immune from any liability because this failures are beyond the intended purpose (racing) so we not not perusing money just information for the public good of safety.

[Brandon]

Finally got back 'my' hub - I use quotes as it's my position it was a new one they finally shipped out instead of returning my original.

What this means is I have no way of confirming/denying any wear or other potential indicators of premature failure.

 

And this means my idea of swapping hubs from side to side is not gonna happen either since my next event is at WGI and there's enough long-duration right hand turns so I'd rather have the confidence of a new hub on the LF.

 

Well, "new MSP hub" at least.  

 

I'm still re-packing my old set of hubs along with a '94-era "known good" so I'll be well covered in the event of a MSP failure regardless...

[RazerX]

Peanut Gallery to be sure.  Just catching up on the thread.  With my engineering hat on, i think there needs to be a gap between the inner races.  This allow there to be a compression force on the bearings to provide centering of bearing to handle the axial and radial loads.  All moving things wear so as the bearings and races wear the compression force from the nut allows the races to move towards each other can keep the bearings under load and centereed.  As soon as the races touch the compression force on the bearings will fall off and could theoretically fall to zero.  However it is likely it would fail long before that. 

 

I have had some hubs go 2 years, some go two races before a minor wobble detected from a 12 and 6 test.  Never has tightening it more, resolved that, and in the once case were i was too lazy to replace it, i put one more day on it  and it was shot.  Huge wobble.  So i admit i don't have any old 'failing' hubs to contribute evidence to this thread but i will save and investigate going forward.  BTW, i have had this issue with OEM grease and repacking.  I just accepted it "S$it happens".  But i have not had a catastrophic hub failure either as i check hubs at least twice a day at the track. 

[Brandon]

One more follow-up:

I did, in fact, receive my "failed" (but designated 'good' by ProParts w/their torque value of 160 ft#) component but I had already installed the replacement one from MSP.

(I'm now the proud owner of 3 of these hubs!  Yay!)

 

As I had already started the process of the swapping, I returned the original MSP RF hub to its location, torqued to 160 and sealed it up.

Installed the replacement MSP LF hub and torqued to 160 and sealed it up.

 

Went to the Glen and ran the test day and a single race group.

 

Only issue identified was on the LF which I (ultimately) misidentified as hub play.

Another R&R of the LF hub shifted my suspicions to the lower ball joint.

Then upon arrival of a more experienced mechanic he identified what was presumed to be inner tie rod play.

 

Anyway....Everything reinstalled back on the spindle!!

No further issues that weekend apart from some bone-headed driving by yours truly....

 

Where does that leave us?

I personally have now ran two different hubs at the LF corner on tracks with serious, long length & high speed right-hand turns; one exhibited noise near the end of the weekend (while 'under torqued' per MSP/ProParts) and the other, while used less than the first, has presented zero issues while torqued at the MSP/PP specified value.  Also I still have a previously noisy hub to disassemble and report back on anything discovered.

 

As I noted before, I may be merely reaching a point of performance/speed that could accelerate wear issues going forward.  But whether that means I should be changing my maintenance procedures (from annually swapping hubs to monthly for example) or deciding on a level of disposability for them is still uncertain.  "Right now" I've never experienced a "wheel separation event" (regardless of where or on which axle it happens) and I hope to mitigate that likelihood by being more meticulous about checking my hubs between sessions.

 

I leave the debate on what the Mazda engineers were trying to accomplish with the hub design to the machinists and metallurgists on the board...but if someone comes up with a viable replacement that we can get SCCA/NASA to sign-off on all the better for us!

[Bench Racer]

Thanks Brandon. Who is ProParts and have you ever asked MSP or Proparts about the design parameters of the inner ends of the inner races touching from the get-go or if there is to be inner race inner end clearance?

 

My  , if in fact by design parameters at the get-go there is to be no inner race inside end clearance, increasing the torque from 140 to 160 lbs foot would do nothing ( the extra torque sure as hell would not squeeze the inner races shorter). If there is to be inner race inside end clearance, increasing the torque from 140 to 160 lbs foot would reshape the cheap outer race (induction hardened a few thousands thick) by a couple thousands and eliminate the 12 to 6 wobble.

 

Paul, please post your MSP response. When you receive the response. NTN never responded to my request for information.

[Parity]

Proparts is the outfit that actually replaces the balls and grease in the heavy duty hubs being sold by MSP. They also receive any returned hubs for investigation. Information has been slow coming but is in the works. MSP in going through Mazda to communicate with NTN concerning any recent changes in process or suppliers which may be affecting quality. I've also asked them what criteria they use to determine if the hub failed or not.

 

It would be very helpful if failed MSP hubs were returned to MSP for evaluation otherwise I'm afraid many of these complaints will be perceived as potential internet banter. Keep in mind MSP expects torque to be 160 ft/lbs or they discount the issue as being too loose.

 

I have a test spindle setup where we check end play and radial play with indicators. Then we remove the balls and grease and visually inspect the races. Anyone with a failed hub who wants an inspection can PM me for info.

 

Beyond that we really need some direction from MSP/NTN.

[AW33COM]

Mazda Miata uses double row angular contact ball bearings at the front.  The inner race is divided into 2 separate inner races with a gap in-between .  Those bearings are best for axial and radial loads, and that’s why they are used in a car.  The double row of balls allows taking axial loads in both directions, with one minor problem:  sensitive to misalignment during installation. 

Automobile wheel bearings are under heavy loads.  In order to make those bearings super strong we preload them (at or before* installation).  Miata front bearings are preloaded.  Preloaded in this case simply means compressed, so that they do not “move” under loads.  That is what makes them super strong.  That compression is technically known as negative clearance or initial clearance/load.  You set the preload when you torque the wheel nut. 

 

The gap in between the inner races is there to allow you to set that very preload.  It is not there, as some suggested, stopping the tension on the ball bearings that could cause binding.  You can actually cause binding (clearance too small or none) by over torquing (cause #1 in street car bearing failures).   Hence, inner races DO NOT really touch, unless they are machined/ordered wrong by the producer, where the preload is not fully set while the inners touch (it would be crazy if that happens on some).  If inner races were made to touch, we would not need two of them.  One would serve the same purpose

 

Now back to real world.  There is no such thing as a heavy duty bearing for the Miata.  Buyers are only throwing money away.  For us racers, there are 3 obvious ways to improve this bearing:

1.  Use better grease by calling one of the bearing companies and asking them.  There are almost 1000 (yes) different greases out there for wheel bearings.  ONLY wheel bearing company will know what is right for us.  I’m positive, the grease will make the bearing better, but I don’t think it will fix our problem.
2. * Laser welded bearings.  There are companies who will take a bearing, preload it perfectly and weld the two inner races for life.   The compression would be fixed and it would remove all installation errors.  Not sure if our hub can be welded like that though. 
3. Redesign the contact angle on the bearing.  I’m not sure if this would help.  Maybe our contact angle is optimal already. 

 

Just wanted to clarify the inner race gap thing as it’s the most important concept on that hub.

[Parity]

AW33,

 

Thanks for the revelation but I think what is being looked into is not the bearing deign (generation 2 style hub) but the possible metallurgical or manufacturing defects which may be contributing to early failure.Again, it's important we qualify the failures with DATA. Several have mentioned the possibility of a redesigned hub. Most likely it would be a generation 1 style similar to the rear. Assuming you want to retain at least the same journal diameters the hub housing would have to be much bigger since it has to house the outer race of the new bearing. We've built racing hubs for another application and can design one in a day but unless NASA and SCCA are open to the idea it would be futile.

 

The top race teams have the most pull with Mazda and needless to say I'm not one of them but here's my .02 on the direction to go.

 

1 - Investigate the current hub supply to determine best product currently available and push for Mazda/NTN quality evaluation (a bit of a stretch here)

2 - Tires. There's a growing consensus the current Hoosier may be contributing..

[jdmrrs]

Tapered roller bearings are better suited for the axial loads generated when racing on sticky tires. 

To improve safety, by reducing the frequency of sub-frame weld cracks, front bearing failures, and rear flange breakage, Spec Miata needs less tire. We can't keep going faster and expect reliability to be unchanged.

[SaulSpeedwell]

PDF page 10 and 11 (marked pages 08 and 09) are worth looking at.  Potential defects aside, the portion of this issue that "we" control and have "caused" are Rs, T, and S,

 

http://www.ntnameric...arings_4601.pdf