Jump to content

Photo

MSP enduro hub failure

- - - - -

  • Please log in to reply
80 replies to this topic

#1
Brandon

Brandon

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 765 posts
  • Location:North Jersey
  • Region:NNJR
  • Car Year:1996
  • Car Number:48SM

Everyone,

Here is what I'm submitting to the builder of MSP's hubs along with pictures of the removal (nut, spindle, caliper/bracket, ball joints) & the failed hub itself.  These were their 'enduro' hub with the long studs (2.25") installed.  It's my understanding MSP has halted the sale of these pending the builder's review of the two failures they've received (mine & another's).

 

  • Installed hub, torqued to 140#, installed nut cap w/silicone
  • Installed new Carbotech XP12 pads & rotors
  • Installed Toyo R-888 (<4/32 depth; unknown HCs) on +30mm wheels with a 5mm spacer

Loaded onto trailer (open) using tow hooks at front & subframe brace attachment point in rear (along side the differential); crossed tie-downs both F&R; not "cranked down" and the car will move counter to the trailer over bumps.  Drove to NJMP from north Jersey (2.5 hours, all Interstate/2-lane highway); easy tow mostly at 75MPH.

 

Friday was a test day (20 minutes each); here are the session summaries.

  1. Brake pad/rotor bedding; no offs or impacts, easy laps (mostly 1:39s)
  2. Low 1:38s, no offs or impacts, paddock tire pressures were within range (32.5-33.8)
  3. Same as #2
  4. Switched to Toyo RRs (~5 HCs), 1:37.5, paddock tire pressures were within range (37.1-38.7); no lockups or impacts

Saturday was a qualifying & race (still on RRs).

  1. Qualified P8 w/a 1:37.0; may have had 2x 4-wheel offs (to the right; left-hand turn); I left the paddock tire pressures too high (+1.8-2.2) so no post-session impound pressures were taken
  2. Race finished P6 with lots of low 1:37s; forgot to take pressures in impound

Sunday was a qualifying & race

  1. Qualified P3 w/a 1:36.5; did not take pressures in impound post-session but discovered a chunk of tread removed (directly to cords) at the tread seam.
  2. Race on SM7s (~5 HCs); finished P6; absent hub failure it might have been P3.  A lap 1 wreck (row 3 or 4) on front straight put us in the mix with a lot of faster cars so I was fighting with ITR/ITS cars throughout the first half of the race

Upon removal of the hub there was a distinct "suction" of the nut cap to the hub that was unusual.  IOW, there was negative pressure inside the hub sucking the cap back until I was able to pull it far enough away to suck in outside air.  This I had not experienced ever before (good seals in these hubs I guess?).  

 

The grease in these hubs is stinky!  Additionally the grease had been thrown all around the nut faces & the inside of the cap.  There was also grease thrown from the rear side of the hub onto the caliper bracket but none appeared to be on the inside edge of the rotor.

 

Looking down the bore of the hub I observed the gap between the two inner races, to me, looks to be a lot narrower relative to my other hubs but I did not take any measurements pre/post installation of either style (MSP or personal).  I'm suspecting the balls are smaller in diameter which enables this gap to be narrower.

 

That's about it for my observations and documentation of the car & sessions.  I'm at a loss to explain the failure considering my past experiences.  Granted I've been getting faster than the 4 previous seasons and so may be "using" more of the car's capabilities which is translating into parts wearing faster (I went through 50% pad thickness on those XP12s!!!).  However in my opinion that could not explain a $225 part failing before completing a single double-race weekend + test day.

 

Here's hoping they're able to identify something that will make these a worthwhile investment!

 

Brandon


Bona fide - A bonafide Spec Miata driver

#2
Brandon

Brandon

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 765 posts
  • Location:North Jersey
  • Region:NNJR
  • Car Year:1996
  • Car Number:48SM

Forgot to ask if there are any other data points I should include which might be helpful for them during their investigation?


Bona fide - A bonafide Spec Miata driver

#3
AW33COM

AW33COM

    Member

  • Members
  • PipPipPip
  • 278 posts
  • Location:NJ
  • Region:NE
  • Car Year:1990
  • Car Number:78
Installed hub, torqued to 140#, installed nut cap w/silicone

 

 

Please elaborate, what is 140#?



#4
Mike L

Mike L

    Member

  • Members
  • PipPipPip
  • 31 posts
  • Location:New Jersey
  • Region:65
  • Car Year:2000
  • Car Number:27
140 ft/lbs
Mike Lo

#5
AW33COM

AW33COM

    Member

  • Members
  • PipPipPip
  • 278 posts
  • Location:NJ
  • Region:NE
  • Car Year:1990
  • Car Number:78

Hmm,  in case that indeed is 140 lbs-ft, then that very well could be the answer to the failure.  

 

I have 0 idea what's inside that MSP hub, but 140lbs-ft seems way too low even for a small car.  Brandon, your bearing was under torqued, and the play (play you could not feel by grabbing the wheel) at the bearing caused excessive wear at an angle, damaging the hub (including machined outer race). 

 

The most common cause for bearing failure is over torquing (so NO, over torquing is not better than under torquing), where the bearing overheats.  I'm not sure if that's the case in SM, as drivers seem to under torque in order to get more "spin".  What they don't know is, wheel hubs at proper torque do not spin perpetually.   They actually have some resistance when you spin them by hand.  That resistance is about the same from car to car.  If you do enough of wheel hubs in your life you will learn to tighten them without knowing actual torque values.  The best exercise for this is to torque the hub from 140 to 200 every 10lbs-ft, and spinning the rotor without the caliper.  You will figure out right away what is right. 

 

Ironically, we may not need better hubs.  We may need a fixed torque wrench that "slips" off at a specific value, which will allow drivers to torque the wheel nut to the right value.  Back in the day cars with tapper bearings used to have pliers that would slip off at a specific torque value allowing everyone to torque the nut correctly. 

 

One other thing.  Besides money, wrench time, and PITA, there is a real safety concern with failed bearings if you ask me. 

 

 

**** I forgot the most important part.   Car HAS TO sit on the ground with wheels on,  when you torque the nut, in order for the clearances to set properly. 



#6
Mike L

Mike L

    Member

  • Members
  • PipPipPip
  • 31 posts
  • Location:New Jersey
  • Region:65
  • Car Year:2000
  • Car Number:27

I believe that 140 ft/lbs is within the range of the FSM.

As mentioned in another thread, the range I was given for the hub nut torque was (123 ft/lbs to 159 ft/lbs). This came from MazdaMotorsports so I trust they know what they're talking about.

I am going toward the high end myself in the future.


Mike Lo

#7
Bench Racer

Bench Racer

    Different strokes for different folks : )

  • Members
  • PipPipPipPipPip
  • 5,508 posts
  • Location:Wauwatosa, WI
  • Region:Milwaukee
  • Car Year:1990
  • Car Number:14

AW33COM, you seem to be full of information. Being I'm a somewhat of a doughting Thomas why not use your given name with your posts which to me would add some creditability. Been putzing with cars since the 1950's which don't mean $hit except maybe different experiences. :bigsquaregrin: 
Thank you
David Dewhurst


  • Ron Alan and DrDomm like this
Broken record - You are starting to sound like a broken record. Donor - Made PayPal donation Bona fide - A bonafide Spec Miata driver

#8
AW33COM

AW33COM

    Member

  • Members
  • PipPipPip
  • 278 posts
  • Location:NJ
  • Region:NE
  • Car Year:1990
  • Car Number:78

MazdaMotorsports could be wrong in this case, because the spec was written 25 years ago and the bearings in SM are not all Koyos back from the 90's.  Also, the question I have for Mazda is, why is there a range for the torque to begin with? And why is it so big?  40 ft-lbf diff is a live or die.   I checked on few Jap cars from the 90's with Koyos and there is no range, but a set value.  Was there more than one bearing spec?

 

Here is the spec for SKF (Timken): http://www.skf.com/b...6360/457377.pdf

That spec clearly shows 159.  So if you have a custom bearing, but you're torquing it according to the book or MazdaMotorsports, you could be off by almost 40.  That's 3 laps and the bearing is gone.  Explains a lot to me.  

 

Hubs require mega attention to detail.

 

Pic for doughting Thomases:

 

34y44li.png



#9
ChrisA

ChrisA

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 660 posts
  • Location:Richmond, VA
  • Region:NCR
  • Car Year:1999
  • Car Number:58
Read the foot note on the SKF pdf for the Miata #87. It says, "Maximum value shown. Torque value varies from 123-159 ft-lbs"

Chris

 

Happiness is a dry martini and a good woman ... or a bad woman.
- George Burns


#10
AW33COM

AW33COM

    Member

  • Members
  • PipPipPip
  • 278 posts
  • Location:NJ
  • Region:NE
  • Car Year:1990
  • Car Number:78

Miata #87 is for 99-2003 only :)



#11
ChrisA

ChrisA

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 660 posts
  • Location:Richmond, VA
  • Region:NCR
  • Car Year:1999
  • Car Number:58
^ And you told Mike Lo that torquing his 2000 to 140-lbs was too low and he should tale it to 200-lbs. That would be out of spec for Mazda and against the guidelines of SKF which says to follow auto manufacturer specs.

Chris

 

Happiness is a dry martini and a good woman ... or a bad woman.
- George Burns


#12
svvs

svvs

    do they sell spec training wheels yet?

  • SMembers
  • PipPipPipPipPip
  • 518 posts
  • Location:NYC
  • Region:NYR
  • Car Year:1991
  • Car Number:12?

:fisheszzz: this is the closest thing I could find to a "jumping the shark" emoticon.

 

The next major (so all the big boys will be out) I can think of is WGI.....a track that I think has traditionally been easy on tires (keep in mind i don't leave the NE so I have limited experience)....maybe it will be easy on hubs as well?  I guess we'll find out and have more data points?


Vick
www.volko.com
Black SM/SM2/"Slap Bracelet Throwback" #12 in the Northeast....if the car was made in the early 90's it should look like it.

1.6L forever!  Bring on your '99's and '01's!

Bona fide - A bonafide Spec Miata driver We have a Winnah! - Won their 1st race... Congratulations! Donor - Made PayPal donation

#13
AW33COM

AW33COM

    Member

  • Members
  • PipPipPip
  • 278 posts
  • Location:NJ
  • Region:NE
  • Car Year:1990
  • Car Number:78

And you told Mike Lo that torquing his 2000 to 140-lbs was too low and he should tale it to 200-lbs.

 

 ?



#14
Bench Racer

Bench Racer

    Different strokes for different folks : )

  • Members
  • PipPipPipPipPip
  • 5,508 posts
  • Location:Wauwatosa, WI
  • Region:Milwaukee
  • Car Year:1990
  • Car Number:14

AW33COM  Being I'm a somewhat of a doughting Thomas why not use your given name with your posts which to me would add some creditability.  :bigsquaregrin: 
Thank you
David Dewhurst

 

Pic for doughting Thomases:

Normally I consider your posts sound.    You selectively missed the point. :bigsquaregrin:


Broken record - You are starting to sound like a broken record. Donor - Made PayPal donation Bona fide - A bonafide Spec Miata driver

#15
Ron Alan

Ron Alan

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 3,732 posts
  • Location:Northern CA
  • Car Year:1995

Normally I consider your posts sound.    You selectively missed the point. :bigsquaregrin:

Now that DD is some funny shit!


Ron

RAmotorsports

 

Donor - Made PayPal donation Bona fide - A bonafide Spec Miata driver

#16
ChrisA

ChrisA

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 660 posts
  • Location:Richmond, VA
  • Region:NCR
  • Car Year:1999
  • Car Number:58

 ?

I hate reading & replying from the phone, too much scrolling about to do... You may have been replying to Brandon but Mike's post was right in there with it. Regardless, the torque was within spec and I believe they are the same for NA.2 and the NBs. I do agree that it's a fairly wide spread.  However, we must remember that the torque spec is also for the nut & spindle, not just the bearing. So while you may think 159-lbs is a low max spec for a bearing, it could be the at upper safe limits for Mazda's hub nut..?


Chris

 

Happiness is a dry martini and a good woman ... or a bad woman.
- George Burns


#17
Brandon

Brandon

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 765 posts
  • Location:North Jersey
  • Region:NNJR
  • Car Year:1996
  • Car Number:48SM

Quick update: spoke w/MSP about the hub.  ProParts disassembled and checked for any damage on mine and found none.  They re-packed and installed it onto a spindle and started measuring play at specific torque values starting at 120# and incrementing by 20.  They observed play at 120, minimal at 140, and nil at 160 (top-end of spec range).

 

MSP appeared to agree they would start to specify to torque their hub to 160#s going forward.

 

I hope to have mine returned sometime next week and I'll be reinstalling both hubs with new nuts torqued to 160#s.


Bona fide - A bonafide Spec Miata driver

#18
Bench Racer

Bench Racer

    Different strokes for different folks : )

  • Members
  • PipPipPipPipPip
  • 5,508 posts
  • Location:Wauwatosa, WI
  • Region:Milwaukee
  • Car Year:1990
  • Car Number:14

Direct from SKF/AW33COM catalog.  

 

All replacement hub units

are not created equal

 

SKF premium quality hubs         “Value” grade hubs

 

Is the Miata hub were dealing with a SKF or other manufacture premium quality hub? Does the SKF premium quality hub have a induction hardened outer race (few thousands deep) or a real outer race. (race that is pressed in to hub)  Do the SKF premium quality hubs these torque values are aimed at have outer races that are induction hardened? What torque values does SKF recomend for "value" grade hubs? Maybe the torque values work fine for SKF premium quality hubs and maybe the torque values do not work fine for the aray of other manufacture hubs used by Mazda and we who procure hubs.

The hub/bearing race gap I measured in the other thread is not marked SKF.

 

EDIT:

Thanks for the info Brandon, I was typing while you were posting.


Broken record - You are starting to sound like a broken record. Donor - Made PayPal donation Bona fide - A bonafide Spec Miata driver

#19
Johnny D

Johnny D

    Veteran Member

  • Moderators
  • 6,121 posts
  • Location:Fremont, CA
  • Region:San Francisco
  • Car Year:1999
  • Car Number:88

Quick update: spoke w/MSP about the hub.  ProParts disassembled and checked for any damage on mine and found none.  They re-packed and installed it onto a spindle and started measuring play at specific torque values starting at 120# and incrementing by 20.  They observed play at 120, minimal at 140, and nil at 160 (top-end of spec range).

 

MSP appeared to agree they would start to specify to torque their hub to 160#s going forward.

 

I hope to have mine returned sometime next week and I'll be reinstalling both hubs with new nuts torqued to 160#s.

 

This is good news.

One more question could be asked...

Does MSP feel their hubs are equal consistant quality??

 

I'd also suggest if you've had a failure.

Do what MSP did.

Check for damage, if not, TQ higher until Nil play.

 

If this is it, what will DD do with his time now?

J~


2011 NASA Western Endurance Racing Championship E3 Champ
We have a Winnah! - Won their 1st race... Congratulations! Beta-Tester - Assisted us with beta testing the website. Donor - Made PayPal donation Bona fide - A bonafide Spec Miata driver Novel Approach - When a paragraph simply won't do... Survive the 25, NASA Thunderhill - Survive the 25, NASA Thunderhill Instigator - Made a topic or post that inspired other Make it Rain - Made Paypal donation of $100+

#20
Bench Racer

Bench Racer

    Different strokes for different folks : )

  • Members
  • PipPipPipPipPip
  • 5,508 posts
  • Location:Wauwatosa, WI
  • Region:Milwaukee
  • Car Year:1990
  • Car Number:14

 

If this is it, what will DD do with his time now?

J~

It took a while for Keith's several thousands of clearance info to sink in. DD just did his last measuring of the OEM NTN hub, by inserting some material between the inner races, tourqued to 150 ft lb, took apart, measured and the clearance between the inner races of the one hub I measured is .005/.006 inches.


Broken record - You are starting to sound like a broken record. Donor - Made PayPal donation Bona fide - A bonafide Spec Miata driver




1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users