Keiser 10" wheel failure

[mmcdermott]

A Keiser 10" aluminum wheel failed on Thursday, June 12 during testing of the 2013 Texas A&M FSAE car. The wheel had a center designs which Keiser supplied to FSAE teams using 10” wheels. The left rear wheel failed during a right hand corner. Fortunately, there were no injuries and minimal damage to the car.

Texas A&M purchased eight new 10” wheels from Keiser in 2011 and used them on the 2011 FH car & the 2012 & 2013 FSAE cars. Texas A&M has not done a lot of off season testing and driver training since 2011. The failure was due to fatigue. After the failure the seven remaining wheels were tested with dye penetrant and all spokes on all wheels had cracks. There is no visible damage to any of the wheels that would cause stress concentrations to initiate cracks. In retrospect, it is obvious that the wheel center does not have a very large section modulus in the direction that supports bending during cornering. Keiser is aware of this problem, has a new design with a higher section modulus, but has not notified users or the old design.

All three cars weigh about the same - 350 to 375 pounds plus driver. The 2011 & 2012 cars cornered at about 1.5g. The 2013 car has an aero package to generate downforce and grip & corners at about 2.0g in fast corners.

I have photos of the failed pieces & a cracked (but still intact) wheel center, but have not figured out how to post them.

[Marshall Grice]

how many times are you guys going to post this?

[MCoach]

If the wheel centers are failing, then it's probably not far off to guess that they are also flexing further than what should considered acceptable on car.

[whiltebeitel]

http://i232.photobucket.com/albums/ee291/whiltebeitel/2013-06-13120825.jpg

http://i232.photobucket.com/albums/ee291/whiltebeitel/centeroffailedaluminumwheelstillattachedtohub13061 2.jpg

http://i232.photobucket.com/albums/ee291/whiltebeitel/outsidepartofcenteroffailedKeiseraluminumwheelstil lboltedtorimhalves130612.jpg

[jlangholzj]

The thing is...is this NOT a visible stress? I didn't need ink to see how bad those splines were failing....at the risk of sounding like an asshole.....shouldn't someone have caught this?

[MCoach]

Originally posted by MCoach:
If the wheel centers are failing, then it's probably not far off to guess that they are also flexing further than what should considered acceptable on car.

What I meant by this, is shouldn't someone have noticed the camber compliance?

[Canuck Racing]

Originally posted by MCoach:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by MCoach:
If the wheel centers are failing, then it's probably not far off to guess that they are also flexing further than what should considered acceptable on car.

What I meant by this, is shouldn't someone have noticed the camber compliance? </div></BLOCKQUOTE>

The failure occurred over three years. Would you notice slow, steady camber compliance increase over three years? On wheels already known to have high amounts of deflection?

[MCoach]

Originally posted by Canuck Racing:
The failure occurred over three years. Would you notice slow, steady camber compliance increase over three years? On wheels already known to have high amounts of deflection?

It was enough deflection on our wheels to prompt us to crack check them.

[Charles Kaneb]

Originally posted by jlangholzj:
The thing is...is this NOT a visible stress? I didn't need ink to see how bad those splines were failing....at the risk of sounding like an asshole.....shouldn't someone have caught this?

The first visible indication that the wheels were cracking was a partial wheel and tire rolling along the pavement without a car attached. The cracks on the other wheels weren't visible without the dye.

[jlangholzj]

Originally posted by Charles Kaneb:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by jlangholzj:
The thing is...is this NOT a visible stress? I didn't need ink to see how bad those splines were failing....at the risk of sounding like an asshole.....shouldn't someone have caught this?

The first visible indication that the wheels were cracking was a partial wheel and tire rolling along the pavement without a car attached. The cracks on the other wheels weren't visible without the dye. </div></BLOCKQUOTE>

that would be quite the visual indication of failure http://fsae.com/groupee_common/emoticons/icon_smile.gif

what I was talking about are the spokes holding the inside hub to the outside shell...to me it looks like there's is some visible deformation in the spokes. Almost like a less-exagerated version of FEA displacement....looks like the center of the wheel center was held and the outside was twisted.

[jlangholzj]

Originally posted by coleasterling:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by MCoach:
It was enough deflection on our wheels to prompt us to crack check them.
How did you test and quantify that?


jlangholzj,

That's just weird lighting. There is/was no deformation visible in the spokes. I would bet that there isn't any measurable deformation either, but I haven't checked them yet.

-Cole </div></BLOCKQUOTE>

ah! Well then I stand corrected http://fsae.com/groupee_common/emoticons/icon_smile.gif

out of curiosity has anyone verified that the thickness of the spokes and the predicted loading cases for the tires are compatible? Ie, is it just a weak casting of aluminum and the centers came from a bad batch or are the spokes too thin for the loading cases?

I've seen teams run more with less...so I'm doubting its the latter...just would be curious to see the results!

[Buckingham]

Ie, is it just a weak casting of aluminum and the centers came from a bad batch or are the spokes too thin for the loading cases?

It's not a bad batch because they failed exactly where they should. Smallest cross sectional area closest to the center.

The spokes are not too thin for the loading case (once), they are too thin for the loading case "N" number of times.

With "S" load (stress), any wheels will fail after "N" number of cycles.

If you increase the cross section, you will reduce "S" and increase "N". But "N" will never equal infinity. Aluminum has no Endurance Limit.

Racing wheels are consumable items that should be checked for dimensional tolerances and inspected for fatigue damage AT REGULAR INTERVALS.

[MCoach]

Originally posted by coleasterling:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by MCoach:
It was enough deflection on our wheels to prompt us to crack check them.
How did you test and quantify that?


jlangholzj,

That's just weird lighting. There is/was no deformation visible in the spokes. I would bet that there isn't any measurable deformation either, but I haven't checked them yet.

-Cole </div></BLOCKQUOTE>


After watching the car move from the set negative camber value to small amounts of positive camber year after year during corner loading, we decided that anything that brings that much visual deflection should probably be checked. This wasn't a case where we strain gauged the wheels, just paranoia and seeing large amounts of deflection.