Hubcentric rings. Street/Track what to use.

[Tangoman]

Can you say this in English?

Lets say you have 200 ft-lbs of torque and you modify your engine and now you have 300 ft-lbs of torque. Under hard acceleration, the 5-stud pattern will need to resist that torque. It also resists vertical and longitudinal loads coming in at the tire patch, which can be 1-2Gs on the road and 4-5Gs worst case at the track.

But, if your Honda spec bolt torque is only 80 ft-lbs, and you expect to track your car and have far more drive/brake torque than a normal NSX as well as higher loads all around, I believe that joint is now underdesigned.

When Honda generates a torque spec (or any engineer for that matter), they take the stud, the lug nuts, all the pieces in between and torque them up to proof load (the point at which a bolt will yield). They'll run 10 ft-lbs, 20 ft-lbs, up to 150 ft-lbs perhaps (perhaps in a Skidmore). Each time they measure the actual bolt tension in the bolt and compare it to the torque needed to get that tension. In reality, torque means nothing - that torque is easy to measure though, and can be correlated back to bolt tension. The engineers might want 50,000 lbs of preload in those bolts (10k for each bolt) so that with an assumed coefficient of friction of 0.3, they can resist 15k of shear loads in the joint. Assuming 1.2k wheel load that would be a safety factor of 12.5 (these are all just numbers chosen for calculations sake).

In the end if they've designed the joint to be at 80 ft-lbs based on their torque-tension study, then it can resist a certain amount of vertical/fore-aft/lateral and brake/drive torques. Lets say they designed that joint to 50% proof. In theory you could take that 80 ft-lb number, ratio it up by 1.5 to get to 75% of proof and you'll get 120 ft-lbs.
The stud will yield at 100% proof, or 160 ft-lbs or so (once again, these numbers are just for calculation sake).

Anyway, it surprises me that everyone is willing to modify their cars but in the end they aren't willing to modify the torque specs required to hold the joints together now that those joints need to take higher loads. Worst case is you do your own testing, you find your yield point of the studs and as long as you stay below that you're fine and your wheel will be able to take more load before it slips and starts loading/fatiguing the studs instead. The joint is designed to allow a stud to strip out/pull apart before the rest of the pieces in between (the hub, etc) yield or deform permanently (called ultimate-torque testing, it is part of everyone's validation sign-off requirements, at least at my work it is - I'm sure Honda would have thought of it).

For reference, grade 8 bolt proof loads
Maybe that still wasn't English :smile: but I think it is important to discuss how a torque spec is developed so everyone understands that it isn't some magic torque spec from the NSX gods that is infallible, it can be and should be modified if the joint needs to take higher loads than what Honda thought it would.

 

[AU_NSX]

... Anyway, it surprises me that everyone is willing to modify their cars but in the end they aren't willing to modify the torque specs required to hold the joints together now that those joints need to take higher loads. ...

I understand exactly what you are talking about and was discussing it with my tyre/wheel supplier last week. I want to get some longer and higher tensile wheel studs to go with some lighter titanium lugs on the track car. However, by your reckoning, this could render another part in the hub as the new weak point?

Additionally, going from grade 4 to say grade 8 stud will give a higher tensile strength but now the stud will also be more susceptible to a brittle failure (admittedly at a higher yield) ! Yes?

I have been tracking my '91 car for 7 years now without even having to do the bearings! Some dedicated track cars are running wide body 265/18 & 315/18 R-comp tyres with the OEM bearings (I am unsure about the studs) so I am under the impression that even upgrading the studs and lugs from OEM, that this will not adversely affect any other parts in the hub mechanism?

In any case, I do not think that any of this goes to explaining why Jim stopped having studs fail when he removed the centring rings from his RPF1's. All engineering design and laws of physics I can think of justify less shear stress through the studs with centring rings!

The only thing I can hypothesise is that without the centring rings, the wheel is able to "move" to distribute a stress (like hitting a curb) across all 5 studs???

However, when you trace the load of the wheel hitting a curb, that should be transferred through the wheel and through the (properly fitting) centring ring to the hub and then the suspension and the studs continue to only have tensile stress not a shear stress...

If the studs are failing with centring rings, are they OEM studs? if so, they should show signs of plastic yielding (stretching) before failing. Are they higher tensile studs which may be more susceptible to brittle failure if torqued too tight considering the wheel may be heating up more than usual at the track? This doesn’t explain why without centring rings no studs failed if the torque used is the same either...

As an engineer, I am puzzled by this problem both Jim and Cody have experienced. And is it confined to just the Enkei RPF1's?

More questions than answers in this post I'm sorry to say...

 

[CL65 Captain]

I don't know if it's just the Enkei RPF01s, I just knew that Coday was using the exact same wheels for the black shop car on the track. I called him and he said "don't know why you are snapping studs, but I don't run the centering rings." And after I pulled them the problem never occurred again. Mystery to me. But I have 40+ track weekends without the centering rings and never again snapped a stud.

I'll tell you what, it really sucked pulling into the pits and noticing two snapped on the left side rear and one on the right side rear. I thought maybe my torque wrench is off and I over torqued them. Replaced them at the track and broke two more next time out. Replaced all four corners with ARP. Next time at the track snapped three more. That's when I called Cody.

 

[Tangoman]

Additionally, going from grade 4 to say grade 8 stud will give a higher tensile strength but now the stud will also be more susceptible to a brittle failure (admittedly at a higher yield) ! Yes?

However, when you trace the load of the wheel hitting a curb, that should be transferred through the wheel and through the (properly fitting) centring ring to the hub and then the suspension and the studs continue to only have tensile stress not a shear stress...

If the studs are failing with centring rings, are they OEM studs? if so, they should show signs of plastic yielding (stretching) before failing.

Are they higher tensile studs which may be more susceptible to brittle failure if torqued too tight considering the wheel may be heating up more than usual at the track? This doesn’t explain why without centring rings no studs failed if the torque used is the same either...

As an engineer, I am puzzled by this problem both Jim and Cody have experienced. And is it confined to just the Enkei RPF1's?

More questions than answers in this post I'm sorry to say...

Yes, moving to a higher hardness and a longer stud will make the joint more resilient to just about everything you could throw at it, but it would be more likely to yield other parts of the system, like the hub itself. I don't think you'll see a stud (even at grade 8) have a brittle failure, it is still a very ductile material. If you get up to Metric 12 stuff then you would be walking the line on that.

The plastic/aluminum centering rings are not able to transfer load. If they did, they would need to be a very close to a press-fit, not a slip fit like they are now. They are used for centering the wheel only to make sure it is in balance - please see this Tirerack tech posting.

As for the evidence of plastic deformation before the stud fails, I disagree - parts fail in fatigue all the time and never get anywhere near yield.

Here are the reasons why a stud should fail in my opinion (in order of the most likely)

1) The stud wasn't torqued high enough to ensure adequate tension in for the joint and as a result, the friction joint slipped and transferred shear and/or bending to a threaded stud. Threads HATE shear and bending, particularly bending and this would fatigue them very quickly (like I said previously, it would not near to be near yield to do it. It wouldn't take long either because if a stud saw 10ksi fully reversing every time the wheel did a revolution you could pick up thousands of cycles very quickly.

2) Something is binding up - perhaps the centering ring spotface on the wheel is not perfectly centered with respect to the tapered holes the lug nuts seat against. When the lugnut is torqued, it actually puts the stud in bending and possibly sends one part of the thread over its yield point. One time isn't a big deal, but repeated wheels on/wheels off action is going to fatigue that stud and fail it.

3) And finally (this is hard to do) the stud was torqued PAST its proof load and yielded, leading to a loss of pre-load and the stud became susceptible to fatigue.

I think 1) is the most likely and that is what I would put my money on. To fix this, I would try torquing to 120 ft-lbs instead of 80 ft-lbs. This should decrease the working load on the bolts by at least 50% which could give you a 6-7x improvement in fatigue life. Using the standard rule of thumb that every 10% reduction in stress reversal doubles fatigue life.

HOWEVER, 2) could explain why a stud would fail ONLY when the centering ring wasn't there. But it is also the hardest to prove. Probably the only way to prove it would be to switch wheels, but who wants to purposefully break their wheel stud and have to replace it?.

 

[TURBO2GO]

Interesting info... I'm torquing at 90 now. Should I go higher I wonder with the arp studs.

 

[nsxtasy]

I spent a good amount of time on the interwebZes trying to get info on polycabonate versus aluminum hubcentric rings. There is a lot of BS out there. Some guys claiming you don't need rigs at all.... some saying plastic ones deform and melt... one guy eve saying he has seen 1000F on the rings and the aluminum ones melt. Now I don't know the melting point of aluminum, but.... Come on....

Wikipedia says it's 1220F. Presumably that's for pure aluminum, and it could be different for the alloys used in hub rings, body panels, etc.

 

[TURBO2GO]

Wikipedia says it's 1220F. Presumably that's for pure aluminum, and it could be different for the alloys used in hub rings, body panels, etc.

Unless its an F1 car, I don't see how most guys would hit 1200 degrees on their hubs at some track day event on basically a street car. I mean the wheels are aluminum. The rotor hats are aluminum and they get a lot hotter. That story just didn't make sense to me.

 

[CL65 Captain]

Unless its an F1 car, I don't see how most guys would hit 1200 degrees on their hubs at some track day event on basically a street car. I mean the wheels are aluminum. The rotor hats are aluminum and they get a lot hotter. That story just didn't make sense to me.

They don't. Front brake rotors may get that high but the hubs never would.