And, exactly what is the warning that the website posts under its calculated torque values?Originally Posted by john f
And, exactly what is the warning that the website posts under its calculated torque values?
Charlie Warner
fatto gatto racing
'Cause there's bugger-all down here on earth!
All bolts have a dry and/or lubricated value. We usually only see them in those applications.
Charlie Warner
fatto gatto racing
'Cause there's bugger-all down here on earth!
I wonder about the assumptions of the calculated torque values in that website. They seem rather high and probably assume some elastic stretch or give to the bolt. You do that too many times and the bolt will fracture. Concerning our wheel studs, I'd rather under torque than over torque. If one of the live axle wheel studs snaps off, you're along for the ride. At least in under torque you can usually sense something is wrong if a wheel goes loose.
Misleading information on the interweb?...really?
aaron
While this doesn't change the assertion of my last post, I ran a torque calculation for a Grade 5, 3/8 fine bolt (dry), suggested torque is 31 ft/lb...sounds reasonable to me
aaron
My point in showing that site was that a 3/4-16 center stud torqued to 120 or 140 lb-ft
is not loaded that great. If you look at other sites you will find approximitly the same numbers. One would think that the center bolt is similar in strength to a grade 8 bolt and when compaired to those numbers my opinion is that I could add 10,20,30% or more to the current torque numbers and not "break" or come close to yielding or deforming the center clamping stud.
"Please keep in mind that the above calculations are merely approximations. You should always consult with the manufacturers for torque data since significant variations may exist even within the same grade and size. "
A statement like this is a typical "cover your a$$" web statement in my opinion. As a mechanical engineer I lothe statements like that. Do the calculations and go with the results. Just my $0.02.
john f
That would make a great epitaph on a tombstone.
What material is the "center bolt" (I assume you mean the stub axle) made of? What grade steel. Rolled threads or lathe-cut threads? Heat treated? To what Rockwell?
You have to ask the manufacturer about these things. If you buy from Doug Learned he can tell you exactly what material spec was used and you can source an exact torque value for your application. If you are using an OEM axle then you have to trust the manufacturer's specs.
It ain't rocket science . . . . but it is science.
Charlie Warner
fatto gatto racing
'Cause there's bugger-all down here on earth!
I wouldn't get too worried about 160 torque unless the studs are real crap.
Let's go through some quick and easy calculations (the numbers are not meant to be really representative of your car, so you will want to use your own at some point):
Assuming that we a re talking about a 3/4-16 stud, to induce 100,000 psi stress in the thread roots, Unbrako recommends (dry, unplated bolt) 400 ft.lbs (some other sources use 390). That will put the stresses at approximately a bit over 1/2 of its tensile strength. No, they don't list the yield strength or fatigue limit, but if the material they use has properties anything like 4340, the fatigue limit at that strength will be somewhere near 96Kpsi (going off of memory, so I may be off a bit).
Simple math tells us that the stress induced by the torque will be at a correspondingly lower level for whatever torque you use, so if you are torquing to 125 ft.lbs, the stress induced will be 125/400*100000 or 31250 psi - not much.
However, you also have to take into account the thread conditions - rolled or cut - and the profile ( mainly what the root is shaped like, since we are all talking V-threads here). Each will have its own characteristic as to how it will induce stress concentrations - a rolled thread is better than a cut thread, etc. Unfortunately, those are factors that are nearly impossible to accurately put number to, so we will ignore them for the moment.
Bending loads also come into play - none of the surfaces you are working with are perfectly parallel when you add everything up, and the nut will never sit perfectly square on the stud threads - it will always be slightly crooked because of frictions at the thread interface - so the load will never be spread perfectly evenly over all the threads (even if we didn't take into account how the stud stretch affects thread fit, which will make stresses even worse). Again, another factor we cannot put hard numbers to.
Wheel load is another factor - you are essentially working with a lever, with the nut being the anchor. For a greatly simplified example, let's assume a 23 inch diameter tire, a 5.0 inch dia hub face (max dia where it sits against the hub) and 1000 lbs lateral force generated at the contact patch, the load induced on the bolt will be 11.5/2.5*1000 or 4600 lbs. Assuming a thread root dia of .672" (for an area of .3546 in sq), that means an additional stress of 4600/.3546 or 12,970 psi, for a total now of about 42,200 psi - still at somewhere under 1/2 the fatigue limit (but more probably a bit over half considering any stress concentrations and any bending loads), and at maybe 1/4 to 1/3rd the yield stress.
Upping the torque to 160 adds another 5000 psi stress at the root, making the total now around 62kpsi - still somewhere below the probable fatigue limit, even considering stress concentrations.
However, I would guess that most studs are heat treated to only somewhere around 150 - 160kpsi (30-40k lower than what Unbrako makes its cap screws to), so the fatigue limit is probably closer to 75-80kpsi - still leaving you with a cushion.
Even if the induced loads are above the fatigue limit, all is not lost as long as they are still well below the yield point - all that will happen is that you will have to life the stud out instead of expecting it to last forever.
Thanks Richard. And, for the record, there have been several runs of "crap" axles marketed in the past. Not deliberately, but by those not up on metalurgy and they went the cheap route. Ralts had a reputation for breaking stub axles to the point that they were replaced if the car took an impact on the wheel or was dropped off the jack.
My main point was that if an increase above the recommended torque settings was needed to keep the wheel from loosening then the real issue is being masked.
Charlie Warner
fatto gatto racing
'Cause there's bugger-all down here on earth!
I agree with you, Charles, and only posed all that to show that slight over torquing isn't all that much of a problem, and if it "cures" the real problem at least temporarily (if the real problem can't be found easily enough), one need not be overly concerned.
That isn't to say that the problem has disappeared - it needs to be found and fixed as soon as possible!
One thing everyone keeps forgetting in this business is that as cars get faster over the years, stresses rise and things that worked fine for the last 50 years now start having problems - sometimes the problem was there all along, but didn't manifest itself at the lower stress levels. Weakest link, etc.
One overlooked area for most guys is the wheel/hub interface. If the surfaces don't mate up perfectly, wiggling can/will occur, and all the extra torque does is delay the loosening.
Check your hub faces to make sure that they are actually flat, and I don't mean check them with a straight edge - check them with a height gauge on a surface plate or trued up in a lathe. It is conceivable on old hubs that they have been displaced to a convex shape over the years, causing the wheel to clamp down tight at the center only.
Check the parallelism of the 2 faces of the rotor hat. If the thickness is tapered (thinner at the outer diameter), the wheel will again clamp down solidly at the center only.
Check the mounting face of the wheel - often they will wear convex shaped (worn down towards the outer diameter) over the years, causing the same clamping issue.
The most usual issue we see, though, is that the wheel sits down on the radius at the base of the centering snout first without ever actually clamping solid on the rest of the hub face. If it is just barely touching, it may take quite a few laps for it to work loose, but when the interference is enough, the wheel will come loose in a couple of laps.
Something I've never figured out why it's done, but all wheels have their hub face machined flat when it should really be a couple 'thou concave to ensure that the wheel contacts the hub at its outer diameter first, and then drawn in to full contact by the wheel nut and washer. The actual depth of the center will vary with different wheel designs and applied torques, but in general, a 2 thou drop will at least ensure proper full contact of the faces.
It's easier, and costs less to machine it flat. That's why. You know that...
Last edited by DaveW; 02.26.12 at 11:52 AM.
Dave Weitzenhof
Thanks for the further enlightenment Charlie and Richard.
Yea - changing that final Z dimension on that one line would really break the budget!It's easier, and costs less to machine it flat. That's why. You know that...
That assumes, of course, that the wheel seating surface was done on a CNC lathe. If not, then a mechanical setup change would be necessary. I am not convinced, considering the low quantities and large variety of wheels produced by most race-wheel manufacturers, that they would use a CNC lathe for that cut.Yea - changing that final Z dimension on that one line would really break the budget!
That's why I said what I said.
Dave Weitzenhof
Actually, the surface should be curved rather than conical, so you'd have to add a few words to the G code.
I never quite understood this approach to center lock wheels. It doesn't seem very efficient to transmit side load torque to the bearing with one fairly flexible stud in the center clamping the wheel to a hub face. You need quite a lot of preload to prevent loss of contact on one side.
Is it a result of conversion of upright designs to center lock? The approach makes more sense if you have four or five (or more) wheel studs around the periphery of the hub, like you see on street cars and trucks (and in NASCAR). Especially given the tapered seat on the lug nuts.
Nathan
I think the practice of center studs started when tires had to be changed during pit stops - it's a lot faster to change just one nut than 4 or 5 - especially without air tools as it was back at the beginning. At least we don't still use knock-offs like what you saw in the '20's thru '60's !
It would be interesting to do a cost/weight/stiffness study of and optimized centerlock setup against a 4-lug. I'd be wiling to bet that the 4-jug can be made better all around - I'd be willing to bet that the 4-lug can be better in all 3 in most/many cases.
Unfortunately, for most current club racers, 4-lug isn't "modern" enough, even if it were shown to be more efficient.
I should have emphasized this approach to centerlock mounting, with a small stud used only for clamping. The other approach is to use a larger diameter spindle and a tapered wheel nut that locks the wheel directly onto the spindle. Then you are not reliant on clamping preload to resist cornering loads in the same way.
If you are going to clamp the wheel to a hub, then four or five smaller studs spaced around the periphery would be less expensive and much more efficient.
Nathan
It's called low-dollar amateur racing
In comparison to the large dia, threaded, hardened steel hub snout, the 3/4" stud arrangement is cheap and doesn't take much tooling-wise to make. It also allows the hub to be made from aluminium - again, a lot less expensive when made in small quantities (though a bit more bulky).
It doesn't seem to be less expensive based on prices I see for Van Diemen parts: $375 for the hub, $100 for the center stud, >$100 for the wheel nut and spacer.
Given no one really does pit stops at this level, wouldn't it be cheaper and work better to have four or five studs threaded into the hub and use lug nuts? Probably lighter, too.
Nathan
So . . . . now we are really reinventing the wheel?
Charlie Warner
fatto gatto racing
'Cause there's bugger-all down here on earth!
Production and retail costs are two entirely different things!
True that!
Sorry, Charles, didn't mean to go off on a tangent, just trying to understand this design approach. Were they originally conversions from a four stud design?
Thanks,
Nathan
Not positive, but it was most likely Bruns with the DB1. Not everybody copied that right away, but it didn't take long for it to become a "must have" item.
Nathan,
When you say 4 or 5 studs is cheaper than one center stud, are you including the cost of hiring someone else to do tire changes?
Having spent lots of time with both systems,
I vastly prefer the 4 bolt approach to centerlock, but one has to use what is available or make it themselves.
In the multi bolt design things are generally at much lower stresses and do not fail like centerlock bolts have an annoying tendency to do.
There are currently 3 users browsing this thread. (0 members and 3 guests)
Posting Permissions
Reply With Quote