center-lock hubs and wheels

[Wack Attack]

I'm researching center-lock hubs and wheels for next year and was wondering if people who have done it can tell me if they/it:
1. reduced total unsprung weight of wheel and hub package.
2. considered a tapered area between the wheel and hub to maximize contact area or went with a splined solution.

[jack]

2: you can also use pins, similiar to F1

accually, i was just wondering where people get the nuts for center-locks? i was going to just look at summit racing or something. any good sources?

[Denny Trimble]

What's that? A purchased part on a WWU car?

http://fsae.com/groupee_common/emoticons/icon_smile.gif

[RaID]

1. not really due to the use of centre-lock hubs
but rather due to the design of the uprights in our case

2. were using a triple drive dowel approach

and we are manufacturing our own centre-lock nuts
that way they suit our design exactly

[Greg H]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by jack:
accually, i was just wondering where people get the nuts for center-locks? <HR></BLOCKQUOTE>
Jack, we are going to center-locks this year and are planning on manufacturing the nuts and possibly the socket ourselves.

[jack]

thats what i was afraid of--i thought we would end up making them. denny guesses right http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Wack Attack]

I don't think F1 uses pins to keep the wheels from falling off. That would slow down tire changes. I think the rear center-lock lugs are left-hand threaded so they tightened under acceleration and the front are right-hand threaded so they tightened under braking.

[Denny Trimble]

I heard the commentators say that F1 teams are required to run a retention device that retains the wheel in case of nut loosening. They were quite baffled at how the Minardi managed to lose a wheel in China...

I'm sure it's automatic and activated by the air gun.

[jack]

here are the pins i was talking about:

http://scarbsf1.com/testing_2003/pict0151.jpg

[John Bucknell]

Jack's got the right picture.

That purple anodized pin is the locking mechanism. The gun pushes it in and the linkage connected to the central pin yanks the peripheral pins in (or in some cases they are just ball bearings). The tire man pulls the pin back out after the wheel change and torquing of the nut.

[95M3Racer]

What type of thread are normally used on these centerlock stub axle/nuts?

Thanks

[95M3Racer]

Also, where have people bought cneterlock nuts?

[Amos]

We are using centre locking wheels this year, we made the wheel nuts - obviously not much use for where to buy them from.
But the thread we are running is M36 fine. (In American language about 1.25 to 1.5 inch)

Cant tell ya if they work for about another week and half when we get it on the track.

Dunno how to post pics, but see http://formula-sae.adelaide.edu.au and you might catch a pic of the wheelnut (if i convince the legend in our team that updates the site).

[rajpr]

HRP sells them. They are about $60-80

[Chris Allbee]

We use the center lock nuts from a dirt racer. I think it may be a sprint or a midget, not for sure, but they are really common in dirt circle racing.

[Joseph]

Here is the center lock nut(Knock-off nut to dirt track racers) and wheel that Chris is talking about. The nut is 7075 Al anodized and only cost 30$ a piece!(Bishop Racing, Oklhoma City OK) But we did have to machine everything else.



http://students.ou.edu/K/Joseph.D.Kliewer-1/pics/Wheel.jpg
http://students.ou.edu/K/Joseph.D.Kliewer-1/pics/2004%20Font%20Hubs.jpg

[alfordda]

Joseph-

Why does the one on the right not have drive pins?

[fade]

because the pic was taken as they were being manufactured. they all got the pins eventually

[jack]

joseph, are those spindles in the picture for both the front and rear? do you thread your CV's onto the spindle, or do you have ones for the rear with splines? i'm having a little trouble figuring out exactly how to attach the CV to the spindle on the inside of the rear upright. i imagine most do it with splines. any pics?? thanks.

[Denny Trimble]

Did you get any flak from the tech inspectors on the "critical suspension fasteners must be Grade 5 or better" rule, regarding aluminum wheel retention nuts?

[Joseph]

Jack, both spindles in the picture are for the front. The rear are the same but without threads and with internal splines for a stub axle. Although that design has been obsoleted out in favor of a lighter more compact design, we also just got a CNC lathe, which changes things a bit.

Denny, No trouble at all. In fact the nuts were kind of overkill. We would have preferred to go with fewer threads and a larger ID but for $30 I was not about to machine my own.

[fsae racer]

For the pictured hubs, how did you handle the brake torque? Did the direction of the threading just use the brake torque to tighten the hub onto whatever it was threaded, or were the threads just for a nut to hold the hub into the upright with brake torque handled by other mounts? For the rear hubs, with internal spline, how did u secure the hub to the shaft axially?

[Wack Attack]

Why do people use centerlock hubs and wheels? I didn't find any real advantage from what I've read and looking at other people's design. It seems that most teams are using pin-drive centerlock setups which do not save any significant weight and add complexity to the design because of the tremendous amount of torque needed to tightened the nut.

[fade]

the brake torque was handled by the caliper mount.
as you can see in the first photo high strength glue holds the stub shaft in the hub

[fsae racer]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by fade:
the brake torque was handled by the caliper mount.
as you can see in the first photo high strength glue holds the stub shaft in the hub <HR></BLOCKQUOTE>

Wait, now I am totally confused. high stregth glue for what? As for the brake torque, I understand that the torque is transfered to the car via the caliper mount on the upright, but how does the hub handle the torque from the tire to the rotor? On the rear, what did you guys use to handle the drive and the brake torque? thanx.

[fsae racer]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Amos:
We are using centre locking wheels this year, we made the wheel nuts - obviously not much use for where to buy them from.
But the thread we are running is M36 fine. (In American language about 1.25 to 1.5 inch)

Cant tell ya if they work for about another week and half when we get it on the track.

Dunno how to post pics, but see http://formula-sae.adelaide.edu.au and you might catch a pic of the wheelnut (if i convince the legend in our team that updates the site). <HR></BLOCKQUOTE>

Amos, where did you get the wheels that you have pictured on your 2004 car? I have been looking for center locking one piece wheels besides the BBS and OZ wheels, but nothing I have found looks as good as the wheels you guys have. Beatiful all around car.

[clausen]

Hi,

We made them ourselves. Are you interested in buying a set? http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Joseph]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR> how does the hub handle the torque from the tire to the rotor? <HR></BLOCKQUOTE>

It's the same drive pins that drive the wheel. Why use a bunch of extra fasteners and mechanical links to transmit torque. Just make a break hat that slides on before the wheel and use the pressure from the center lock nut to hold the whole thing together. Notice the shiny hat and floating rotor just behind the wheel.

The high strength glue is proprietary info, not to be discussed in an open forum.

[Denny Trimble]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by clausen:
Hi,

We made them ourselves. Are you interested in buying a set? http://fsae.com/groupee_common/emoticons/icon_smile.gif <HR></BLOCKQUOTE>

Nice! How were they made? Cast, or multi-piece?

[clausen]

Denny,

They're 2 piece machined and spun aluminium. The design is more or less the same as chalmers wheels. The idea was to also use a carbon inner part, but we never had time. Next years guys might.

They still give as a huge negative offset with room for big brakes right out against the machined part.

And they look pretty cool http://fsae.com/groupee_common/emoticons/icon_smile.gif So I guess they were worth doing.

[Denny Trimble]

Very well-done, you're giving me ideas... http://fsae.com/groupee_common/emoticons/icon_smile.gif

[clausen]

Well we can't take credit for the idea. We were only copying Chalmers http://fsae.com/groupee_common/emoticons/icon_smile.gif

[ODUformulaRacer]

we used the center nut design on the rear of our car and have had no problems once we started cranking them down to 75 ft*#. we modified the ends of a sprint car axle to mount to the taylor tripod housing. then we just spun the nut on crank it down and go. kiaser wheels are actually designed for that spline and we got the hook up from a manufacture who cut off the ends of broken axles. we then just welded on a flange and bolted up the housings. we did however use reverse thread on each side and mark the thread and nut with a paint pin for quick checks

[fsae racer]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Joseph:
<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR> how does the hub handle the torque from the tire to the rotor? <HR></BLOCKQUOTE>

It's the same drive pins that drive the wheel. Why use a bunch of extra fasteners and mechanical links to transmit torque. Just make a break hat that slides on before the wheel and use the pressure from the center lock nut to hold the whole thing together. Notice the shiny hat and floating rotor just behind the wheel.

The high strength glue is proprietary info, not to be discussed in an open forum. <HR></BLOCKQUOTE>


What size are those wheel nuts that you guys got from bishop? thread and socket size would be hugely appreciated.

[Marwan]

Originally posted by Wack Attack:
I'm researching center-lock hubs and wheels for next year and was wondering if people who have done it can tell me if they/it:
1. reduced total unsprung weight of wheel and hub package.
2. considered a tapered area between the wheel and hub to maximize contact area or went with a splined solution.

Hello, wouldn't mind raising this question again...

For those using/used single lock nuts, how did you go about fastening your wheel centers to your stub axles (or hub?)?

The issue I'm seeing is if, for example, the wheel center is Al and the stub axle is steel then a spline solution would seem to be tricky.


Originally posted by RaID:
2. were using a triple drive dowel approach

What is this exactly? Like pins or something?

I only really started looking into this stuff tonight so any other advice would be appreciated too http://fsae.com/groupee_common/emoticons/icon_smile.gif

Cheers,

Marwan

[Chris Allbee]

When using single lock nuts we fastened the wheel center to the hub with...a single nut. I'm not sure exactly what you are looking for with that one.

You are right in your assumption with the dowel approach. Its just a fancy word for "locating pins".

One thing to keep in mind when machining your own threaded parts: do not ignore accepted standards. If your threads are not to the standards you want (close tolerance, etc.) it won't hold torque like it should, or like you expect it should. Always provide for a positive stop, preferably with less than a 1/16th of a turn before the stop is hit when backing off.

Aluminum splines would be large and ungainly, not sure if it would be worth it in this case, but thats for you to decide in your application.

[KU_Racing]

A few people seem confused about how a center lock system transmits wheel torque to the car. The wheels either have a taper-splined ID that fits into the splined pattern on the hub, or machined areas in the backs of the wheels that fit over dowel pins on the stub axle to transmit torque. The center lock nut itself is not loaded by the drivetrain or braking torque- its only purpose is to retain the wheel onto the stub axle.

Also, somebody mentioned something about how the threads on one pair of wheels are reversed. This is correct, but the common practice is to design the system so that all the nuts see a 'tightening' moment under vehicle acceleration, i.e., the center lock nuts on the right side of the car are left-hand thread.

For the teams that are running center locks, what rationale made you decide to go with a center lock system? Seems to me that the only reason for center locks in most series is to speed tire changes, so what made you decide to go with center locks in a series where there arent any, especially considering the additional cost and design issues??

[Scorpio]

Originally posted by KU_Racing:
For the teams that are running center locks, what rationale made you decide to go with a center lock system? Seems to me that the only reason for center locks in most series is to speed tire changes, so what made you decide to go with center locks in a series where there arent any, especially considering the additional cost and design issues??

Weight reduction.

[Chris Allbee]

In response to no tire changes in FSAE. I recall a moment in 2004 I believe it was when we were in line for the autocross event and rain moved in. Track conditions went from dry to damp to wet very quickly and many teams scrambled to change their wheels in line. With the center lock system we were able to change to our intermediates and get out onto the track before it was completely soaked while everyone was still trying to get their tires off. Granted, doesn't happen all the time, but when it does...

[Chris_S]

Originally posted by Scorpio:
[QUOTE]Originally posted by KU_Racing:
For the teams that are running center locks, what rationale made you decide to go with a center lock system? Seems to me that the only reason for center locks in most series is to speed tire changes, so what made you decide to go with center locks in a series where there arent any, especially considering the additional cost and design issues??

Be prepared to change tyres quickly then, cordless impact gun to get the nuts off (faster than spinning the nuts off by hand). Then have someone with the right torque wrench to tighten the nuts once the wheels are changed.

[Chris Allbee]

or just use a big ass breaker bar and a quick foot. and god forbid you actually use a torque wrench on an FSAE vehicle.

[Marwan]

Originally posted by Chris Allbee:
When using single lock nuts we fastened the wheel center to the hub with...a single nut. I'm not sure exactly what you are looking for with that one.

You are right in your assumption with the dowel approach. Its just a fancy word for "locating pins".

One thing to keep in mind when machining your own threaded parts: do not ignore accepted standards. If your threads are not to the standards you want (close tolerance, etc.) it won't hold torque like it should, or like you expect it should. Always provide for a positive stop, preferably with less than a 1/16th of a turn before the stop is hit when backing off.

Aluminum splines would be large and ungainly, not sure if it would be worth it in this case, but thats for you to decide in your application.

Alright, my wording was a bit off then. What I meant was how did you transmit the torque between the hub and the wheel, e.g. tapered spline, pins/dowel etc.

At the moment we are using pins/dowel approach. I'd like to implement a single nut just for simplicity and speed in use. Granted, there will be issues in design, analysis, manufacture etc but if a robust, cost effective solution is there then I believe that implementing a single nut system is quite suitable in FSAE. Even though we may not do many tyre changes at the comp, we'll still do a lot leading up to and after it.

Cheers,

Marwan

[Chris Allbee]

We use the pin/dowel approach. It has some benefits over splines in that it is generally cheaper to ream a set of holes and press pins in than have a set of splines cut. Obviously the splines could provide a robust solution as well if cost wasn't as much of a concern.

[KU_Racing]

I understand the argument about saving time to some degree, but I know for a fact that we have rolled up to an event with the wrong tires on, and gotten the right ones on in say, 2 minutes. Not nearly as fast as center locks, but fast enough for 99% of FSAE situations.

For the weight reduction argument, would you be willing to share how much weight you saved between the two setups??

[Marwan]

Originally posted by KU_Racing:
I understand the argument about saving time to some degree, but I know for a fact that we have rolled up to an event with the wrong tires on, and gotten the right ones on in say, 2 minutes. Not nearly as fast as center locks, but fast enough for 99% of FSAE situations.

For the weight reduction argument, would you be willing to share how much weight you saved between the two setups??

To be honest, I wouldn't expect that much saving at all. It's more of usability decision, but you do make a good point about how much you are actually saving.

[J. Vinella]

It is obviously to see that a center lock is faster to change than a 4 lug wheel, but you mustn't forget the retention devise as required by the rules. Each cotter pin or whatever device you use will also add time to the wheel change. Once to disengage the devise at the beginning of the change and once to put it back in at the end. You can add up the operations of a center lock and 4 lug to see the advantages.

[KU_Racing]

what retention device do you mean? I am not familiar with this section of the rules, as we have only ever run a 4-lug setup.

[J. Vinella]

3.2.2.1 Wheels

...

Any wheel mounting system that uses a single retaining nut must
incorporate a device to retain the nut and the wheel in the event that the
nut loosens.

[Wesley]

Think cotter pin on a castle nut - basically its there to keep your wheel from completely detaching.

[Scorpio]

Originally posted by KU_Racing:
I understand the argument about saving time to some degree, but I know for a fact that we have rolled up to an event with the wrong tires on, and gotten the right ones on in say, 2 minutes. Not nearly as fast as center locks, but fast enough for 99% of FSAE situations.

For the weight reduction argument, would you be willing to share how much weight you saved between the two setups??

0.6 lb

[KU_Racing]

yeah that rule makes sense... I was a little confused at first, because the way I read J. Vinella's post, i was thinking that the rules implied that a secondary lock was necessary for all wheel fasteners, which would really suck if there were 4 on each wheel.

Scorpio, is that .6 lb at each corner, or .6 lb over the whole weight of the car?

I guess I need to also consider materials... If we were to switch to a single center nut, I suspect our system would get heavier (starting with a hub/upright setup that I believe is one of the lightest around)

~2.5 lbs unsprung AND rotating mass is a pretty good savings though. Has anyone measured how much a center lock wheel weighs compared to a 4-lug wheel? I would guess that A center lock wheel would need to be heavier, since you cant hollow out the center section and the shoulder that the nut seats on has to be pretty beefy to be stiff.

[Chris Allbee]

about 1.1 pounds for a wheel center

[KU_Racing]

is that for a center lock or for a 4-lug

[Chris Allbee]

center lock

[rjwoods77]

Hyper technologies makes really nice forged magnesium wheel centers for 27 tooth sprint splines that are incredibly lightweight.

[poweredbyvdub]

it looks like cutting threads into the centerlock hub is pretty much standard practice, but is there any way to get like a blank hub off of something with rolled threads?

for those who cut threads, any special considerations to take into account when cutting the threads?

ive seen some aluminum setups, and how do those stand up to the torque on a centerlock nut? what alloys do you guys run on those things? it seems like going with a steel centerlock hub wouldn't save much weight, but thats just a gut feeling assumption.

[skillet]

there are K factors for designing threaded and rolled threads.
we run centerlocks because its cool of course http://fsae.com/groupee_common/emoticons/icon_smile.gif

[KU_Racing]

Sweet. I want our car to have center locks so that I can expense out one of those custom carbon and titanium lock nut impact guns that F1 pit crews use. http://fsae.com/groupee_common/emoticons/icon_cool.gif

[Jersey Tom]

Using steel nuts this year for wear resistance. I'm just too wary of people destroying aluminum threads.

Centerlocks.. design I did this year I feel is easier and cheaper to manufacture, easier to mount / dismount, and looks tight.

http://durning.colorado.edu/sae/Media/thumbnails/wheelassy4.jpg (http://durning.colorado.edu/sae/Media/wheelassy4.jpg)http://durning.colorado.edu/sae/Media/thumbnails/wheelassy6.jpg (http://durning.colorado.edu/sae/Media/wheelassy6.jpg)

The hub itself is only 0.8 lb and is universal to all 4 wheels of the car. Threads are cut, 1 1/4 - 18 UNEF, Class 2. Hub is 4130 hardened to Rc35, nut is 1144, Cond N. Drive pins are also 1144.

Using 5 dowel pins to transmit torque. Hell of a lot easier and cheaper to do at this level on a one-off than a spline.

[Chris Allbee]

Nah...center-lock is SOOOO last year! This year we need a .028" all Aluminum gas tank. I think I read somewhere that it has a high enough burst strength....

[KU_Racing]

you read correctly. http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Dave K]

Quick question: we are running a center-lock setup this year for the first time. I am trying to figure the packaging for the rear.. since the stub axle goes through the splined spindle, you need a nut to secure that as well as the large locknut on the outside of the wheel. This small retaining nut (7/8 or 15/16) needs to be pretty deep inside the spindle... my question is if you guys are running this kind of setup, how do you retain that small inner nut? do you just suck it up and jam a cotter pin way down there, or is there another way to retain it? i saw someone said they use some glue to keep the stub axle in the hub...

[Dave K]

Here's the one i am talking about...

[Dave K]

using the picture, here's the one i am talking about...

http://img81.imageshack.us/img81/6265/wheelwp8.jpg

[BenB]

that looks like a serrated lock nut to me with the little notches around the outside. That should work but those normally are hard on threads.

Nylon locking is not a good way to do it because the brake heat can melt the nylon. I think Loktite (or glue as you mentioned) might also have the same problem. Loktite should give temperature specs, and you can probably estimate the temperature here or measure it after doing a lot of laps with your car.

Safety wire might be just as good as a cotter pin, but would probably be easier to thread in there. These would not keep the nut from loosening, but would keep it from coming completely off.

I would be interested if somebody has any other ideas.

[Travis]

Hello all,

I am Travis with Taylor Race Engineering. I have taken a picture of our wheel hub and centerlock axle combination to show you a common and effective way to retain the stub axle nut into the wheel hub. Can someone help me post a picture? or Email me and I can send you the picture and you can post it.

Travis Townsley
Taylor Race Engineering
travis@taylor-race.com

[rjwoods77]

Dave,

That nut looks like it is just pulling the cv joint snout in tight. Using a all metal locking nut or a nylon lock nut will suffice. The joint wont get hot enough to melt the nylon and the shaft material is hard enough to use an all metal locking nut. Almost every fwd car has a 30mm nut to do the sme job without any extra locking mechanisms like a cotter pin or such. You will be more than safe and withing the rules with a sps technologies flex-loc nut or the standard NAS/AN all metal ellipitcal locking nut. As far as what the nut is engageing you need to just run a washer underneath the nut that engages a shoulder within the splined stub axle. Just for you information DMI supplied us our splined stubs at a decent price with quick turnaround. They were made our of 7075-t56 like their sprint axles are.

[exFSAE]

Nylock works just fine.

Checking its tightness now and then is always a good idea tho.

[Dave K]

Eh, i have never measured the heat that is generated in there, but i think nylocks or loctite would be kinda sketcy... just because there is no positive stopping, just in case. The safety wire sounded pretty good to me, that should work fine....

[Dave K]

btw, Travis:

www.imageshack.us (http://www.imageshack.us)

go there and you can host images and stuff http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Chris Allbee]

Thats a picture of our '04 car. That was actually our first attempt at the center-lock and you have it right with assuming the stub axle is the thing being bolted in the middle there. The stub axle is then attatched to a tripod housing (Taylor Race standard setup). We used a nylock for the positive locking. No glue.

[Kyle Walther]

If you are going to center lock wheels look at completely redesigning you hub/cv/upright system. the whole stub shaft hub combo is very unnecessary and it is pretty easy to just combine the two if you have the machining abilities. It will produce a much lighter and stiffer part because of the geometries required to house the cv inside the hub. the car pictured is our 04 car and after we did this we quickly switched to the integrated hub/cv in 05. and actually saved money by eliminating the splined bits.

[edit] dang chris got to it first. hey congrats on the job!

[Travis]

Dave,

Thanks for the help with the image hosting.

http://img69.imageshack.us/img69/1603/mvc775fwv6.th.jpg (http://img69.imageshack.us/my.php?image=mvc775fwv6.jpg)

This picture shows the common solution to positively retaining the stub axle nut. I can't take credit for this design, but wanted to pass it along. As you can see, the wheel hub is drilled and tapped in four places for a 10-32 bolt. When using a six sided nut, there should always be a hole available to install the 10-32 bolt. This bolt will keep the nut from backing off.

Travis
Taylor Race Engineering
travis@taylor-race.com

[J. Vinella]

From the picture you would then need to safety wire that bolt, or have a nyloc or jetnut on the back to provide retention as per the rules.

Still I like the design.

[fade]

using a nylock or flexloc nut will work for the drive axle. crimp locks will likely destroy the threads. if you are using a bearing assembly without a preload spacer then you may want to look at a different style like keyed locknut.

[Dave K]

I am a bit skeptical about the nylock, but if others use it, then it must work. What do you guys usually torque them to? On our stub axle, we have a 5/8-18 thread (taylor race polaris stub axle). We normally torque until tight, then back it off about a quarter or half turn or so to make sure the bearings still spin nice.

[fade]

The bearing used in the picture you posted is a dual row angular contact whose inner race determines the minimum preload. It sounds like you are using dual row tapered rollers. The design in the pic uses a full height Nylock, your method of preload would rely soley on the locking mechanism of the nut. Using a preload spacer (see "tune to win") would allow you to properly preload the threads in the joint. You might look at keyed spanner nuts on the back side of the upright if you want to keep your preload adjustment as is.

[BBolze]

I've been pondering how professional racing ensures proper wheel bearing preload with center-locks. Fade was eluding to it but I was wondering if anyone knew what for example Le Mans cars do?

I can't think of an application more demanding on center-locks than LMP and would probably be the designs to look at for good insights. With what must be over 1000 wheel changes by all the teams in those 24 hrs, all done by an impact gun (not the most precise equipment compared to a torque wrench) how do you ensure the bearing preload is correct so that the bearings don't burn up on mulsanne?

I've got two idea's that come to mind;
Someone had mentioned that the choice of going to centerlock allows you to integrate many functions of the rotating assembly (in the case of rear axle, load path as well) into fewer parts. One way to do this would be to machine a shoulder into the axle mid span which rests on the wheelbearing (outboard shoulder) while threads on both ends of the axle allow 2 nuts, a wheel nut and an axle nut. The shoulder effectively separates the two preload torques from eachother.

I don't like this idea for the rear axle though as it makes good engineering practices such as taking the threads out of the load path extremely tricky.

Another way to do it would be the two half axle as shown in the '04 picture above, however this increases parts counts and necessitates a 2 piece axle ... never going to be as rigid or stiff as a one piece axle which is the highest priority of an axle.

Does anyone have something to share which offers a third alternative?

[Z]

BBolze,

Some comments. Firstly, if you are only going to "snug up" the bearings for preload then you should have the inner races NOT rotating, such as on the spindle of an undriven front axle. If the inner races and shaft rotate (eg. a driven axle) then a spacer should be fitted between the inner races and everything clamped VERY tight. Adjust spacer length to set preload. Without the clamping, the rotating load together with the small clearance between shaft OD and inner race ID, allows the races to "walk" around the shaft, ultimately causing fretting and trouble.

Secondly, a conventional centre-lock hub is an abomination. It relies purely on friction between the mating flanges to stop the wheel moving, regardless of how many "pegs" are fitted. It just takes a bit of "orbital" fretting motion to undo the nut. As long as I can remember there has been at least one top level race car per year lose a wheel because the centre-lock nut came loose.

Rather than two flat mating flanges with axial pegs (with the pegs needing clearance in their holes to allow assembly, but then allowing orbital motion), a better solution is a "conical" or "face spline". Here radial teeth on the two parts are pressed together by the nut, thus constraining all degrees of freedom and preventing any motion. I and others have been pushing this idea for decades, but no one wants to change...

Z

[shark.ashwa]

Originally posted by Z:
a better solution is a "conical" or "face spline". Here radial teeth on the two parts are pressed together by the nut, thus constraining all degrees of freedom and preventing any motion. I and others have been pushing this idea for decades, but no one wants to change...

Z

Z,
Can you post a picture or put me on the right path. I cant seem to understand the restraining mechanism here.

Cheers.

[Z]

Shark,

A face spline is also known as a Hirth Joint (http://en.wikipedia.org/wiki/Hirth_joint), after Hellmuth Hirth, a German engineer who used them for built-up crankshafts for racing engines in the 1930s (eg. Auto Union, etc.).

A conical spline is a tapered spline, similar to the joint between the Pitman arm and the shaft poking out the bottom of most steering boxes. In this case the taper is very narrow (~6deg half-angle), similar to a Morse taper, so very hard to separate! Opening out the half-angle to about 45deg or more makes separation easy. A "face" spline is simply a conical spline with half-angle = 90deg, so the mating cones are effectively flat, but with the radial teeth.

A conical spline only needs three teeth (per side of the joint) to sufficiently control all 6 degrees of freedom. Each tooth has two flanks, or contact zones, giving 6 n-lines, with the nut providing "force closure". A 3 tooth face spline can be made quite easily with a hacksaw (just 3 cuts) and file because it doesn't need much accuracy to mesh properly. More teeth = over constrained, so requires more accuracy and deflection of contact zones for complete meshing, but it makes a very strong joint (see pics and ref.1 in above URL).

These joints are quite common, just not in motorsport anymore http://fsae.com/groupee_common/emoticons/icon_confused.gif

Z

[Canuck Racing]

Originally posted by Z:
BBolze,

Some comments. Firstly, if you are only going to "snug up" the bearings for preload then you should have the inner races NOT rotating, such as on the spindle of an undriven front axle. If the inner races and shaft rotate (eg. a driven axle) then a spacer should be fitted between the inner races and everything clamped VERY tight. Adjust spacer length to set preload. Without the clamping, the rotating load together with the small clearance between shaft OD and inner race ID, allows the races to "walk" around the shaft, ultimately causing fretting and trouble.

Secondly, a conventional centre-lock hub is an abomination. It relies purely on friction between the mating flanges to stop the wheel moving, regardless of how many "pegs" are fitted. It just takes a bit of "orbital" fretting motion to undo the nut. As long as I can remember there has been at least one top level race car per year lose a wheel because the centre-lock nut came loose.

Rather than two flat mating flanges with axial pegs (with the pegs needing clearance in their holes to allow assembly, but then allowing orbital motion), a better solution is a "conical" or "face spline". Here radial teeth on the two parts are pressed together by the nut, thus constraining all degrees of freedom and preventing any motion. I and others have been pushing this idea for decades, but no one wants to change...

Z

What?

Are you being serious? Normally you have some good insight Z but you may have gone off the deep end here.

[Z]

Canuck,

Where? And why?

Z

[Zac]

I much prefer tri-coidal polygons to splines.

[Canuck Racing]

Sorry, I guess I should have explained my views. I'm basing this on every PROFESSIONAL hub design I've ever seen, and that's quite a few.

First of all, I'm not sure I understand how you set pre-load on the bearing with the wheel nut, or why you would want to, so I'll chalk that one up to misunderstanding the question.

"Pegs" or "drive pins" are what drive the wheel in any set-up I've seen. The wheel nuts are typically torqued to 2000ft/lbs (yes, you read that right. The "impact wrenches" cost $30K each.) and are opposite handed threads on each side of the car so they are tightened if any slipping occurs. If properly done, positive retention methods should hold the wheel on the hub should the nut fail. It won't be pretty, but it'll get you back to the pits.

The face spline seems like a good idea, but i've seen many a pit crew at work. They don't line up anything. Those teeth would be absolutely destroyed after a few tire changes and, assuming you could even get the wheel on, wobbly as all hell. You'd also have to change out hubs every race. Not really practical at $10K a pop.

[Chapo]

I see no reason why this could not work. The wheel would still be lined up (90% of the way) through the center lock thread. As the wheel is then pushed onto the hub the taper on the teeth will do the final location so that the two objects are concetric. The nut is then tightened as per usual and a a positive retention mechanisim is still used.

I think that it may even be possible to improve the packaging of the whole setup using this drive method. However the cost of manufacturing could be much higher as it requires the cutting of gears, not just bored holes for the insertion of pins.

[Z]

Originally posted by Canuck Racing:
Sorry, I guess I should have explained my views. I'm basing this on every PROFESSIONAL hub design I've ever seen, and that's quite a few.

Canuck,

Yes, and I'm basing my views on watching all those PROFESSIONALLY designed wheels fall off, and that's more than quite a few! And yes, I read that right - "2000ft/lbs" with a "$30K wrench" - and still they fall off...

But first to put this other issue away, I wasn't suggesting that pre-load should be set with the wheel nut. Rather, I was briefly pointing out something that appears in most bearing design manuals. I can go into that further if anyone wants...

Now, regarding the conventional "centre-lock hub", this is a great example of how fashion (and stupidity!) dominates engineering design. I remember talking to Wolfgang Weiss (spelling?) back in about year 2000. Wolfgang is a German engineer with some involvement in motorsport. He told me he had spent years watching Porsches chew up their centre-locks, and he also spent years trying to convince the teams to use face-splines. No interest! The really bizarre thing about all of this is that the face-spline is (was?) one of Germany's favourite sons (see Hirth Joint in above post)!!! http://fsae.com/groupee_common/emoticons/icon_confused.gif

Along with the many racing engines of the 1930s that used face splines, just about all the German aero-engines of that period also used them (one piece conrods, built-up crank). In the USA Pratt & Whitney originally built up their cranks using axial splines (similar to a peg drive hub). About mid-WWII (a bit of reverse engineering maybe?) they switched to face splines, because they worked MUCH BETTER (stiffer, stronger, etc).

I could go on but the following explains what I'm getting at (my emphasis).


"Pegs" or "drive pins" are what drive the wheel in any set-up I've seen. The wheel nuts are typically torqued to 2000ft/lbs (yes, you read that right. The "impact wrenches" cost $30K each.) and are opposite handed threads on each side of the car so they are tightened if any slipping occurs. If properly done, positive retention methods should hold the wheel on the hub should the nut fail...etc...

To briefly repeat what I wrote earlier, Canuck is describing two flat plates that are clamped together SO !!! TIGHTLY that it is HOPED that they don't slip. Of course they do slip, only just enough for the pegs to move around in their necessarily larger holes, but that is enough to undo the nut and lose the wheel (at least once a year in high profile races, but many, many more times in amateur racing).


... i've seen many a pit crew at work. They don't line up anything. Those teeth would be absolutely destroyed after a few tire changes and, assuming you could even get the wheel on, wobbly as all hell. You'd also have to change out hubs every race. Not really practical at $10K a pop.

Utter crap!

How does the pit crew get the wheels on a conventional hub unless they "line up anything". A face spline's teeth are V shaped (rounded crests) so they line up automatically.

"Those teeth would be absolutely destroyed" a lot more slowly than the conventional pegs/pins.

The joint would be "wobbly as all hell"!!!!! http://fsae.com/groupee_common/emoticons/icon_eek.gif

(Sorry, I've got to stop. This is like trying to put brains into statues...)

Z

[Z]

Originally posted by Chapo:
I see no reason why this could not work...
Thanks Chapo (just read your post), ... I'm slowly climbing down off the walls ...

(Ommmmmmm ... think happy thoughts Z ... Ommmmmmm ....)

And there are so many of these joints on agricultural stuff...

And beach umbrellas...

And ...

(Oops, go to your happy place Z ...)

Z

[swong46]

Just to be sure, is this what we are talking about for the face splines?

http://www.gkndriveline.com/drivelinecms/export/sites/driveline/en/media-centre/news/gkn-news/images/0036_-_Axial.JPG_1054758069.jpg

The wheel would be splined to center onto the splined hub?


When talking about preload, most are preloading the bearings when securing the hub to the upright. I would imagine that either the upright or hub would be a slip fit for the bearings to allow you to insert a large spacer inside so the inner races would not be pushed out? Is this the correct way to go about the design?

[Z]

Swong,

Yes, there is a face spline in there, just visible at upper centre of the picture.

The dark blue bell-shaped part at the right is the removable CV joint. This is located to the orange hub at the left by the face spline in the middle, all pulled together by the central greyish bolt. The wheel (not shown) would be bolted to the hub from the far left.

This is a typical (except for the face spline) automotive "double-row angular-contact ball-bearing" type non-serviceable unit. In this particular case the pre-loading of the inner races is done by "swaging" the inner section of the hub over the rightmost inner race (so it can't be disassembled). The inner race would be selected for the right width, then a tight push fit onto the hub, and the swaging then jams it up tight against the shoulder of the hub. The face spline teeth are then machined (or pressed?) into this swaged section.

Note that bearing preload and fixing (here via swaging) is entirely separate to wheel location and fixing (here via normal bolts/studs). This is as it should be.

As for "correct way to go about design", well I guess that is a matter of opinion. Old fashioned tapered-roller bearing type hubs would last a lot longer than the one in the picture (it doesn't take much "overload" before the angular-contact balls start whining). But these sealed units are designed and supplied as a complete package by the bearing companies, so overall are cheaper. Such is progress...

Most bearing suppliers have handbooks available that describe all the little design details, although not always very clearly. Your school should have these books. If not ask the bearing suppliers, mentioning that it is for your school. Also talk to the "Tech Reps" at the bearing companies. Face to face is best, because they can show you different layouts, pros and cons, etc.

Z

[Canuck Racing]

Why would a precision ground spline have less play than a precision ground drive pin?

Anyway, I hate forum arguments so I guess I'll just say that if it's such a good idea, why hasn't anybody switched?

Case and point: http://www.youtube.com/watch?v...U1E&feature=youtu.be (http://www.youtube.com/watch?v=-Mq3gpBWU1E&feature=youtu.be)

[nowhere fast]

Originally posted by Canuck Racing:
Why would a precision ground spline have less play than a precision ground drive pin?

The angle of the 'teeth'.
When tightening the wheel, the joint will not bottom out until both sides of a tooth are in contact, thus there is no play. As Z pointed out; this is not quite true for all teeth when there are many splines due to over constraint.

A pin must be undersized by some amount to allow the wheel to be removed. This under-sizing results in rotational play of the wheel.

[ibanezplayer]

Do Force India's wheels satisfy this?
http://scarbsf1.files.wordpress.com/2010/02/fi_hub.jpg

http://scarbsf1.files.wordpress.com/2010/02/fi_wheel.jpg

[Z]

ibanezplayer, Hard to tell from the photos, but that looks like a conventional peg/pin drive (effectively an axial spline). The wheel has some extra material machined away, turning the holes into slots, but it looks like all the mating surfaces on the pegs are axial (see below for more explanation). I guess the noses of the pegs are rounded and tapered simply to allow easier and quicker fitting.
~~~~~o0o~~~~~


Originally posted by Canuck Racing:
Why would a precision ground spline have less play than a precision ground drive pin?
Canuck, Nathan's got it right.

An axial spline is a conical spline with half-angle = 0 degrees. This unique angle allows the two parts to slide indefinitely (relative to each other) along the axis, without ever making complete contact. This is USEFUL, for example, in gearboxes where it might be necessary to transmit torque with the sliding part in different positions along the shaft.

It is USELESS for a completely constrained 6DoF joint. Hence the conventional peg/axial-spline hub joint requires two flanges, perpendicular to the axis, that come into contact to constrain all DoFs. The axial-spline surfaces (pegs) require clearance for easy fitting, so ultimately the joint relies solely on friction between the flanges to control all 6DoFs.

Canuck, your stressing of the fact that the conventional joint requires a "$30K wrench" to apply the enormous "2000ft/lb" torque in the hope that it "should" work, is proof positive of what a stupid joint it is!

Incidentally, the P&W aero engine I referred to above initially had HEAVY PRESS FIT axial splines in the built-up crank, and still they didn't work as well as the later face splines.


... if it's such a good idea, why hasn't anybody switched?
Because contrary to the views of those in motorsport, it is actually a minority pastime, peopled mostly by egotistical, brain-dead buffoons, who do NOT have to achieve any level of success, such as winning, to stay indefinitely in the sport, so instead they spend most of their time looking for excuses for why they are losing.

A typical excuse is "Awww geez..., we can't afford the $30k wrench, so our wheels keep falling off...".

Z

[Z]

Should anyone want to try a conical spline type wheel attachment, here are some things to consider.

1. Very important! The nut is only required to provide "force closure". That means it must NOT have a conical seat as is usual, because this seat constrains the wheel in the radial direction, and fights the splines. The nut mating face should be FLAT!

2. For structural reasons, a nut with large thread diameter with a coarse thread of only a few turns is preferable. A similar sized bolt can also be used, eg. hub threaded internally (inside the splines), and the bolt hollow for less weight.

3. Cone angle. A face spline with 90 deg half-angle has the machining advantage that both sides of the joint are the same (it is a hermaphrodite, both male and female). A cone half-angle of 45-80 deg (male hub, female wheel) may be better because of greater bearing area on the soft aluminium wheel. Also, if the cone apex lies on the mating plane of the nut, then differential thermal expansion (Al/steel) doesn't effect pre-load.

4. Number of teeth. One or two DON'T work. Three teeth work exceptionally well, but might be a bit bulky, and so heavy. Four is bad... Ten or more can give a thin lightweight joint, but requires reasonably high accuracy (ie. no good with a hacksaw, or rough castings, but no problem with machine tools).

5. Fewer, larger teeth are more tolerant of dirt. The pebbles get pushed into the clearance zones (ie. the valleys, see below). Many fine teeth don't like dirt, because the boulders have no place to hide. (This is like coarse vs fine threaded fasteners.)

6. Tooth profile. Very important! The opposed teeth must CONTACT ON THEIR FLANKS ONLY! No peak to valley contact allowed! So undercut the valleys (generous radius) and chop off the peaks. For a three tooth spline a 60 deg included angle works fine. For many teeth a smaller angle (tall narrow teeth) is good because it allows the teeth to bend slightly for more complete meshing.

7. (Edit+) The teeth do NOT have to be symmetric or equal sized on the two parts of the joint. For example, narrower steel teeth can mesh with wider aluminium teeth, making their shear strengths more equal. Also the teeth can be "saw-toothed" (/1/1/1/1) if the torque to be transmitted is expected to be asymmetric. However, this probably doesn't make much difference, and symmetric teeth would be the best place to start.

Z

[swong46]

Z, is this what you are talking about basically? As you tighten the nut from the left side, the splines mesh up and lock together. And just by looking at the model, it is easy to see that a hack job can work using a 3 spline setup.

Yes, I know the model is very bad but I just made it very quick so I understand this debate a little better.

https://lh6.googleusercontent.com/-HJffTA1InUc/TsSATwpCbNI/AAAAAAAABKI/RI6BJVRDsWc/s1280/Sample.JPG

[Z]

Originally posted by swong46:
Z, is this what you are talking about basically?
Swong,

Not quite. Your teeth have a rectangular profile, so as you tighten the joint the sides of the teeth remain loose, and only their tops make contact.

Redraw the teeth with a triangular profile ("A" shaped). Also make them bigger so easier to see. Now when the two parts touch, the contact is at the sides of each tooth (ie. at "/" and "\"). So 3 teeth with 2 contacts each = 6 contact zones, for complete constraint of 6 x DoF.

I'll try to figure out how to post pictures, but that might take some time. Not much hardware here for getting the pictures into this interweb box.

Is that "google-whatever" place you use free? Is it easy to use (ie. not too many hurdles)?

Edit: I went to googleusercontent and "no such place". Can someone please suggest a good image hosting site?

Z

[swong46]

Gotcha, I understand now. As for a photo hosting thing, use Picasa. It is included with all gmail accounts.

Once you log in, click photos on the very top

[Big Bird]

Do you all mind if I ask a really dumb question?

Why?

What is the purpose of a centrelock wheel in FSAE? All the professional examples I see above are for racing series where pitstop time is critical. Since when has wheel change been a time-critical issue for us?

I can quote an example where a centrelock wheel has cost a podium finish, and that was our mob in Detroit 2005. Centrelock came loose, Endurance over...

I think all the above posts are very interesting and there is some good engineering discussion going on. But I just think it is all about solving a problem that isn't really ours. Can anyone convince me otherwise, I'm certainly willing to listen.

[Z]

Geoff,

Why centre-locks?

I'd say the reason FSAE teams use them is that "they are on all the F1 cars" (RMIT'05??? http://fsae.com/groupee_common/emoticons/icon_wink.gif).

The main thrust of my ranting is that if a team really wants to use them, then please don't do it like F1. Why?

I can quote an example where a centrelock wheel has cost a podium finish, and that was our mob in Detroit 2005. Centrelock came loose, Endurance over...
Geoff, I'm sorry that it was your mob that proves my point! http://fsae.com/groupee_common/emoticons/icon_frown.gif

Much of what I have written above is stuff I have been pushing for years, but with no results. Just felt I may as well get it out here. There was an article in Race Tech a few years ago about "wheels still falling off...". I wrote to the editor and he replied "Yes, I'd like a short article about these conical splines...". I sent the article ... and ... nothing ...

The thing is, a well done centre-lock on a conical spline would be a slightly lighter and stronger connection than the standard multi-bolted flange. And possibly more time consuming for the team, although that depends on their facilities.

It would certainly be educational, given that conical splines are perhaps not well known, and by comparison the normal centre-locks are abominations. Sure, the team wouldn't be solving an important FSAE problem. But solving a problem that is plaguing much of professional motorsport might be fun.

Anyway, BBolze dragged up this old 2007 thread and asked the question, so I hope we've gone some way to answering it for you... http://fsae.com/groupee_common/emoticons/icon_smile.gif

(And regarding bearing preload, BBolze, better do it separately to the wheel fixing.)

Z

[ben]

I tend to agree with Z - lots of stuff in racing is done due to fashion. No one will criticise you for maintaining the status quo.

My only concern Z is that needing a flat faced nut to do the Hirsch nut up means you need a locking nut. A centrelock just uses the friction at the conical interface with a spring loaded clip to physically stop it falling off if it backs off.

What locking mechanism would you use?

Ben

[ben]

Incidentally one of our teams this year has drilled a massive countersink on their drive peg holes to get them to line up quicker.

Ben

[Z]

Originally posted by ben:
A centrelock just uses the friction at the conical interface with a spring loaded clip to physically stop it falling off if it backs off.
Ben, it is the "conical interface" that is the problem. The vertical wheel forces rotate around the joint and, together with the small clearance around the pegs, cause the wheel to move in an "orbital" fashion wrt hub (once friction is overcome). This, together with nut thread clearance, and nut conical interface clearance, acts a bit like a gear inside a slightly larger internal ring gear (eg. a "harmonic drive"), and unscrews the nut. It's a bit more complicated than that, because of LH or RH tread, and where the bigger clearance is, etc., but the end result is the same - lost wheels.

The conical spline positively prevents the "orbital" motion so the nut never has the unscrewing forces applied to it. All those old 1930 racing engines and WWII aero engines had their face splines done up with a simple threaded fastener, torqued appropriately, with no further locking mechanism.

Quote "What locking mechanism would you use?" Just for peace of mind I might put a spring "R" pin after the nut (like pic on earlier pages). Or if using a large diameter hollow bolt (screwed into internally threaded hub) I might somehow lock the bolt-head to the wheel.

But the important point is that no matter what locking mechanisms are currently used on the conventional joints, the orbital fretting forces chew the begeezus out of them! And then the wheel falls off... http://fsae.com/groupee_common/emoticons/icon_frown.gif

Stop the orbital hub/wheel fretting, and you don't have to fret.http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

[swong46]

Why not just use 3 slightly tapered drive pins? In my mind this would be just as effective.

[Z]

Swong, you could not take the wheel off. Think 3x Morse tapers!

Z

[ben]

http://www.voithturbo.de/hirth...ing_product-data.htm (http://www.voithturbo.de/hirth-coupling_product-data.htm)

http://www.tedisa-hirth.com/in.../serie_standard.html (http://www.tedisa-hirth.com/ingles/serie_standard.html)

[Big Bird]

Originally posted by Z:
Geoff,

Why centre-locks?

I'd say the reason FSAE teams use them is that "they are on all the F1 cars" (RMIT'05??? http://fsae.com/groupee_common/emoticons/icon_wink.gif).

The main thrust of my ranting is that if a team really wants to use them, then please don't do it like F1. Why?

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">I can quote an example where a centrelock wheel has cost a podium finish, and that was our mob in Detroit 2005. Centrelock came loose, Endurance over...

Geoff, I'm sorry that it was your mob that proves my point! http://fsae.com/groupee_common/emoticons/icon_frown.gif

Z </div></BLOCKQUOTE>

I'm pretty sure the reason we went for centrelocks back in 04-06 is that they were "lighter and stiffer". We went back to a three stud design in 07, because it was "lighter and stiffer"...

No prob about using our mob as an example, this is a learning venture and we learn from our mistakes as well as we learn from others.

I generally don't like centrelocks, as the loads from the contact patch all need to be transmitted through the interface of the wheel centre and whatever it is attached to (axle, hub, whatever). The closer to the axle centreline, the higher the contact forces. In the case of a centrelock, we are taking tyre forces from a large diameter, transferring them s moments to a small diameter axle (and therefore high contact forces) and then feeding the forces into the uprights via bearings. I just think the larger the diameter where wheel loads are transferred into the hub, the lower those forces and therefore the less material required.
We need a draw function on these forums so I can explain myself, and I couldn't be bothered drawing something and scanning it and then hosting it somewhere.
Thats enough.
I'll shut up now.

[Z]

Originally posted by Big Bird:
We need a draw function on these forums so I can explain myself, and I couldn't be bothered drawing something and scanning it and then hosting it somewhere.
Geoff, absolutely 100% agreed!

And I've already got the pictures drawn, AND in this box (from that draft article I wrote a while ago)! Well, so far I've jumped over most of the hurdles to upload these pics ... mumble, mumble ... http://fsae.com/groupee_common/emoticons/icon_mad.gif

Anyway, the link in Ben's post above has quite a few good pics of Hirth joints. As shown, they can also be used to strengthen multi-bolted joints, like your three stud wheel fixing.

Z

[ben]

Originally posted by Big Bird:
Do you all mind if I ask a really dumb question?

Why?

What is the purpose of a centrelock wheel in FSAE? All the professional examples I see above are for racing series where pitstop time is critical. Since when has wheel change been a time-critical issue for us?

I can quote an example where a centrelock wheel has cost a podium finish, and that was our mob in Detroit 2005. Centrelock came loose, Endurance over...

I think all the above posts are very interesting and there is some good engineering discussion going on. But I just think it is all about solving a problem that isn't really ours. Can anyone convince me otherwise, I'm certainly willing to listen.

I think you're right to a degree Geoff, but we're back to what the competition's about and how to define it.

In reality the "weekend autocrosser" probably wants his car to be "like an IndyCar" or "Like F1" So on that basis centre-lock wheels have a sound marketing justification.

I think they can also be useful if the weather is changeable and you need to swap tyres quickly, although admittedly that is rare.

Ben

[mech5496]

Originally posted by Big Bird:
Do you all mind if I ask a really dumb question?

Why?

What is the purpose of a centrelock wheel in FSAE? All the professional examples I see above are for racing series where pitstop time is critical. Since when has wheel change been a time-critical issue for us?

I can quote an example where a centrelock wheel has cost a podium finish, and that was our mob in Detroit 2005. Centrelock came loose, Endurance over...

I think all the above posts are very interesting and there is some good engineering discussion going on. But I just think it is all about solving a problem that isn't really ours. Can anyone convince me otherwise, I'm certainly willing to listen.

Why? Well.... FSG 2011, late afternoon, minutes before the Autocross track closes, no attempt yet. We are patiently waiting in our line with slicks on the car and fingers crossed not to rain. Then, an FSAE best-friend comes to visit: Mr. Murphy with his laws, and as you already might guess it started pouring cats n dogs. Now consider a team with a 4-lug wheel mount system and one with centerlocks changing from slicks to wets...Suddenly speed becomes very very important!

Z, I have proposed something similar for our wheels a couple of years ago, but we decided to yse a conventional peg-drive system in sake of simplicity and reliability, as we have already used those for a few years...I admit it sounds tempting now that I hear it again, we only have to test some on our old car and see how it turns out!

[Z]

Trying to post some drawings... grrrrr ... http://fsae.com/groupee_common/emoticons/icon_mad.gif

https://lh3.googleusercontent.com/-OxoQkxGAJTg/TsdHPH_zomI/AAAAAAAAADc/97BEpyJYCes/s800/SplineFg1.jpg

https://lh4.googleusercontent.com/-hZ9rKLbKN-o/TsdHnL8gf8I/AAAAAAAAADg/lPvxJYUkM7o/s800/SplineFg2.jpg

Got it!!!!!!!!!! http://fsae.com/groupee_common/emoticons/icon_smile.gif

And now I've got to have a drink......

Z

[Z]

Some comments regarding the above sketches.

Figure 1b. "Conical Spline", is roughly what I would use for a "centre-lock" wheel attachment. But perhaps with more and finer teeth (fewer teeth = easier to draw http://fsae.com/groupee_common/emoticons/icon_wink.gif).

At bottom right of Figure 2 is a sketch of an octahedral spaceframe that represents the six n-lines of the three-tooth face-spline. This spaceframe is a fully triangulated rigid structure. The support structures of "simulator motion platforms" are often this shape, just with 6 linear actuators (eg. hyd. rams) where the 6 n-lines are.

(Edit) The "three hacksaw cuts" required to make a "three-tooth face-spline" are shown in the main part of Figure 2. Each cut forms a flank on two different teeth, the two flanks thus being coplanar. This type of joint can be easily hand-made from round bar or thick-wall tube. It is useful if you want a quickly removeable bracket (ie. one bolt) that is sturdy and relocates to the same place every time. (End Edit)

Z

[Big Bird]

Why? Well.... FSG 2011, late afternoon, minutes before the Autocross track closes, no attempt yet. We are patiently waiting in our line with slicks on the car and fingers crossed not to rain. Then, an FSAE best-friend comes to visit: Mr. Murphy with his laws, and as you already might guess it started pouring cats n dogs. Now consider a team with a 4-lug wheel mount system and one with centerlocks changing from slicks to wets...Suddenly speed becomes very very important!

OK, gunna play devils advocate here, and maybe be a little provocative.

The question I'd be asking in the above scenario would be "why the hell were we sitting in the AutoX queue minutes before close without a time on the board". I think the answer to that question would bear a lot more fruit than what technical fix we can come up with in case it happens.

I'd expect the above scenario would have come up due to:
- Failing one or more of scrutineering / noise / brake / tilt test
- Having an underprepared and maybe untested vehicle that wasn't working reliably
- Some form of team / organizational breakdown due to overstressed and underslept team members.

So putting my systems thinking cap on (because the noise inside my brain tells me to keep linking everything to everything else), and referring to my guiding principle that "everything that happens at the end of a project is a direct result of a decision made at the start", I'll provoke some argument by saying:
Maybe the decision to use centrelock wheels was actually a contributing factor to the team being stuck in that autocross queue in the first place

The crux of my argument is that we decide what is important enough to allocate our limited resources to. Designing and implementing centrelock wheels is something that costs an amount of time and money, (and I'd argue more time and money than hubs & studs), and both of these resources could be used for other purposes. Add this to maybe a few other "niceties" that are above and beyond the most basic brown go-kart we can get away with, and maybe we might end up with a bigger task than resources will allow. If the team failed to make it to the autocross queue in time then I'd say that the core of the problem was that the team was in fact overcommitted, and did not manage time, people and money well enough.

I understand that the intent of the competition is all about learning. Often this is interpreted as learning how to design components, or to do FEA/CFD etc, or how to make stuff. If you listen to some of the senior organizers though you'll hear them speak quite a bit about students learning how to manage and deliver a project, how to use design process, and how to think holistically and systemically. So apologies for derailing the thread, but I'm just waving the systems thinking banner here. I'll stop now http://fsae.com/groupee_common/emoticons/icon_smile.gif

[custerjw]

I've been lurking around the forum for a while, I had a thought about Z's proposed solution (face/conical splines) that made me want to join the discussion. I had an issue that would make me a bit hesitant to use that system. When you're putting the wheels on the hubs with those splines, you torque up the whole unit to get the appropriate preload on the splines, but at that point you're starting to load the threads of the lug into shear. My thought is though, when you start applying loads to the hub, that force of the face splines pressing against each other would create a normal force that would try to force the joint apart.

That force will be reacted by the lug holding the whole unit together, but the repeated loading cycle will put a fatigue stress on the threads. Z, I'm curious what your thoughts are on those forces acting on the threads, and if there's a way around the issue (or how big of a deal you think it is). I guess you could spec out the threads in order to handle a ton of force, but that doesn't seem to really solve the issue, merely just a brute force way to delay the issue.

I really appreciate everyone's insight across the entire forum. I hope I can bring something valuable eventually as well.

John Custer
Bearcat Motorsports 09-13

[Adambomb]

Originally posted by custerjw:
That force will be reacted by the lug holding the whole unit together, but the repeated loading cycle will put a fatigue stress on the threads. Z, I'm curious what your thoughts are on those forces acting on the threads, and if there's a way around the issue (or how big of a deal you think it is). I guess you could spec out the threads in order to handle a ton of force, but that doesn't seem to really solve the issue, merely just a brute force way to delay the issue.


That reminds me a lot of a discussion I had with an individual about overturning moments. Our choice of outer CVs forced us to run a somewhat smaller distance between our wheel bearings, and this individual was just convinced that "the whole thing would blow up" under cornering. I told him that we had taken the cornering moment into consideration, the wheel bearings were adequate, and the retainers were designed for the loading. He was still convinced that it would "just blow up," although he could never give a reason why other than "it just will." It was fine. Believing that your engineering data is correct takes a leap of faith in some respects, but science doesn't lie! Although if your assumptions or calculations are wrong...

So in other words, it's simple...you just take that into account as you have described. "FBD that $hit," and from there it's just engineering! Besides, for simple packaging you would likely want a fairly large retaining nut anyway, so probably not a big issue.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Geoff, as a student of your other posts, I hear what you're saying! One thing that this discussion has taught me was that I'm sure glad we didn't switch to centerlocks like many people on our team over the years have talked about. It is 90% for "style points" IMO, as I concur that for FSAE it would be a rare set of circumstances indeed where centerlocks could give you a competitive advantage, especially in terms of reliability. Although this is a very enlightening discussion, especially if you look at FSAE as a "breeding grounds" for other levels of motorsports. Besides, I feel free to contribute to this forum because I feel that knowing is only half the battle, it's the application that wins or loses! Nothing wrong with adding a bit of entropy to the collective mindset http://fsae.com/groupee_common/emoticons/icon_wink.gif


Originally posted by Z:

Because contrary to the views of those in motorsport, it is actually a minority pastime, peopled mostly by egotistical, brain-dead buffoons, who do NOT have to achieve any level of success, such as winning, to stay indefinitely in the sport, so instead they spend most of their time looking for excuses for why they are losing.

A typical excuse is "Awww geez..., we can't afford the $30k wrench, so our wheels keep falling off...".

In my short number of years in motorsports I'd say this is a completely accurate description. More often than not, decisions are made based on style (what is XXX series doing), whose idea it was (that guy was an a-hole, therefore his design/calculations were wrong), or any number of highly subjective egotistical reasons that haven't got a damned thing to do with competitive advantage. Most people in motorsports aren't humble enough to be truly objective and open minded, perhaps even more so in FSAE. I'd surmise that the "standard method" of doing centerlocks persists because it works "well enough" to most peoples' expectations, is easy enough to manufacture, and easy enough to understand. But 10 years from now they could be seen as being completely antiquated if someone finds a better way to do it that is simple enough to execute and understand, and enough of an improvement (and a visible improvement for that matter) that it catches on.

[Z]

Originally posted by Big Bird:
OK, gunna play devils advocate here, and maybe be a little provocative.
Hey, that's my job!!!
(Geoff, I thought you were supposed to be "good cop". http://fsae.com/groupee_common/emoticons/icon_smile.gif)

Regarding the holistic approach, I am in complete agreement. My brown go-kart would probably have off-the-shelf production car hubs. Preferably oversized with taper-rollers, then machine as much unnecessary steel off them as is feasible. Then lots of extreme testing... If they break, buy another and machine less.
~~~~~o0o~~~~~

John,

When you fit a cylinder head, the head-bolts/studs stretch (quite a lot) and develop a preload. Every time the gases explode they push the roof of the chamber up a bit, but the bolts/studs hardly stretch any more. The roof movement is absorbed in the depth of the head. The fatigue of the bolts is purely a function of any stretching/relaxing they do. If no (or minimal) change in length, then no fatigue.

Repeating what I wrote earlier, Pratt & Whitney (along with many others) found that the Hirth joint on their highly loaded cranks was far better than the previous axial spline. Do it right and it works great.
~~~~~o0o~~~~~

Adambomb,

Well, I probably was a bit OTT with that comment... http://fsae.com/groupee_common/emoticons/icon_smile.gif

I am convinced, however, that if motorsport was a bit more like war, then the cars would be much, much better (faster, more efficient, lower fuel use, you pick). Ok, maybe having heads LITERALLY on the chopping block would not be nice. But some kind of "you're a regular loser, so you're out!" would make it much more interesting.

BUT!!! The problem is, most motorsport, even at the very top, is sort of self-governed. So no team is going to sack themselves, no matter how incompetent they are. So we end up with a system where massive $$$s are spent each year with absolutely nothing to show for it (except that a small number of people get to go to the circus every weekend).

If the rules encouraged a direction towards socially beneficial developments, and a few heads were regularly put on the metaphorical chopping block, then we'd all be flying around in fully sustainable sky-cars... http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

[Big Bird]

Aah Z, you have brought out the worst in me. You've set an example and I can't help but follow - however meek my attempts at provocation may be.

But the main point is - THIS IS FUN!!! Its just like old times. I come to the forums these days and I actually learn stuff.

So with our renewed acquaintance with informed debate, maybe we should set our collective minds to the big question facing us all:
What is formula for optimal chassis? Please respond urgently.
Of course, I should open a new thread for that one...

Sorry, back to centrelocks. Now where were we?

[mech5496]

[Offtopic mode on]
Geoff, no offence but as you say there's no such thing as "best solution", it's a matter of design choices, priorities and resources. We designed our hub only with one nut for both retaining the wheel and setting the preload, so centerlock was by far the simplest way to go...(not quite sure if anyone will understand my description as it is pretty stupid but I will try to upload photos at some point). BTW we were stuck there because failing noise test in the first place...
/[offtopic mode off]

[custerjw]

Adambomb:

Although if your assumptions or calculations are wrong...
Bingo. After consulting with our resident grad student, I realized I hadn't been considering the whole system correctly. Mostly wasn't thinking about the friction force along the face correctly. As Z responded, executed correctly, it should work great.


When you fit a cylinder head, the head-bolts/studs stretch (quite a lot) and develop a preload. Every time the gases explode they push the roof of the chamber up a bit, but the bolts/studs hardly stretch any more. The roof movement is absorbed in the depth of the head. The fatigue of the bolts is purely a function of any stretching/relaxing they do. If no (or minimal) change in length, then no fatigue.

Okay, that makes sense. Practical applications are a great thing.

This whole internet thing's pretty amazing... Ask a question, and within a day, have responses from halfway across the country, and halfway across the world. Thanks guys.

John Custer
Bearcat Motorsports 09-13

[Z]

Originally posted by Big Bird:
...the big question facing us all:
What is formula for optimal chassis? Please respond urgently.
Geoff, I hinted at the answer in my previous post, namely;

Formula-for-Optimal-Chassis = (Student's-Neck + Chopping-Block)**Lose-and-Axe-Comes-Down! x ASAP. http://fsae.com/groupee_common/emoticons/icon_smile.gif

I was going to expand on that theme on your "Reasoning..." thread, but I just left it with the one Sun Tzu quote (the kiddies might be a bit sensitive these days).

What I'm getting at is that motorsport, and FSAE, has a lot of "carrot", ie. the teams are rewarded if they win (they get a cup, and fame...). But there is very little "stick", so the perennial losers still have a lot of fun. With no "cost" associated with performing poorly, it is no surprise that so many teams do perform poorly. Hey, they still get to go to the circus at the end of the year...

I think the "optimal" chassis for a newbie team is the one they would build if they were told by their school; "If you fail to score reasonable points in EVERY SINGLE EVENT, and ESPECIALLY ENDURANCE (!!!), then your rotting carcasses will be nailed to the walls of the school as positive encouragement to next year's team. An Endurance winning team is an optimally happy team. Be happy. Optimally..."

With such an optimistic and positive mission statement/formula, I think most newbie teams might build a brown go-kart and do quite well. http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

Of course, the above is just a blunter version of what you have been saying all along...

Which leaves the question, how to get them to listen?

Z
(Edit: Spelnig!)

[PatClarke]

Trouble is Erik, for every winner in FSAE, you are always going to have 99 losers...Or perhaps in these dreadfully PC times, 99 'non winners'.

Initial thoughts are the effect of bringing in the tar and feathers and guillotine would ultimately be to dumb down the competition, though a look at history in the first half of the last century showed the absolute cost of losing really progressed the areas of mechanical science we are interested in.

Even for the 'losers' who could probably now be described as the winners!

Pat

[Z]

Originally posted by PatClarke:
... though a look at history in the first half of the last century showed the absolute cost of losing really progressed the areas of mechanical science we are interested in.
Yes, Pat, that is what I was getting at.

And although we both missed the "big one", we also both witnessed the technological progress that was a direct result of the cold war, when there was only the threat of nastiness. For example, man walking on the moon! I found that quite interesting... http://fsae.com/groupee_common/emoticons/icon_smile.gif

By comparison, I note that Ben Bowlby's (spelling?) recent and rather interesting "Delta Wing" Indycar concept was all but ignored by the organisers. To me the really interesting thing is that it looks remarkably like the Lockheed SR-71 Blackbird of the 1960's, which is still the fastest plane ever built! (Of course, the Ruskies might argue that point, and claim that the MiG-25 Foxbat was just as fast. Both were back-of-the envelope sketches in the late 1950s, and operational in 1962! Ah, the good old days...http://fsae.com/groupee_common/emoticons/icon_smile.gif)

But now we're off-topic...

Z

[Mike Cook]

I don't know, maybe I'm naive, but a few things come to mind.

First, Z, how long have the been using centerlocks in racing? 30, 40 years? Sure, over the millions of tire changes a few wheels have fallen off. Does this mean the design is flawed? Or is it more likely that the mechanic who put the wheel on messed up? Because it seems to me if the design was fundamentally flawed a lot more wheels would fall off.

The face spline seems like a good idea. In fact I have a flange sitting on my work bench right now that utilizes a face spline. Cool! I just think that maybe if this design was put on every race car in the NEXT 40 years we would still see wheels fall off.

The face spline seems more difficult to manufacture in my opinion (although I guess this could be debatable) and I think a centerlock with proper preload will not move.

Mike

[Z]

Originally posted by Mike Cook:
... is it more likely that the mechanic who put the wheel on messed up? Because it seems to me if the design was fundamentally flawed a lot more wheels would fall off.
Mike, the mechanic is using a "$30k wrench" and still at least once per year, in the very TOP formulae (with presumably the best mechanics), does a wheel fall off. In the lower formula, and amateur racing, it is extremely common.

Here's one way to see what an abomination the joint is;

Take two lengths of flat bar. Drill a hole in each, anywhere along its length. Now get someone to bolt them together "as tight as possible", with, of course, a single nut & bolt. Then bet the other person you can "undo the joint with your bare hands".

Ok, you have to make sure the bars are long enough, or the bolt small enough, but usually it is a simple matter of a push and a pull in the right directions, or maybe a whack on the concrete floor, to loosen the joint. With practice you can make it look like a magic trick - "abracadabra" and it's undone!

Note that if the bars are very rough and rusty, then friction is greater between them, and the joint is harder to undo. Fine face splines are bit like an orderly version of the rough rust. Smooth, shiny flat bars are very easy to undo, like a smooth, shiny hub on a top spec race car.

Z

[Mike Cook]

With enough leverage you can move the world.

[Canuck Racing]

...and now add a few more bolts (sans nuts on the far side) to act as drive pins such that you can only wiggle the bars a few mm at each end. Try and untorque the first bolt.

Also, in regards to the motorsports industry, yes everyone has an ego problem. But they all have the desire to win. Unlike in FSAE, too many losses costs them thier job, and it's not so easy to find a new one when you lost your last one for underpreforming. So any comments about people being reluctant to change in order to go faster (even just a tenth of a second) are completly illogical and un-based. Otherwise Audi would still be running the R8.

Oh, and if ten years from now everybody switchs to splines, please feel free to pull this up and rub it in my face. It will help keep us egotistical motorsports professionals in check.

[ben]

Originally posted by Canuck Racing:
Otherwise Audi would still be running the R8.


Bad example.

The rules were framed at Le Mans to guarantee a performance advantage for diesels so there was zero risk in doing it, provided the car didn't break. Not breaking is easier to achieve if you don't need to be at the limit of the technology to go quicker than your opponents.

Ben

[Canuck Racing]

Originally posted by ben:

Bad example.

The rules were framed at Le Mans to guarantee a performance advantage for diesels so there was zero risk in doing it, provided the car didn't break. Not breaking is easier to achieve if you don't need to be at the limit of the technology to go quicker than your opponents.

Ben

How? They changed in order to win right? They didn't just accept that the R8 worked and let it stagnate into obsolescence, such as Z is claiming hub design has.

[ben]

They changed in order to realise a marketing goal. Nothing wrong with that, but a bad example if we're talking purely about the superiority of concepts from a pure engineering point of view which we have been.

Ben

[Z]

Originally posted by Canuck Racing:
... in regards to the motorsports industry ... too many losses costs them thier job, and it's not so easy to find a new one when you lost your last one for underpreforming.
Canuck,

Maybe I was a bit harsh with the "buffoons" comment, but what I have been saying in my recent comments directly contradicts your above quote. People in racing DO NOT lose their jobs for underperforming.

Honestly, I've spent 40 years watching the same clowns going around, team to team, always climbing the ladder thanks to their increasing "experience", until they are the tech director of a team that will never, ever win!

Consider some examples of teams in F1. Both Lotus and Brabham were hugely successful, then suddenly they were gone. On the other hand, both Minardi and Arrows were hugely UNsuccessful, never looking like they would get off the back of the grid, yet they hung around for decades.

To be "successful" in motorsport you just have to keep your sponsors happy. There is always a marketplace for the less wealthy sponsors, namely at the back of the grid. These sponsors get to pay less money, but they still regularly go to the circus, get pit passes, rub shoulders with the "bold and the beautiful", etc., etc. But teams that concentrate too much on winning, and don't chase sponsorship hard enough, can disappear overnight.

It is similar with the individuals. Those that cover their arses, white-ant their colleagues, and grovel to their bosses, stay around for decades. "Gee, boss, if that wanker in marketing worked harder we could afford that $30k wrench, then our wheels wouldn't fall off, and with your brilliant designs we would start winning...".

Except they never win...

Z

PS. Even if you put some slightly loose drive pins in those flat bars, I bet I could undo the bolt. http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Mike Cook]

Originally posted by Z:


Maybe I was a bit harsh with the "buffoons" comment, but what I have been saying in my recent comments directly contradicts your above quote. People in racing DO NOT lose their jobs for underperforming.



I don't know Z, this sounds like the majority of the jobs out there. I still don't understand it, but generally the shit rises to the top.

[Nicky]

Dunno if this link was posted earlier, atleast the first 4 pages didn't have it. Do excuse me if it was already posted.

http://www.youtube.com/watch?v...oSNM&feature=related (http://www.youtube.com/watch?v=PeIqXgaoSNM&feature=related)
The first link shows the locking mechanism on the nut itself. It's very clearly shown.(1min, 24sec)

[ben]

Happy Z presumably:

http://sphotos-e.ak.fbcdn.net/hphotos-ak-prn1/534818_409300165756695_2113586278_n.jpg

Ben

[Racer-X]

Originally posted by ben:
Happy Z presumably:

http://sphotos-e.ak.fbcdn.net/hphotos-ak-prn1/534818_409300165756695_2113586278_n.jpg

Ben

Those are neat. I wonder how long the splines/teeth will last on there though.

[Drew Price]

Ooooohhhhh... I get it. That's pretty clever.

[Markus]

You can achieve the same with an "easier" (at least from the manufacturing point of view) solution too: Who says you need full face-spline? http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Z]

Ben,

Oooo... yes! Z happy now! http://fsae.com/groupee_common/emoticons/icon_smile.gif http://fsae.com/groupee_common/emoticons/icon_smile.gif http://fsae.com/groupee_common/emoticons/icon_smile.gif

Do you know who, where... ?
~~~o0o~~~

Racer-X,

I'm sure the teeth would last a lot longer than a conventional peg-drive's mounting face and holes, which are subject to "fretting".
~~~o0o~~~

Just for the record...

The image shows one half of a "Hirth joint", aka a "face spline", or "conical spline with half-cone-angle of 90 degrees". As such, it is a hermaphroditic joint where both sides are the same (boy=girl). So the "axle-hub" to which this mounts would look similar.

This particular one has 20 teeth that seem to have a flank angle of about 45 degrees (inluded angle of teeth = 90 degrees?), possibly for ease of machining with conventional milling bit. Narrower teeth are also common, with included angle down to 30 degrees (or less).

If I was making that joint, then I might cut a little more off the teeth peaks, and/or undercut the valleys a bit (with round nose mill).
Any peak-to-valley contact = bad!
Only flank-to-flank contact = good. http://fsae.com/groupee_common/emoticons/icon_smile.gif

The centre-lock nut, which presses the two halves of the joint together, should have a flat mounting surface (ie. at 90 degrees to axle axis), NOT the usual conical surface. This appears to be the case as seen at top-left of image, but not entirely clear?

As a final BTW, I recently saw a similar joint on some interweb image of an electric motor for someone's hybrid car.

It's an oldie, but a goodie! http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

[Z]

Originally posted by Markus:
... Who says you need full face-spline? http://fsae.com/groupee_common/emoticons/icon_wink.gif
Markus,

Hmmm, well, you need at least 3 teeth, each with two contacting flanks (for control of all 6 DoFs) ...

Z

[ben]

Originally posted by Z:
Ben,

Oooo... yes! Z happy now! http://fsae.com/groupee_common/emoticons/icon_smile.gif http://fsae.com/groupee_common/emoticons/icon_smile.gif http://fsae.com/groupee_common/emoticons/icon_smile.gif

Do you know who, where... ?


Z

University of Auckland: http://www.facebook.com/UoAfsae

Nuts look flat-faced too:

http://sphotos-f.ak.fbcdn.net/hphotos-ak-prn1/527029_449709078382470_2103869950_n.jpg

Ben

[Markus]

Originally posted by Z:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Markus:
... Who says you need full face-spline? http://fsae.com/groupee_common/emoticons/icon_wink.gif
Markus,

Hmmm, well, you need at least 3 teeth, each with two contacting flanks (for control of all 6 DoFs) ...

Z </div></BLOCKQUOTE>
And 3 is what we have. And a conical washer with a rotational dof for centering.

I'm not going to share more about the design because it's not my own. But it's proven reliable since 2009 with zero count on lost wheels due to centerlock loosening / failure.