Wheel Bearing Help

[erau_paul]

I have reading various sources on bearings (this forum, Bosch Automotive Handbook, Machinery Handbook, this forum) but non have really satisfied me in understanding wheel bearings.

I was hoping to use ceramic angular contact bearings because of their lower weight and even more importantly lower coefficient of friction. However I cannot seem to find them in a more desirable 40 degree contact angle... most are 15 / 25 degrees.

Also when I am solving my free body diagram to choose my bearings loads I am confused about weather the bearing experiences (or requires) a axial preload.

My advisor told me that he thinks I can ignore the forces from suspension geometry (for example, jacking forces), aligning torque (on the rear wheels), and overturning moment. Then I should just apply a fairly large factor of safety (2-3). I was going to calculate my loads on the bearings by applying all of the weight of the car to a single wheel, giving 1.5G lateral and 1.5G longitudinal forces (for the entire car) and then coming up with a number from there.

I thought I was beginning to understand everything and then I dove into this thread (http://fsae.com/eve/forums/a/tpc/f/125607348/m/31510472931?r=74510823931#74510823931). It only served to confuse me. Are teams using the double row angular contact bearings?

Thanks guys!

[exFSAE]

So.. you want to use ceramic for some marginal savings in weight and friction.. but you're willing to make a load assumption of 100% of the car weight on a single wheel and a 200-300% overload margin to boot? You will probably have much more weight savings with a realistic worst case load scenario... if you pick a condition that's the worst physically possible then a factor of safety of 1.2 is sufficiently high (and has worked for me just fine, even as a 'rookie' design engineer in college). 20% overload margin is huge.

Also, why are you keen on a 40 degree contact angle as opposed to 15 or 25? What is that contact angle there for? More load capacity in one direction than another, no? Do you expect very high axial loads, or radial loads? (Hint: the radial loads will probably be way way higher than axial and that should point you toward some types of bearings more than others).

With regard to double row angular contact ball bearings, they certainly work. If they can handle Formula Ford loads, they sure can handle FSAE loads. Bit heavy though.

[Silente]

I don't know exatcly about preload.

But anyway to choose bearings we normally calculate how much load tire experiences at contact patch in a certain situation (cornering, braking, both) and with a certain acceleration. Then you can choose how much safety you want on your loads with a safety factor, but honestly to think to have all your car weight on just on wheel seems to be too much.

Then, knowing basic things like tire radius, the offset between bearing and wheel center and using, for example, SKF formulas (they give them in their site) you can calculate if your bearing C and C0 can be ok. As you probably know it's a life calculation, you will have a life duration for a certain average angular speed.

We normally use double row angular bearing, they are not very light but they work and they are quite cheap. but not ceramic stuff. I think they are quite expensive too...

[erau_paul]

Thanks for the input guys.

I guess one of my primary design goals is a very reliable car. Using ceramic bearings improves weight and rolling resistance without any impact on reliability, whereas going with a smaller bearing does not improve reliability. I understand that this may not be the best approach but my understand of wheel loads is not great... Our team is very inexperienced. We did not compete last year because of the EcoCar project and now I am the only team member with design experience. I am having to take the brunt of the design work and taking the time to carefully choose a wheel bearing really isn't there.

I decided on my factor of safety using these recommendations. http://www.roymech.co.uk/Usefu.../Safety_Factors.html (http://www.roymech.co.uk/Useful_Tables/ARM/Safety_Factors.html)

Is it okay to use 2 single row angular contact bearings that are spaced out more (this would seem more efficient to me than a single double row angular contact bearing because of the greater spacing). exFSAE, I would think that the radial loading would dominate the bearing selection for 2 reasons: 1. The coefficient of friction is greater than 1, so the car can experience higher axial loads than radial. 2. The radial loads are supported by 2 bearings whereas the axial loads are only supported by 1 bearing in each direction.

I appreciate the feedback.

[erau_paul]

If you guys notice in this (http://www.taylor-race.com/pdf/ACFD06.pdf) generic wheel assembly from Taylor Race Engineering, they use a double row angular bearing and the bearing supports all of the clamping force from the wheel nut.

I don't understand on their design why they use a nut to hold the spindle on and then another 'contraption' to hold the wheel on. I also don't quite understand the mechanism that holds the wheel on.

Thanks!

[nowhere fast]

I have used deep groove ball bearings.

If you calculate the load at each bearing (rather than just at the contact patch) you will likely find that the radial loads are significantly higher than the axial loads so I doubt a 40 degree contact angle is suitable.

As mentioned above your factor of safety is excessive. Remember that bearing load calculations are based on the expected life of the bearing; the bearings will not fail if overloaded occasionally (within reason) but their lifespan will be reduced.

[moose]

Originally posted by erau_paul:
If you guys notice in this (http://www.taylor-race.com/pdf/ACFD06.pdf) generic wheel assembly from Taylor Race Engineering, they use a double row angular bearing and the bearing supports all of the clamping force from the wheel nut.

I don't understand on their design why they use a nut to hold the spindle on and then another 'contraption' to hold the wheel on. I also don't quite understand the mechanism that holds the wheel on.

Thanks!

The extra "contraption" from what I understand is for center-lock wheels - giving a back-side taper for them to mount to. Haven't used/seen one, hopefully if Scotty sees this he can comment - otherwise I'd shoot him a quick email, he can probably help out quickly. If you're going without a centerlock, then there is just a standard bolt pattern on the inner flange.

[cjanota]

I helped out with some of these calculations last year. Previously we used a double-row angular contact that was packaged in a single unit and found that they were getting loose in camber fairly shortly.

When we resolved the contact patch forces to the bearings, we found that the largest load (by far) was radially due to the the overturning moment from lateral acceleration. The easiest way to lessen that component of the load is to space out the 2 rows to increase the lever-arm. We initially designed for angular contact bearings but they were special order so we got some deep grooves.

As to axial preload, your best bet is to talk to the bearing manufacturer. SKF said that the angular contacts needed about 20Lbs preload (if my memory is correct) whereas the deep grooves didn't need any preload.

[Analogue]

Figuring the best contact angle is not as simple as converting a moment into a force couple, which of course, would act radially. A back to back bearing arrangement can behave as if the bearings are further apart than they really are, since the forces through the balls act through points outside the bearings.

We probably all agree that the moment loading has the largest impact on FSAE wheel bearings. Moment is RXF, and these are both vector quantities. R is increased as the contact angle is increased, and the component of F that acts normal to R is decreased as the angle increases.

I think you will find that if your bearings are close together, as they would be in a typical double row arrangement, then a larger contact angle will increase moment capacity. As the bearings get further apart, the optimum contact angle decreases. From my rough calculations, a 45 degree angle is optimum for a bearing with balls operating at a radius of D/2, and with the bearings spaced apart distance D. Note that this is starting to become a bit on the wide side for an FSAE application, so high contact angles are not a bad thing. exFSAE, I disagree that 15 or 25 degrees is a step in the right direction.

That was the math part, here's the practical part: don't spend 1000 bucks or more and days or weeks of your engineering effort on ceramic bearings. If your team is young and understaffed, efforts focused on any number of other things will move you up far more places. Overloading a steel bearing will cause noise, and you will probably still finish your race. Overloading a ceramic bearing is catastrophic. I don't think this is a good match for your goal of reliability.

Our team has been using VW Rabbit bearings for years. They are a double row angular contact, sealed, and can be had for 25 bucks each. This particular size is pretty popular on small formula cars. Not much to using them, just press in and clamp the inner races together. They have some shortcomings-- they seem to have a little bit too much clearance when they are not pressed into a cast iron VW upright, probably the same situation cjanota describes. I heard from a team at FSAE West that they pop these apart, grind a few tenths off, and put them back together. I know the FAG ones we have been using cannot be taken apart without damage, but I haven't popped open one of the many other brands. Anybody have experience with tightening these up?

[Nick Renold]

Originally posted by Analogue:
...VW Rabbit bearings...
They have some shortcomings-- they seem to have a little bit too much clearance when they are not pressed into a cast iron VW upright, probably the same situation cjanota describes. I heard from a team at FSAE West that they pop these apart, grind a few tenths off, and put them back together. I know the FAG ones we have been using cannot be taken apart without damage, but I haven't popped open one of the many other brands. Anybody have experience with tightening these up?

We just assembled our hub/bearing/spindle setup from Taylor race and also noticed that the FAG bearings have more slop than we like. I'm thinking about getting a set of these bearings that we can pop apart and grind down the inner race width for a tighter spindle setup.

Does anyone know which brand I should buy to try this with?

Am I also right to assume I should go with unsealed bearings on this?