Bearing Design Life

[Dash]

Just curious what some other people pick as a design lifetime for their wheel bearings. I was thinking something around maybe 500 hours ( I know that is quite a long time for a racecar ). The only problem that this is giving me is that because our car is relatively heavy ( 530lbs + ) the loads that I am calculating are leading me to some rather large and heavy bearings, about 45mm ID taper roller bearings. Even when I half that life to approx 250 hours, it doesn't change the load rating all that much. Just curious if anyone would mind giving me a sanity check on this solution. I'm still in the iterative design stage, so I'm hoping this number will whittle down as I get better defined dimensions on other parts.

[Dash]

Just curious what some other people pick as a design lifetime for their wheel bearings. I was thinking something around maybe 500 hours ( I know that is quite a long time for a racecar ). The only problem that this is giving me is that because our car is relatively heavy ( 530lbs + ) the loads that I am calculating are leading me to some rather large and heavy bearings, about 45mm ID taper roller bearings. Even when I half that life to approx 250 hours, it doesn't change the load rating all that much. Just curious if anyone would mind giving me a sanity check on this solution. I'm still in the iterative design stage, so I'm hoping this number will whittle down as I get better defined dimensions on other parts.

[Hector]

45 mm isn't that big. I've seen a lot of teams run a pair of deep groove ball bearings on each corner. They had a much bigger diameter than you're posting.

Also, 250 hours?!? most teams are very lucky to get over 100 hours of loaded (AKA driving balls out, not just idling) driving on one of these cars. I'd design for much lower life and plan on replacing the bearings if you exceed that life. Remember, this is a race car. Long lasting is unfortunately very heavy.

[Dash]

I was just considering the fact if you got to run like two weeks of testing for a few hours a day, that would easily use up the bearings if you set the lifetime too low.

[Pete Marsh]

Be sure to do your life calcs at the correct speed, which is probably much lower than the nominal speed rating in the bearing book.

Also rate your loads duration for what the car experiences. Surely it isn't at maximum design load for every second of that 500 hours? That should bring you back closer to what you will commonly see on other cars.

I agree 45mm tapered rollers are too big, but ball bearings need to be much bigger than rollers to carry the same load.

Check out, and re calc, some OEM wheel bearing selections for a sanity check of how small they can be. Might make you think twice about those 20" wheels on your old classic though!

Pete

[Sam. B.]

What about the rigidity of the bearings, the hub and the assembly? If your calculations gives you a bearing of 10 mm ID for a 250 hours lifetime will you go with a 10 mm bearing? I would say no because I don't want a hub flexible like a cheese. Tel me if I am wrong but I think this whole bearing selection thing is a compromise between rigidity and mass.

By the way, what kind of hub rigidity are you guys usually targeting? We often see team targeting a torsion rigidity for the frame of let's say 3000N*m/deg. But what about the rigidity of the hub and upright assembly? Some teams have numbers? I think it's all based on the tire data and a change in toe of 0.5° due to flexibility will probably have a big impact on the performance of the car, but I don't have any references to confirm or invalidate what I think. What do you guys think about that?

Sam

[Simon Dingle]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">By the way, what kind of hub rigidity are you guys usually targeting? We often see team targeting a torsion rigidity for the frame of let's say 3000N*m/deg. But what about the rigidity of the hub and upright assembly? Some teams have numbers? I think it's all based on the tire data and a change in toe of 0.5° due to flexibility will probably have a big impact on the performance of the car, but I don't have any references to confirm or invalidate what I think. What do you guys think about that? </div></BLOCKQUOTE>

Back when I was doing this we had targets for stiffness in terms of camber change and toe change per g of acceleration. I suspect that most teams have targets for suspension stiffness as well as chassis stiffness, there's certainly no point in having a stiff chassis if your suspension assembly has all the stiffness of over-cooked pasta. Unfortunately I can't remember what sort of values we had, but I would expect a team's suspension lead to be able to give ball-park figures for how much change in camber or toe is acceptable and then base your suspension stiffness values on that.

[Dash]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Sam. B.:
What about the rigidity of the bearings, the hub and the assembly? If your calculations gives you a bearing of 10 mm ID for a 250 hours lifetime will you go with a 10 mm bearing? I would say no because I don't want a hub flexible like a cheese. Tel me if I am wrong but I think this whole bearing selection thing is a compromise between rigidity and mass.
Sam </div></BLOCKQUOTE>

I am also doing the spindle design. Basically I am just looking for a starting point on that. I'll find the smallest bearings that will take the load, do some calculations for the spindle to determine the smallest acceptable OD, and then make a bearing decision based on that. If the smallest acceptable OD is bigger than the calculated ID needed, then I have to increase the bearing size. Or the other way around could be true, and I would need to increase the spindle size to match the bearing. All the while I will be checking for compliance in the system.

[Sam. B.]

Nice input Simon!

The idea of (° of camber or toe/g) is interesting because it's possible to make some comparison between car of different weight. If you ever find or remember the sort of values you where using, please share it. Anyway, I'll base my design on the tire data and will try to find some literature. If you know some good references about the subject, share that.

Sam

[Dash]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Pete Marsh:
Be sure to do your life calcs at the correct speed, which is probably much lower than the nominal speed rating in the bearing book.

Also rate your loads duration for what the car experiences. Surely it isn't at maximum design load for every second of that 500 hours? That should bring you back closer to what you will commonly see on other cars.

I agree 45mm tapered rollers are too big, but ball bearings need to be much bigger than rollers to carry the same load.

Check out, and re calc, some OEM wheel bearing selections for a sanity check of how small they can be. Might make you think twice about those 20" wheels on your old classic though!

Pete </div></BLOCKQUOTE>


Double Posting here, but just wanted to comment on this one as well.

As per Shigley's Mechanical Engineering Design book, the life is rated in revolutions. I'm assuming an average of 30 mph for the lifetime that I am choosing.

I'll look more into the variable loading conditions calculations. As of yesterday Afternoon, I found two or three bearings that were approximately 35mm ID that hit the criteria that I have. I'm assuming that much smaller than that, and the spindle will start to flex more than I want.

[Nick Renold]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Sam. B.:
Nice input Simon!

The idea of (° of camber or toe/g) is interesting because it's possible to make some comparison between car of different weight. If you ever find or remember the sort of values you where using, please share it. Anyway, I'll base my design on the tire data and will try to find some literature. If you know some good references about the subject, share that.

Sam </div></BLOCKQUOTE>

RCVD has a few good pages on the topic, what you want is a compliance budget. On our car, we designed for a compliance budget of 0.1 deg/g in camber and a smaller toe compliance which I do not remember. This worked out to 0.03 deg/g given for each of a-arms, upright, and hub. The wheel was not considered, but for completeness this would be a good idea. Considering the designs I have seen for other teams' uprights and hubs, I would expect them to have similar ballpark figures.

[rrobb]

I would say the primary factors, in order of importance, are:
1) Axial deflection
2) Axial load
3) packaging
4) weight

I think any bearing that meets # 1) would be more than sufficient in all other load criteria, and have plenty of life.

Obviously, a quick sanity check calc should be preformed so your car doesn't fall apart after 10 minutes, but don't spend 4 million hours on it.