Running bearings on aluminum

[The AFX Master]

Hi.

I'm planning to switch a bunch of parts of our drivetrain from 4130/4340 to 7075T6 aluminum. What's your experience sitting bearings against aluminum?. did you left a part due to surface fatigue at bearing surfaces?.

Another thing, someone here uses sinusoidal splines in order to attach aluminum with steel torque transmitting parts (hubs afaik), do you have galling problems?, do you know some kind of bibliography in order to know how to calculate these splines?, i'm learning some tribology in two or three books, but the subject seems to be far complicated than a plane-plane contact.
And i doubt seriously about the capacity of FEA to handle these multiple contact problems accurately without an expert behind it.
regards

[Jersey Tom]

What parts are you planning on switching exactly? I'd be real wary of switching from 4340 (which I imagine is case or thru-hardened to Rc50+) to aluminum.

[James Waltman]

Pictures now. Explanations below.
http://dot.etec.wwu.edu/fsae/HostedPics/Viking35/Thumbnail_Galleries/Main_V35_Gallery/images/DSC01532.jpg
http://dot.etec.wwu.edu/fsae/HostedPics/Viking35/Thumbnail_Galleries/Main_V35_Gallery/images/DSC01520.jpg
http://dot.etec.wwu.edu/fsae/images/CV%20joint%20tripod%20with%20hub_JPG.jpg
http://dot.etec.wwu.edu/fsae/v26/image004.jpg

The first three are from the '04 WWU car.
The last one is from about '96 or so (Viking 26)

Maybe Travis can comment on how he did the calculations for them.
I don't think there were any though.
The major diameter of the spline is basically the same size as the shaft.
The minor diameter of the spline is dictated by the taper of the end mill.
This is a critical part – the taper. It allows the parts to fit together very well.
We used a 7? end mill. You program the tool path based on a simple profile and as the end mill steps down it makes the taper. Same thing for the hub side.

They work. They work very well. I can't help you much with calculating any of it though.

[murpia]

Nicely manufactured parts, but how did you deal with correctly preloading the bearings? I assume (maybe I'm wrong?) that the taper is not accurate enough to ensure you always get the same spacing between bearing faces. Also, if you tighten up on shims etc. then the more preload in the bearing the more slop in the spline? Seems to me like the two requirements fight each other, which doesn't happen with a parallel spline.

Regards, Ian

[jaker]

The brake caliper in the last picture - is it a purchased part or one that WWU manufactured? If it was purchased, who makes it?

[The AFX Master]

What parts are you planning on switching exactly? I'd be real wary of switching from 4340 (which I imagine is case or thru-hardened to Rc50+) to aluminum.

Two pieces that carry the differential each side in order to take torque and seal the diff.the left one is connected to the sprocket. each one have a deep groove ball bearing attached to the chassis fixed differential carriers. I've seen on Detroit this year a lot of drivetrains made out of aluminum, so, putting bearings on it certainly don't seems to be a big problem. Our current parts are clearly overbuilt, so, overall 7075 properties are enough to withstand the operation conditions of the part, except for the bearing surfaces that make me to scare

[Chris Allbee]

Just a point of technicality here...are you intending to use the aluminum as a race (bearing surface) or just as a mounting shaft? If using a bearing with its own inner race, such as a deep groove ball bearing, then you shouldn't have a problem pressing the bearing onto your aluminum piece and maintaining reliability. However, if what you are considering is to use the aluminum as the bearing surface for a needle roller then you will have to consider a hard anodize or shrink fitting a race over the aluminum first. I wouldn't recommend using the aluminum as a bearing surface though without taking one of those options.

[The AFX Master]

Aluminum as a receway for needle bearings is discarded. I'm talking about deep groove ball bearings, so i'm talking about how the fitting of the steel over aluminum would damage the last one due to the different elasticity modulus of those materials. What i mean saying bearing surface, is the surface intended to fit a bearing unit, not to roll any needle/roller unit

[drivetrainUW-Platt]

Most people run custom aluminum housings with the torsens, so I dont see the problem you are trying to address.

[Travis Garrison]

On stress analysis, there was some FEA done. All potentially contacting surfaces were defined as surface contacts (not bonded or attached with any kind of bearing constraint) and a tension load applied via a bolted connection constraint. It seemed legitimate to me at the time...still seems that way now. Is there something I'm missing there?

On preload, we didn't exactly pre-load but one set did have a second LARGE nut behind the hub to snug everything up and keep everything in place. The same thing would work for tapered roller bearings.

Galling was not a serious issue on the tapered spline, but like most FSAE cars who knows what it would do over 50,000 miles.

They do make pre-load more akward (requiring a seperate nut) but they are much easier to produce than a typical spline. My FEA showed them to be in the ball park of a similiarly sized traditional spline, close enough to work, but nothing that is going to save weight (beyond the ability to actually use aluminum on your drivetrain). The aluminum will wear a little with multiple teardowns, and if your tensioning bolt comes loose, game over.

So those are the pros and cons as I saw them, drop me a line at garrisont -at- gmail.com if you want to discuss FEA. (If you run SW I might be able to give you a cosmos model to start with).

[The AFX Master]

Thanks for your advice Travis. I have almost a year learning FEA and i know to handle contact problems con CW. My doubt was precisely that you already did.. if FEA is a confident method to validate the sline design. But you know, we always need to support by hand calculations all things done by FEA.

[Travis Garrison]

"we always need to support by hand calculations all things done by FEA" very true....but this is one of those "good luck with that" cases http://fsae.com/groupee_common/emoticons/icon_smile.gif Machinery Handbook has equations for a three (or more?) lobe spline that you can use as a reality check in FEA, model that, check agains those equations. If the two agree reasonably well move on the a full sinusoidal spline, that way you'll at least know the contact constraint is working ~as intended.

[Chris Allbee]

For the pressing of the steel on to the aluminum, there are standard equations for calculating the installation stresses on shafts and hubs based on the modulus and geometry of the two pieces. Any good mechanical design book should have them.

[Jersey Tom]

Ohhh. If you're just doing a differential carrier than yea you're fine. And pressing bearings in is no big deal. Usually I like about .0005-.0007" interference fit per inch of shaft diameter.

You'll be usin a LOT of stock to make those things from billet! When we used to do this sort of thing we used a mix of 6061 and 7075, the former used to weld separate parts together so it wasn't such a machining nightmare.

[The AFX Master]

did you weld T6 alloys? http://fsae.com/groupee_common/emoticons/icon_eek.gif.. You have certainly a normalized "O" condition in the welding zone, and the loads that a drivetrain need to carry on.. are way large.

[Mike Flitcraft]

Originally posted by The AFX Master:
Aluminum as a receway for needle bearings is discarded. I'm talking about deep groove ball bearings, so i'm talking about how the fitting of the steel over aluminum would damage the last one due to the different elasticity modulus of those materials. What i mean saying bearing surface, is the surface intended to fit a bearing unit, not to roll any needle/roller unit

Press in a bearing race, and no worries. You also know that the races are matched to the bearings, less to worry about there.

[Jersey Tom]

Yup, we welded -T6. Yes, it anneals the area around the weld, and the 4043 you can't treat it back to strength. I don't recall if they welded it with that or 5356 and if they heat treated it again. I'd wager not. They did some sketchy stuff back in the day.

Regardless, however it was designed, it held up.

[Brian Evans]

6061-T6 typicaly age hardens to a
T4 condition after about a month post welding.

Brian

[murpia]

Originally posted by Travis Garrison:On preload, we didn't exactly pre-load but one set did have a second LARGE nut behind the hub to snug everything up and keep everything in place. The same thing would work for tapered roller bearings.

OK so if I understand you correctly the tapered spline goes tight, then you use a threaded sleeve with a nut to generate the bearing preload? Then you have the same issue don't you, as the threaded sleeve needs to be prleoaded tight around the splined axle?

Maybe you can post another picture please as I don't think I understand this correctly, sorry...

Regards, Ian

[Travis Garrison]

Ian, I can't seem to find a good picture and I'm away from my computer so no CAD for the moment...the basic idea is:

Hub

Splined end of shaft
threaded end of shaft with nut (obviously larger than the spline)
bearings
step/seat on shaft

Make sense? Same shaft but the preload nut doesn't impact tension on the splined interface.

[murpia]

Thanks Travis, makes sense now!

Regards, Ian

[G.Arce]

As we are planing similiar situations of usin needle bearings on aluminum (the needle bearing races around the exterior diameter of an Al tube), how was your experience?

By anodizing we could improve the "galling" of the sliding surface, what do you think?

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

[Owen Thomas]

As long as the tolerances on the bearing face are to spec, you should be fine. Literally every bearing on our car (outside the engine) is pressed into aluminum, and it has been done that way for at least 3 years. No complaints yet.

Two things to be wary of when hard anodizing: It reduces fatigue life dramatically, and there is some material removed in the process. Neither of these will likely have an effect, since these cars don't run very long and the material removed is in the order of a micro meter (maybe less?). I would say it is a good idea to anodize aluminum parts just for the extra corrosion resistance and the slick look.

[G.Arce]

Thank you very much Owen!

I will try to gather more information about the topic, but thank you for your experience.

Originally posted by Owen Thomas:
As long as the tolerances on the bearing face are to spec, you should be fine. Literally every bearing on our car (outside the engine) is pressed into aluminum, and it has been done that way for at least 3 years. No complaints yet.

Two things to be wary of when hard anodizing: It reduces fatigue life dramatically, and there is some material removed in the process. Neither of these will likely have an effect, since these cars don't run very long and the material removed is in the order of a micro meter (maybe less?). I would say it is a good idea to anodize aluminum parts just for the extra corrosion resistance and the slick look.