On 2010-12-03, jim beam > wrote:
> On 12/03/2010 06:44 AM, Brent wrote:
>> On 2010-12-03, jim > wrote:
>>
>>> of course you are - that's why ball bearings are more commonly used,
>>> because of their misalignment tolerance.
>>
>> It is clear you don't understand the geometeries involved or the
>> designs. Ball bearings are used because they are CHEAP and EASY to
>> install in a manufacturing envirionment. Your angular ball bearings are
>> tolerant of the cheap design. If you use tapered roller bearings
>> correctly you get a much more robust design that is also much more
>> expensive, if for nothing else the cost of the bearings.
>
> 1. there's not a significant difference in cost.
>
><http://www.autopartsworld.com/results/?PN=9265&N=0&VN=4294967193+4294966864+4294965966+4 294967247+4294966883&Nr=AND%28universal:0%29>
>
><http://www.autopartsworld.com/results/?N=0&VN=4294967193+4294967031+4294967025+429496693 7+4294966998&Nr=AND%28universal:0%29&PN=9265>

It's not just the cost of the bearing itself. But the machining and
and assembly. I was VERY clear about that. Plus there are TWO bearing
assemblies in the tapered roller bearing design. Also looking at retail
cost is irrelevant compared to what manufacturers pay.

> 2. "cheap design" is not loading that can be inside a 40°+ arc.

Define this and describe why it is meaningful to a wheel bearing set up
that has been done properly with a bearing that can withstand both axial
and radial loading.

> 3. you're still not getting it on /why/ tapered rollers only have a
> small angle of application.

Whatever wierd ass loading you're using to make a usenet point simply
does not apply. Bearings are fixtured to control the loads that they
see. If you aren't going to use a bearing correctly then all bets are
off. What it seems you want to do is take a single tapered roller
bearing, press it into a knuckle and then press the hub into the the
bearing just like is done with the ball bearings. This is not the way
to use them, they aren't designed to be used that way. And yes, if
used that way they will fail because they aren't fixtured properly
and thus will see wierd ass loads outside their design. The races will
indeed pivot relative to each other and not distribute the load on the
rollers correctly. But, if used correctly, in a design so made, they
won't see those loads, only the axial and radial components as they are
designed to handle.

It's a cheap way to assemble so what is used instead is the angular
contact ball bearings so the cheap machining, cheap assembly, and
cheaper bearings can be used. To use the tapered roller bearings which
can withstand higher loads as you admitted, the design is different and
more expensive. I keep repeating that. You must use bearings correctly.
You're comparing an incorrect use of tapered rollers to a correct use of
angular ball.

>>>> If you're loading
>>>> through only part of the cup or cone to get the uneven loading
>>>> you are describing you've done a ****ty design and/or the shaft in the
>>>> inner race isn't concentric with the bearing mount holding the outer
>>>> race.
>>
>>> nope, it's a practical reality. i don't know what kind of tolerance you
>>> think you're getting on the average steering knuckle [for instance], but
>>> it sure ain't the sub-micron tolerance you're getting on the bearing.

>> Now it's really clear you have no understanding. Make the steering
>> knuckle hole or spindle off by an inch location wise, hell make it 15
>> degrees off angle wise too while you're at it, what's in the inner race
>> of the bearing is still concentric with the outer race of the bearing
>> within the tolerance of the bearing. The bearing is fixtured in the
>> knuckle the hub is fixtured in the bearing. Or the bearings are fixtured
>> on the spindle and the hub is fixtured on the bearings. There's nothing
>> on the other side of the wheel that's not concentric creating a load
>> because it's only fixtured on one side.
>>
>> The only way you can get off centered is with a spindle (or knuckle
>> hole) that that has two sections that aren't concentric with an
>> inner/outer wheel bearing design. This requires actually being
>> monumentally stupid and doing the machining in two steps
>> with re-fixturing of the workpiece in between to create such error. Only
>> an idiot would allow such a thing. The entire spindle (or knuckle hole)
>> will be machined with the part fixtured once.
>
> uh, i don't know how you drive, but when i drive, i go around corners.
> when i go around corners, the loading axis on my wheels changes. hence
> the bearings get loaded "off axis". if you can't understand that, this
> conversation hasn't gone anywhere.

I refer you to two posts ago where I told you that at the bearing the
load is broken down into two vectors. Axial and radial load. That's what
the bearing sees because it sits in a fixtured environment. It doesn't
care about the loading axis at the tire. It sees what that vector breaks
down into. That's why bearings are rated for their axial and radial
loads and not rated for every degree from zero to 360.

Also Tapered roller bearings were used for DECADES succesfully with much
poorer manufacturing tolerances than today's. They were used properly.
Your argument comes down to a usenet irrelevance of taking a design for
an angular contact ball bearing and shoving a tapered roller bearing in
it. That's irrelevant. Bearings need to be fixtured according to their
design. When you put together a _proper_ tapered wheel bearing design
you'll have something that is _more expensive_ than a proper angular
ball bearing design. The tapered roller design will be able to withstand
higher loads.

>> Spindles with inner and outer tapered roller bearings were manufactured
>> successfully for decades without the benefit of modern CNC machine
>> tools. The spindle is a machined shaft. Making a machined shaft with
>> both ends concentric is not rocket science. Neither is making a bore
>> that has both ends concentric. I'll tell you right now the outer side of
>> that steering knuckle bore that holds an angular contact ball bearing is
>> sufficently concentric with the inside part of it that you can put
>> tapered roller bearings on both sides of it and have proper loading and
>> alignment. The reason? You couldn't press in your angular contact ball
>> bearing (easily and without damaging it) if it wasn't concentric. If the
>> hole was wavy or stepped the bearing would jam upon being pressed in.
>
> see above. you need to understand the basics before you start talking
> about anything else.

Now it's very clear you want to shove a tapered roller into a angular
ball design. YOU CANNOT DO THAT. The entire design has to properly
fixture the bearing(s).

>>>> The races shouldn't be loaded like that in any kind of bearing
>>>> except those designed to have the inner race rotate relative to the
>>>> outer.

>>> dude, you can't /not/ have them misaligned to some extent. at least,
>>> not with any loading that causes elasticity [which is by definition
>>> inevitable], and without spectacular precision in the machining of the
>>> parts to which they're fitted.

>> Dude, you have ONE FIXED END. They will ALWAYS be concentric within the
>> tolerance of the bearing. There is no other fixed end to cause a
>> misalignment. Machining concentric shafts and straight holes is not
>> "spectacular precision". You couldn't even get your angular ball bearing
>> in the hole if it wasn't straight. The concentricity of the assembly
>> between the hub's bearing mounting and the knuckle/spindle's bearing
>> mounting is completely driven by the bearing. There's no other end to be
>> off center to create a load.
>>
>> Now if you're going to argue that where the wheel is mounted is off
>> center from the end that's mounted to the bearing, you're back at being
>> cheap ass again and no matter how tolerant of said misaligment your
>> bearing is the car is going to go thump-thump down the road.
>>
>> If you cannot understand basic tolerancing there's no point in
>> continuing. I'm not even going to bother reading the rest. three
>> strikes.

> it's ironic that you try to talk of "tolerancing", yet don't seem to
> understand the fundamentals of load application. without that,
> "tolerancing" is utterly irrelevant.

Now you're just doing declarations. Clearly you're out of your league
here. And I spelled it correctly so I don't know what your problem is.
It seems that you simply lack the knowledge of how to do more than one
design. You're thinking parts swapping. You made that clear with your
idiotic comparison of retail bearing prices. You don't even grasp how a
tapered roller bearing design is different.

Here is the angular ball type:
http://www.autopartslib.com/wp-conte...ts-diagram.png

Here is tapered roller:
http://www.4x4abc.com/G-Class/img/wh...ring_CV72.jpeg

See the difference?

Here's tapered roller as a single press on piece, notice the proportions
of hub and mounting.
http://www.timken.com/en-us/products...pindleHub.aspx

Are you also noting that heavier vehicles are still using tapered
roller? Why do you think that might be?

Why would a truck or a 4x4 off road vehicle want to use tapered roller
bearings if what you say was the least bit true?