What is the best dimension for Raceway? 2

[FreddyMusic]

One of my customer is able to check dimension of raceway,
we send some brg of Reputable bearing manufacturer for trial.
The result is strange. Raceway dimension/ball diameter are not same?

SKF I.R. 0.515 O.R. 0.525
FAG I.R. 0.52 O.R. 0.54
NSK I.R. 0.515 O.R. 0.52

What is the best dimension for Raceway? Any thumb rule for this?
Or any design viewpoint from bearing type or applications?

 

[electricpete]

I assume you are talking about bearings with the same AFBMA number from different suppliers.

The external dimensions necessary for mounting are of course controlled by the standard, but the internal paramters such as number of rolling elements and raceway dimensions are not controllled by the standard (other than indirectly through load-carrying requirements).

One basic design decision is how many rolling elements. If the suppliers of two bearings each choose different number of rolling elements, that the races and rollling elements will likely be almost the same size among those two manufacturesr. If the suppliers each choose the same number of rolling elements, those internal dimensions will likely be very close.

It is of course no concern to the end user since you are not expected to disassemble a bearing and replace individual parts. It is a minor complication to guys who monitor rolling element vibration since you have to get the fault frequency from a database that includes not only the AFBMA number of the specific bearing but also the manufacturer.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[diamondjim]

Generally the closer the numbers
approach the .50 factor, the higher
the static load will be. The closer
it nears .50, the higher the turning
torque is required to turn the bearings
and generally greater precision is
required or greater clearance is
required to allow low bearing
turning torque. The contact stress
are also reduced as the numbers
approaches the .50 factor. This figure
is simply the conformity radius divided
by the ball or roller diameter.
.50 represents perfect conformity,
i.e. the conformity radius equals
the ball RADIUS. The contact stresses
are generally limited by the inner race
conformity ratio and many manufacturers
allow a greater conformity ratio in the
outer race to equalise these stresses.

So if you want lower torque, you want a
higher ratio. If you want static capacity
to increase, you want a lower ratio.

 

[FreddyMusic]

Thanks DJ for comments.

Some further concerns, I assume on deep groove ball bearing.

1.Why not approach 0.5 as much as possible? (Only for heavy duty)

Why manufacturer approach perfect value as close as they could?
With latest craft grinding +Super finish…etc. Not possible?
Afraid of problem on assembly? Define tolerance of Race ratio only plus.
Make a small radius near race and shoulder is well-known.
Even 2 points contacts on I.R. + 2 points contacts on O.R. seems better for Load capacity.
I understand spherical roller is perfect choice.
But 0.5 -> 0.52 still makes big stress difference.

2.Max. Hertz contact stress Limit on Inner ring race / balls

Generally, this is only statically viewpoint from one bearing section.
Anyone consider also the race circumference difference between outer / inner race?
I wonder, does it necessary to equalize outer race to inner race stress?
Statically Hertz contact stress seems far to experiential A. Palmgren life theory.

 

[diamondjim]

You can appreciate that the balls overheat
with respect to the raceways and the .50
condition would lock up the bearing or
create edge loading in a hurry. Many of the
deep groove ball bearings are 4 point contact
to accomodate both axial and radial loads.
4 point contact bearings could not exist with
the .50 conformity factor. Some manufacturer's
use the same conformity factor in the both the
inner and outer races to make the grinding
process identical. This also makes it easier
to calculate the clearances in the bearings.
I have seen conformity radii as high as .70
to reduce the turning torque and decrease the
heat in the bearings where the loads are light
and have high revolutions or a slight preload.

Yes the inner race has a negative curvature
while the outer has a positive curvature which
helps to envelope the raceway. That is why
the outer race can have a greater conformity
factor than the inner race. This is the
theory that Palmgren and others developed
when you examine the depths of the hertzian
contact stresses on each of the races.

You mention spherical roller bearings. Anytime
you approach line to line contact instead of
point contact, the static capacity increases
as well as the turning torque. The tapered
roller bearings allows for loads both axially
and radially loading.

In general, ball bearings are cheaper to manufacture
than roller bearings.

 

[FreddyMusic]

Big thanks to Jim for share valuable experience.

Two points still confuse me. (Forget about Lube or EHD for a while )

1> I assume, SKF (I.R. 0.515 , O.R. 0.525) is correct for general industry application (Not for Hi-speed).
What is the initial design intention for a slightly bigger O.R.?
To have an equalized stress on outer ring and inner ring?
OR
To have an equalized life for outer ring and inner ring? Or even balls ?


2> NASA/TM—2001-210462 mentions about their angular contact ball brg.
Conformity percent: Inner = 54% , Outer= 52 % , reach DN = 2.5 million or higher.
What‘s your concern or comments for this example?
Inner race ratio bigger than outer race ratio


My intention of ball bearing assume is to have a physical model, as simple as possible.
I believe, Ball bearing is enough complicated for me.
Nevertheless, all spindle brg. ABEC7-9 brg, roller element are also balls.

 

[diamondjim]

As to the NASA TM, I assume it would be a
mistake. The thinking about having the outer
greater than the inner is that almost all
bearing fail on the inner race because of
the higher stresses imposed because of the
negative curvature of its raceway.
By using a larger conformity on the outer,
they are trying to allow easier turning
torque in the bearing.
Basically .51 is the limit that most customer's
use for the conformity radius. I think most
catalogue ratings would be for .52.

I am not that familiar with the NASA spec but if
it only for very high rpm bearings they may be
trying to overcome the centrifical forces
which would impose a higher load on the
outer race?

 

[FreddyMusic]

DJ, I examine report once again.
They use oil-jet and oil-mist for cooling and lubrication at inner race / balls, where generate vast heat.
Perhaps they are not more care about outer race ratio.
or perhaps they try to to equlize temperature of inner and outer ring.

High Speed + Low temperature is they are looking for.

 

[diamondjim]

There is some confusion in the terms
roller and rolling. Ball and Roller
bearings are rolling bearings. Also
there are Bearing Roller elements in a
Roller Bearing which are commonly called
simply Rollers verses Balls in a
Ball Bearing. The reason I make this
distinction, often people want to buy
only the rollers or balls for a bearing
and not the complete bearing assembly.
Ball or Rollers hardly ever fail before
the race ways. The points on the raceways
see many balls or rollers pass thru the
load zone but the other times during rotation
these rolling elements see little load.
I hope this makes sense.

The only thing I find confusing in Static
and Dynamic Capacity is that the Static is
often based on the Square of the Rolling
Element Diameters and the Dynamic Capacity is
based on the cube power of the Rolling Element
Diameters. I have always been intriqued by
this phenomenom.

 

[FreddyMusic]

I think this has been standardized by ISO 76 and 281.
I am afraid, we have no change to remark it.

 

[diamondjim]

I am not proposing changing these but
am trying to understand the physical
reasoning between the differences.

 

[diamondjim]

Do you have this article?
as well as Palmgren's literature?

Stribeck, R., "Characteristics of Plain and Roller
Bearings", Zeit. V.D.I., 46, 1902

Also of interest:

http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19950020354_1995120354.pdf

 

[FreddyMusic]

Diamondjim,

Thanks for the information.
I vote one more " diamond "