Fatigue life ~ 1/load^3 - theoretical basis? 2

[electricpete]

I was just thinking about the life equation.
L10 = 1E6 * (C/P)^3 revolutions.

Did it come from empirical data?

If the mechanism is fatigue, doesn't fatigue S*N accumulate proportional to load which would suggest a relationship (C/P)^1 instead of (C/P)^3 ?

It would seem maybe the explanation is the bearing is operating outside the linear range for fatigue.
Does a bearing typically operate with stresses above the elastic range? Down near the endurance range?

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

 

[Tmoose]

The programs used for FAG/Barden calculations turn red when the contact stress exceeds the material fatigue limits, thus strongly encouraging designs for whom L10 life is meaningless as long as excellent lubrication and sealing are maintained.

FAG has been saying that if the the equivalent stress is less than ~1/8 the basic dynamic stress rating then life can be infinite.

SKF lists the endurance or fatigue limit right along side the other load ratings. Usually less than the 1/8 thumb rule.

 

[electricpete]

Thanks. That makes some sense. But...

at the low end of loading, fatigue cycle accumulation rate is proportional to P^0 (life proportional to P^(-0) )

at the high end of loading, fatigue cycle accumulation rate is proportionalt to P^0 (life proportional to P^(-1) )

So for a range including both high and low loading we might fit a life curve something between P^(0) and ^(-1).

I would pick P^(k) where -1 < k < 0.

But the standard relation is P^(-3)... how is that consistent? Did Lungren whatshisname make it up?



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

 

[diamondjim]

Palgren basically said that life
varies by the cube of the load.
So if your load is 1/2 the rated
load you life would increase by
the 2 cubed or by 8 times.
I think it was by 3 for ball bearings
and 3.333 for roller bearings.

 

[electricpete]

That's what he said, but isn't the basis fatigue failure and doesn't fatigue cycle accumulation rate go somewhere between the 0th power (below endurance limit) and 1st power (above endurance limit) instead of 3rd power?

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

 

[diamondjim]

Palmgren based his work on spalling of
the raceways being approx .0001 times
the roller diameter which is not really
the same phenomenom. This is only at
the surface and not at the core of the
material. The surface material might be
in the range of 63Rc and the core may be
only in the 24 Rc range. Really not
certain what endurance limit is imposed
on an non isomorphic product.