It is common that the endurance limit increases with increase in the stress ratio R. The fatigue limit may be lower than the yield strength when the R = -1. Cyclic hardening could also play a role in increasing the fatigue limit.

I agree with you all. Iman Shakeri

The doubts get in this direction:

In our tests stress amplitude are below YS. Neverthless the maximum stress are above YS! (since R = 0.1, when stress are close to the YS of the material)

Is that common? Or should I consider my curve only as stress amplitude and despise the maximum stress above the YS?

Or should the maximum stress be lower than YS?

That's a lot of doubts.

Thanks for helping me.

The maximum stress should be lower than YS. I wander what was the maximum cycles in your fatigue tests?

Hello Felipe and all,

The best way to look into this problem is devising a thought experiment. Let's suppose you're dealing with SAE1040 for the sake of the argument, although you can play with any material of your choice.

The material properties would then be E=200GPa, UTS=621MPa, Yield=345MPa, K'=1666,n'=0.18 (these two defining the cyclic stress-strain curve), Sf'=1538, b=-0.14, Ef'=0.61,c=-0.57 (these four defining the Coffin-Manson strain-based fatigue curve; or simply E-N curve).

In terms of loading let's imagine you are cycling between Smin=40MPa, Smax=400MPa (R=0.1, as you wanted).

As some plasticity is taking place, using Neuber's Correction we can land at slightly different numbers:

Smin=16.259MPa, Smax=375.4MPa. (I'm not going to provide here the details on how Neuber Correction works)

Now if you use Brown-Miller's method (for example) in order to find the expected fatigue life you'll get Nf=1.55 million, i.e., you can repeat this cycle (16.3MPa/375.4MPa) for more than 1e6 (which is often taken as the endurance threshold or fatigue limit) until you get a fatigue crack.

Now, you can push the maximum stress a bit higher (if you like), up to 450MPa. Then unload it to 90MPa, which sets R=0.2. Using the same material (SAE4140) and the same method (Brown-Miller) you end up with Life = 1.423 million cycles. Obviously the nominal stress of 450MPa becomes 410.3MPa, which is still higher than Yield=345MPa. (and Smin becomes 51.16MPa after plasticity correction).

So, coming back to the question, you can certainly survive a long life (or even infinite life) under special circumstances. The maths (as I quickly demonstrated) can confirm it.

I hope I have helped.

Regards,

Giovanni.

I think there is no such thing like fatigue limit.

We can only talk about the limit value for some base (say, 2.10^6) as about an ordinate on the fatigue curve.

Moreover under irregular loading the limit desappear (if it had been initially - not for all materials, situations, ets)

Privet Gadolina,

Although the subject is debatable, I tend to agree with you, i.e., there is no such thing as fatigue limit. We'd better off saying you're safe or not in regards to a given or adopted target life. In my exercise, I just wanted to show the material can indeed withstand a long life despite experiencing plasticity. At those numbers I provided (Nf=1.55e6 and Nf=1.423e6) some people can be tempted to jump to conclusions and assume the material didn't fail at 1e6, therefore, it may never fail (infinite life). However, this reasoning is erroneous as my numbers clearly lead to finite life, although high cycle fatigue regime.

Bolshoi Spasibo,

Giovanni.

Glad to hear that!

Goog luck.

Can you send me your paper with those results, I will give a reference in some of my future papers?

Dear Felipe,

what is you are mistaken: instead AMPLITUDE of stress you have taken erroneously the RANGE?

RANGE=2*AMPLITUDE

That might has leaded to that paradoxal result!

Hello Gadolina,

I have not published a paper on this subject. I've just done some calcs (using my own codes) with the material properties available for the SAE1040 (taken from the fe-safe library).

Regards

Giovanni.

OK, thanks!

Hello Felipe Pereira Finamor ,

In the uniaxial tensile fatigue test with R=0.1, the fatigue limit in the range was 220 MPa, so the fatigue limit in the amplitude would be 121 MPa. The yield strength was 170 MPa, so I think your results were a common case.

Thank you all for your contribution. Those answers are very helpfull.