Calculating Strain Life Cycles for LCF calculations

[Doezer99]

Hi,

Im hoping somebody can help me out here.

Im trying to calculate the corresponding strain cycle, using two elastic stresses
and taking into account the Neuber rule which allows calculation of plastic strains from elastic stresses

If I take a stress location that cycles from Load 1 (tensile) down to Load 2 (compressive) for a number of cycles, say 200 for example.
Lets say BOTH loads are taking the stress well into the plastic region. So the tensile ELASTIC stress is, say 50% higher than yield stress,
and the compressive elastic stress is , say 30% higher than compressive yield stress.

When I calculate the two strain ranges using the Neuber rule.
Is it then correct to add both plastic strains to to get the full strain range?
Im not so sure! That seems VERY conservative. Because when the stress relaxes after Load 1, and moves back down the elastic curve,
the strain doesnt return to zero. it returns to [Plastic Strain for Load 1] - [Plastic Strain for Load 2].
It is the difference between the two strains.

But that seems too non-conservative, if you get me. That is too far the other way...

I understand its really difficult to understand me without some graphs. I will follwo with some a.s.a.p.

thanks
Donal

 

[rb1957]

you are i imagine talking about composite fatigue (not metal fatigue) ?

if composite, well, that's it's own body of knowledge. How composite fatigue is affected by compression, and plastic compression at that, is something i'm not familiar with. if composite, i do "know" that we do less fatigue (well, classical fatigue) calcs with them, relying more on BVID development, or rather non development.

if metal, well, traditionally we discount compression in fatigue. you might compare fully reversed loading with zero-max loading.

but then I imagine you are also talking about low cycle fatigue ? again not something I'm familiar with.

But remember the paperclip example of fatigue ... only a few fully reversed plastic cycles are needed to produce failure.

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[Doezer99]

Hi

yes I am talking about metal fatigue specifically.

Low Cycle Fatigue (Strain-Life Method) calculations.
The compressive does have an effect on this.

But I get confused when trying to work out the strain range for a cycle that is like this for example:

Smax (elastic) = 1.5 * Yield
Smin (elastic) = -1.3 * Yield

SO. its not symmetric.
On the first cycle the material yiel;ds plastically to a high value.
Then on the compressive cycle it goes back down the elastic curve (2 * [Y.S./E]),
and then it yields in compression.
And it repeats back up to max plastic strain then back down to min strain.

SO from that point on, I think the cycle is just
elastic strain + plastic compressive strain only
2 x [Y.S./E]) + eps - plastic

where eps plastic is the plastic component of the strain calculated from teh compressive load.

DOes the max tensile value come into it any more?? i dont see how??

I understand this is rdiculously confusing without a graph.
im tied up here and cant give it time at the moment... will post later...

sorry

 

[LiftDivergence]

For strain-life analysis you need to index the SWT (or some other method) correct e-N curve with the cyclically stable local notch strain range.

As you mention the material may have hardening/softening or relaxation properties which you can account for with a local notch strain analysis using a variety of methods.

When I calculate the two strain ranges using the Neuber rule.
Is it then correct to add both plastic strains to to get the full strain range?

No, you need more than just the plastic strain. You need total strain (elastic + plastic).

You need to be using cyclic stress-strain curves (for the compressive side maybe you can using Masing hypothesis if you don't have the compressive data). But for you e-N analysis the delta_Epsilon you use needs to be defined by your cyclically stable hysteresis loop max strain range (from minimum elastic+plastic to maximum elastic+plastic).

Some good references:
- Search "eFatigue" there are some free notes on strain-life analysis:

https://www.efatigue.com/training/

- You can also check out this very useful NACA paper which talks about computing local notch strain and how to determine the equivalent notch strain range:

https://ntrs.nasa.gov/citations/19750024434

Keep em' Flying
//Fight Corrosion!