Fatigue Analysis vs Transient Stress - Direct Integration

[Anonymous]

Hello,

I performed both a Fatigue Analysis and a Transient Stree-Direction Integration on a model.  The FA indicated failure within 35 cycles.  I then used the same load curve over 35 cycles in the TS-DI analysis and the result was a maximum stress well below the yield strength and ultimate tensile stregnth of the material.  For example, the node with the lowest Log10 life had a stress approximately equal to 1/3 the ultimate tensile strength.

Should I expect the TS-DI to show a stress value greater then the ultimate tensile strength for the node above?

The above TS-DI analysis leads me to believe no failure at 35 cycles.  Can a TS-DI analysis also be used to preduct failure? 

Regards,

PSteven

[S.LI]

Hi, PSteven,

FA and TS-DI talk totally different things.

TS-DI shows you the stress/strain and FA tells when the structure/part could fail due to the accumulation of damage.

The damage here is a general concept and could be caused by lots of reasons.

Usually, the maximum stress is far below yield stress in high cycle loading (HCL) fatigue analysis. But after around 1e5~1e7 cycles, it could fail.

In low cycle loading fatigue, the maximum stress could be higher than yield stress,  but it definitely is lower than ultimate stress.

If the maximum stress is larger than ultimate stress, that's called "Fracture analysis", not fatigue analysis.

[Anonymous]

Hi S. Li,

Thanks for you post.  The LSS simulation I ran in order to perform the Fatigue Analysis had a maximum stress greater then the ultimate strength for the particular loading I was applying.  So what you are saying makes sense, an accumulation of damage.

I suppose the stress in the TS-DI is an average stress, since the same load and load curve was applied that would create a maximum stress greater then the ultimate strength.

Regards,

PSteven

[S.LI]

Hi, PSteven,

"The LSS simulation I ran in order to perform the Fatigue Analysis had a maximum stress greater then the ultimate strength for the particular loading I was applying."

I don't think this is right. The maximum stress for Fatigue Analysis should be always smaller than ultimate stress, otherwise, we know the life is less than 1. Please correct me here.

In Fatigue Analysis, the most important thing is "cycle". No matter what the magnitude is, its contribution is always counted in.

---

PSteven wrote:

Hi S. Li,

Thanks for you post.  The LSS simulation I ran in order to perform the Fatigue Analysis had a maximum stress greater then the ultimate strength for the particular loading I was applying.  So what you are saying makes sense, an accumulation of damage.

I suppose the stress in the TS-DI is an average stress, since the same load and load curve was applied that would create a maximum stress greater then the ultimate strength.

Regards,

PSteven

---

[Anonymous]

Hi S.LI,

This is my understanding:

In order to perform a Fatigue Analysis an LSS needs to be ran first, so that the Fatigue Analysis option is available.  The LSS I ran had a maximum stress (55k lbf/in2) that was greater than the ultimate tensile stress (Al A356 Sand Cast from the parts library).

I understand what you are saying, that by definition if stress is greater then ultimate then it shoudl fail.  However, when I ran the FA I had about 35 cycles before failure.  The LSS did not use load curves, obviously, but the FA did.  Perhaps for the load curves used in the FA the stress did not reach the maximum that it did in the LSS.