I recommend the Walker equation. It has an adjustable exponent that is specific for every alloy. You can find the values for it in the literature.

There are many mean stress correction schemes. All work for some material but none work for all materials. Dowling and some of his students wrote a couple of excellent papers comparing some of the most popular and provided generic guidance by metal category. Here is his Google Scholar search result:

https://scholar.google.com/citations?user=KdKo__QAAAAJ&hl=en

Those in Fatigue & Fracture of Engineering Materials & Structures were particularly useful in my previous life.

The most important thing when picking which mean stress correct method to use is determining which one models your data within the range of the data you have. Do not extrapolate.

Dear Fabian Dwenger

Thank you for your answer. The Presentation contains some valuable information. I also referred literature written by Prof. Dowling and from the given suggetions decided to select using "Walker equation as the mean stress correction".

Dear Murat Tiryakioḡlu Sir and Doug A. Hall Sir,

First of all thank you both for your suggestions.

I was also reading "Mechanical Behavior of Materials" by Prof. Dowling regarding mean stress correction and it was mentioned that Walker equation provides good fit to practical data so as Murat Tiryakioḡlu said I've decided to select "Walker equation" as the method of mean stress correction.

To obtain material constant/adjustable exponent (gamma) I refereed the following two research papers.

1. "Mean stress effects in stress-life and strain-life fatigue(2009)",Norman E. Dowling

Which includes a empirical equation for steels to obtain value for gamma.

2. "Mean stress effects in stress-life fatigue and the Walker equation"(2008), This paper includes gamma values for several steels, stainless steels and aluminium alloys as Murat Tiryakioḡlu have mentioned.

I will also take look on the reference Doug A. Hall has provided to learn more about this.

Also Doug A. Hall Sir, What do you meant by " The most important thing when picking which mean stress correct method to use is determining which one models your data within the range of the data you have. Do not extrapolate. "?

Thank you.

Best Regards,

Rashiga

Rashiga Walallawita -

The Walker equation is very general and works in many situations. It is one of the default models for analyzing fatigue data presented in MMPDS (previously known as MIL-HDBK-5.) It's straightforward to find the four parameters needed for the Walker equation to fit your specific data set. There are no general solutions - we cannot say that gamma = 0.6 for all aluminums, for example. After you have found the curve fit for your data, plot you data against that curve to get a feel for how well the model works. Optimizing the values of the curve fit parameters does not mean it fits the data well.

"Do not extrapolate." Do not use any curve fitted to data to make predictions about what will happen outside the range of the data. If your test data is for R-Ratios between -1 to 0, don't assume it will automatically predict what happens for R=0.5. If your test data is for test failures are for 10,000 cycles to 10,000,000 cycles, don't assume you can predict the stress that can be safe for 100 cycles. If your test stresses range from 20 ksi to 60 ksi, don't assume you can predict how long your part will last if the stress is 100 ksi.

I've seen a lot of fatigue data and fitting an equation to your data means that the model is suitable for the conditions you have tested, which may not cover the conditions of your problem. Any time you go outside the range of your data you need to understand the assumptions of your model, the likelihood that those assumptions are wrong (I guarantee that some are), and the consequences of the errors in your model. Fortunately, it is safe to claim that the life at 10 ksi will be longer than the life at 20 ksi. And you can make conservative assumptions to handle R=0.5.

Doug A. Hall ,Thank you Sir for the explanation!

If you are looking at lives below 100,000 cycles, you should probably not be using stress-based approaches anyway. At those low lives, since there is a lot of plasticity, you should use strain life and SWT for mean stress correction. Unfortunately strain life analysis packages can be expensive.

https://re-test.com/fatigue-analysis-techniques

Dear Doug A. Hall . Sir,

I have two questions regarding fatigue analysis,

Q1) Some fatigue simulation software allows to generate S-N curve data from tensile test data such as UTS and Yield stress. When there is no S-N data present is it better to use these generated S-N data?

Will I face any issue In terms of publishing these simulation results in a journal?

Q2) Due to sudden changes in geometry there can be stress concentrations, where hardening can occur in these small regions. Can this plastic hardening increase the fatigue life?

Thank you.

Rashiga Walallawita,

A1) I have used SN curve estimation methods in the past when I had absolutely no SN data. However, this should be a last resort approach, applying large margins of safety, as the methods fail to predict consistently safe fatigue curves.

See Basan, Ruesa, Franulovic, & Marohnic, Some Considerations on the Evaulation of Methods for the Estimation of Fatigue Parameters from Monotonic Properties, Procedia Engineering, Volume 101, 2015, pages 18-25

A2) Yes. No. Maybe.

Dear Sir,

Thank you for your reply. Also thank you for the mentioned research paper.

Best Regards,

Rashiga

As far as I can tell SWT and Walker are indicated for brittle materials and Morrow for ductile materials. Goodman is over conservative and Gerber provides wrong results for compressive stresses. At LCF regime the most common methods for fatigue life assessment are: Fatemi-Socie, SWT and Brown-Miller. Brown-Miller uses the Morrow correction on the right hand side. Fatemi-Socie uses the maximum normal stress and the yielding limit combined (1+k*Snmax/Sy) on the left hand side. This combination modifies the shear strain amplitude. Both FS and BM are indicated for ductile metals. SWT is indicated for brittle materials.

Dear Giovanni De Morais Teixeira ,

Thank you very much for your recommendation.

By any chance do you have any reference document mentioning about the usage of each mean stress correction method?

Any help would be much appreciated.

Thank you.