Dear Gang Zhu.

Fatigue only occurs on engineering structures subjected to cyclic loading or repeated loading such as pavements, bridges, railway, airport runway, etc. Fatigue life assessment is related to the service life of the structure. Service life is a period whereby the engineering structure can fulfill its design requirements so that major repairs are not required. Fatigue life assessment is necessary when an engineering structure fails in fulfilling its design requirements. The fatigue assessment can be conducted by visual investigation (Cracking as evidence of distress, deflections, deformations), etc. However, for complex engineering components, the load testing of structures is required for the final decision. Moreover, some empirical models or semi-empirical models can be used to predict the engineering structure fatigue life. Besides, fatigue life is affected by the overloading, physical attacks, chemical deterioration, quality of material, Workmanship, quality of design.

Fatigue analysis refers to one of three methodologies: local strain or strain life, commonly referred to as the crack initiation method, which is concerned only with crack initiation (E-N, or sigma nominal); stress life, commonly referred to as total life (S-N, or nominal stress); and crack growth or damage tolerance analysis, which is concerned with the number of cycles until fracture.

The method for calculating fatigue life is sometimes called the Five Box Trick, including material, loading, and geometry inputs, and analysis and results. The three main inputs for fatigue life analyses are processed using various life estimation tools depending on whether the analysis is for crack initiation, total life, or crack growth.

Best regards,

you can use the limit states criteria to estimate the fatigue life of the structures. check the standards in your country on limit states criteria

This link may be useful

https://www.sciencedirect.com/topics/engineering/fatigue-assessment

You can use FITNET method.

http://eurofitnet.org/fitnet/index.html

Article FITNET Fitness-for-Service Procedure: An Overview

It is describe fracture, fatigue, creep and corrosion of a construction.

You just remenbered us the main question in fatigue of materials.

Going back to the roots, I will consider the following facts.

1. In which regime I wat to make the estimation.

2. Which approach applies better in this regime? Mechanical, mathematical or Statistical?

3. Which kind of experimental information is available. It is a common specimen, material, geometry? or everything is new?

4. There are several models to evaluate the data. Choosing one or two whose requirements can be satisfied with the available information.

5. How to interpret the results? Most difficult if the material and the specimen are not common.

I will appreciate your comments.

First of all, thank you for dealing with an original topic. My field of work; Mechanical tests of composite materials are about low speed impact response and fatigue behavior. We try to experiment according to ASTM standards regarding fatigue. But in complex structures, it is not possible to find everything in standards. For this, it is necessary to look at the studies in the literature. In general, before starting the fatigue life tests of the materials, it is necessary to obtain the tensile test results and determine their maximum strength. Considering the maximum stress values ​​obtained, it seems a more appropriate method to determine the stress ranges required for the fatigue test. In preliminary studies, choosing the stress values ​​in the range of 20-70% of the maximum stress values ​​to obtain the S-N diagram gives healthier results in terms of fatigue life predictions. I would like to state that I made these comments in the light of the experiences I gained as a result of my experimental studies and literature research. I hope I have understood your question correctly and have provided useful ideas for your work. All the best.

Accuracy is pretty much difficult to reach.

If you need to design a structure against fatigue you can make use of S-N curves in standards, each of them following different approaches.

If you have a structure and you want to assess or verify the design of an existing component then you can still make use of the standards above. For better accuracy it is best to measure a reprentative load history and run fatigue tests accordingly. This can be done following local or global approaches.

Similiar considerations can be done using fracture mechanics for determining inspection intervals aiming to a safe use of the structure.

If it is a welded structure, it is recommended to use the traction stress method. I have done some simple work about this issue before. You can find it through this link below：

10.1016/j.tws.2020.107348

Welcome to discuss together.

This link may be useful

Article Multilaser spot tracking technology for bridge structure dis...

My PhD work on uncertainty in the fatigue crack growth models currently recommended in BS 7910: Article Consistent and Coherent Treatment of Uncertainties and Depen...

This considers how best to combine multiple laboratory test datasets, and how to quantify the multi-variate uncertainty in the corresponding model parameters.

It depends on whether the structure is steel of concrete. Even for concrete structures, it is always the concrete member is associated with reinforcement. So i think it's critical to evaluate both from material level and structural level.

I would suggest considering a continuum damage model, considering that the accumulation of damage due to environmental conditions (Service loading) is a random phenomenon. Also due to the fatigue complexity, the models vary depending on the particular application and the reliability factors that need to be considered.