Fatigue Strength Estimation Based on the Maximum Likelihood Method

Abstract

Fatigue strength is one of the core principles for designing mechanical components. It has been a constant concern for engineers, as mechanical failure occurs due to loading exceeding the fatigue strength. This concern has led to a necessity to develop new approaches to estimate the reliability of mechanical components. The conventional method that is used to test fatigue strength is the staircase method. However, the staircase method’s ability to calculate fatigue strength is potentially unreliable. The bias and scatter associated with fatigue testing shows the limitations of fatigue strength estimation when the staircase approach is used. The conventional methods of fatigue limit determination also have key flaws, in that they are subjective, time consuming, and costly. This research aims to develop a method that would reliably estimate the fatigue strength of materials, whilst using a lower amount of test results. The present study is intended to formulate and analyze a proposed method of estimating fatigue strength by utilizing a smaller number of tests. The aforementioned method that this paper aims to formulate, mainly focuses on a probabilistic estimation based on the Maximum Likelihood Procedure. The proposed method is applied to existing fatigue test data and its effectiveness is compared with other methods.

Description

Keywords

Fatigue strength, Maximum likelihood estimation, S-N curve

Citation

@inproceedings{Lee2023, title={Fatigue Strength Estimation Based on the Maximum Likelihood Method}, author={Sungchil Lee and Odbileg Norovrinchen and Chinguunbileg Sumiyadorj}, year={2023}, booktitle={Proceedings of the second International Conference on Resources and Technology (RESAT 2023)}, pages={98-106}, issn={2352-5401}, isbn={978-94-6463-318-4}, url={https://doi.org/10.2991/978-94-6463-318-4_8}, doi={10.2991/978-94-6463-318-4_8}, publisher={Atlantis Press} }

Collections