Exploration of data science techniques to predict fatigue strength of steel from composition and processing parameters

Ankit Agrawal, Parijat Deshpande, Ahmet Cecen, B. P. Gautham, Alok Choudhary, Surya R. Kalidindi

Integrating materials and manufacturing innovation · 2014 · 274 citations · 30 references

DOIFull text

Open access

Concepts

TL;DR

Data analytics tools are increasingly used to predict steel fatigue strength, but documented success remains limited, especially for extreme properties like cyclic fatigue where physics‑based models fall short. The study investigates applying data science techniques—feature selection and predictive modeling—to steel fatigue strength using NIMS public domain data, aiming to develop a systematic end‑to‑end materials informatics framework. They use NIMS data to apply feature selection and predictive modeling, building a systematic end‑to‑end framework for materials informatics. Neural networks, decision trees, and multivariate polynomial regression achieved R² > 0.97, outperforming prior methods, and enabled ranking of composition and process parameters for predicting fatigue strength, with predictive models successfully developed.

Abstract

Abstract This paper describes the use of data analytics tools for predicting the fatigue strength of steels. Several physics-based as well as data-driven approaches have been used to arrive at correlations between various properties of alloys and their compositions and manufacturing process parameters. Data-driven approaches are of significant interest to materials engineers especially in arriving at extreme value properties such as cyclic fatigue, where the current state-of-the-art physics based models have severe limitations. Unfortunately, there is limited amount of documented success in these efforts. In this paper, we explore the application of different data science techniques, including feature selection and predictive modeling, to the fatigue properties of steels, utilizing the data from the National Institute for Material Science (NIMS) public domain database, and present a systematic end-to-end framework for exploring materials informatics. Results demonstrate that several advanced data analytics techniques such as neural networks, decision trees, and multivariate polynomial regression can achieve significant improvement in the prediction accuracy over previous efforts, with R 2 values over 0.97. The results have successfully demonstrated the utility of such data mining tools for ranking the composition and process parameters in the order of their potential for predicting fatigue strength of steels, and actually develop predictive models for the same.

References

30