A new method for complexity determination by using fractals and its applications in material surface characteristics

Abstract

In this article, a new method for complexity determination by using fractals in combination with an artificial intelligent approach is proposed and its application in laser hardening technology is detailed. In particular, nanoindentation tests were applied as a way to investigate the hardness properties of tool steel alloys with respect to both marginal and relevant changes in laser hardening parameters. Specifically, process duration and temperature were considered, together with nanoindentation, later related to surface characteristics by image analysis and Hurst exponent determination. Three different Machine Learning algorithms (Random Forest, Support Vector Machine and k-Nearest Neighbors) were used and predictions compared with measures in terms of mean, variability and linear correlation. Evidences confirmed the general applicability of this method, based on integrating fractals for microstructure analysis and machine learning for their deep understanding, in material science and process engineering.