On the Empirical Study of Elemental Analysis and Metal Testing using XRF Spectrum Analysis Algorithm

Girish Balasubramanian, Senthil Arumugam Muthukumaraswamy

This paper proposes an algorithm to perform quick X-ray Fluorescence (XRF) spectrum analysis to be used for metal testing and quality control by empirical studies. The metal industry usually deals with large volumes of scrap and metal during recycling. High speed sorting and testing of these metals allow for increased workflow while ensuring accurate identification of these metals. In this paper, An XRF spectrometer (X-MET8000 by Oxford Instruments) was used to gather raw spectra from some alloys. These spectra were fed into a MATLAB script that was designed to implement an algorithm which quickly identifies the elemental content of the sample by comparing it with the spectra input using Pearson correlation coefficients. Spectra from different alloys were analysed using the proposed XRF analyser algorithm in order to accelerate the identification of the elemental content of the sample. The proposed algorithm is successfully proved for the empirical method to which effectively identifies the elemental composition of the sample. This algorithm can also be used for testing of hazardous material. The elemental composition is then compared with the standard benchmark samples of the required alloy for verification as a measure of quality control and a pass is given for matching the composition of the alloy. The computational efficiency of the proposed algorithm is also tested using the MATLAB profiler and was observed to be around 0.003 seconds for quality control and 0.056 seconds for alloy identification.