Método para quantificação do teor carbono em aços por meio da técnica de WDS

Abstract

Carbon detection by microanalysis in scanning electron microscopy can be carried out using EDS (Energy Dispersive Spectrometry) or WDS (Wavelength Dispersive Spectrometry) techniques. Quantitative analysis, however, for carbon levels typically found in steels can only be performed by WDS, and yet, the analysis is subject to significant distortions when compared with the real value. This is mainly due to the effects as absorption, contamination and peak overlaps. Therefore, in this work it was developed a measurement procedure for carbon in steels applying the WDS technique, which takes into account the inherent quantification interference. For that, steel samples with six known levels of carbon weight percentage were used to build up a calibration curve. Carbon contents ranged from 0.016% to 0.808% and were determined by optical emission spectrometry. The samples were heat treated in order to achieve a homogeneous carbon distribution. The interactions between electron beam and the sample were first analyzed by the simulation software CASINO, which allowed to observe that the higher the applied voltage, the greater the absorption of X-rays generated by carbon in the iron matrix. The following step was to proceed the WDS microanalysis to create curves correlating the carbon content of the samples to the carbon K radiation intensity acquired by the detector, for the accelerating beam voltages of 5 kV, 10 kV, 20 kV and 30 kV. Based on the performed simulation and the characteristics of the obtained curves, it was found that the voltage of 10 kV was more suitable for setting the procedure. Replicating the method and testing the obtained calibration curve with a sample of known carbon content, but different from those used to build up the calibration curve, it was found a relative error of 0.47%. Finally, to validate the method, a sample of 0.592%C, which was subjected to a decarburization process, was used to compare WDS carbon profile to the microhardness profile. The results showed a strong relationship between the curves and carbon values compatible with the alloy composition.