Statistical signal processing of optical emission spectroscopy for micro inclusion detection in steel

In order to improve the quality of molten steel, a new algorithm for rapid determination of inclusion characteristics is studied. The principal factor that decides high-grade steel quality depends on size distribution of oxide inclusion or total oxygen in steel. Element composition and size distribution of inclusion cannot be determined by the current method of oxide inclusion determination at the same time. A new technology for inclusion determination should be developed since the analyzing time of inclusion in steel is too long. To reduce the analyzing time, the technology should be based on modified optical emission spectroscopy coupled with equipment for pulse discriminator preamplification, data acquisition function, and data processing. In this paper the determination of alumina inclusion size distribution, inclusion composition, and steel cleanness are studied by the new method. The iron HIP samples with alumina inclusion generate pulses in the right part of Gaussian fit distribution. These pulses indicate alumina inclusion signals and the pulse intensities are related to the inclusion size distribution. The relationships between pulse intensity and inclusion size are also discussed