Measurement techniques for meltdown control in a stainless steel AC electric arc furnace

A method for vibrational evaluations of spectrogram data from pilot plant measurements was established for obtaining a consistent analysis of the molten scrap state. Not knowing the meltdown progress inside the furnace results in an inefficient power scheduling, which affects refractory life, process time and energy losses. Hence, there exists potential to control the furnace power so that it is kept high while there is sufficient amounts of scrap inside furnace to absorb the input energy. Thus, harmonic distortions of the electric current and measurements of the sound and vibrations emitted from the furnace were correlated to the molten state of the scrap. By using this method, a reduction in melting time in the order of minutes per heat can be obtained. The method was found to give stable signals and it was judged to be a good candidate for implementation in industry. Furthermore, the developed method is universal and may be applied for other types of signals and combined into intelligent soft sensor systems.
QC 20190107