Statistical Processing of a Voltage Signal of the Magnetic Barkhausen-noise Emitted by Quenched and Tempered Specimens of C45 Steel

The paper treats the application of the micromagnetic method to the determination of hardness of C45 hardened-and-tempered steel on the basis of a voltage signal of the magnetic Barkhausen noise, The specimens were hardened and tempered at different temperatures so that small differences in hardness of individual specimens and, consequently, small differences in captured voltage signals were obtained. The magnetic Barkhausen noise occurs in magnetisation of steel and can be captured by a measuring coil as the voltage induced in the measuring coil. The captured voltage signals were used to determine a power frequency spectrum and the variance was selected as a characteristic value. A statistical analysis of the voltage-signal variance was used to assess the reproducibility of the latter at the specimens treated in the same way in a temperature range from TT1 = 220 °C to TT29 = 500 °C. In addition to the assessment of reproducibility of results the discriminatory power of the variance in determining the hardness achieved was determined by means of the statistical t-test. The statistical treatment of the Barkhausennoise voltage-signal variance was used to assess two neighbouring specimens showing a difference in the tempering temperature, i.e., ∆T1 = 10°C, ∆T2 = 20°C and ∆T3 = 40°C respectively. The statistical analysis of the discriminating power of the variances considered two neighbouring tempering temperatures in the same temperature range, i.e., between TT1 = 220 °C to TT29 = 500 °C. Introduction: In the assessment of machine parts microhardness variation and residual-stresses gradient in the surface layer are very important. Recently to this aim various non-destructive testing methods have been increasingly applied, particularly because of the direct applicability of the methods to materials testing. The automated production of machine parts requires on-line monitoring of the state of material, therefore, the methods applied should be sufficiently reliable, fast and reproducible. One of such methods is the micromagnetic method based on the Barkhausen noise. From the viewpoint of physics, the micromagnetic method is based on the fact that a ferromagnetic material, when magnetised by the alternating current contains small magnetic domains. When an external magnetic field affects a ferromagnetic material, a movement of magnetic-domain walls occurs, which produces changes in the size and shape of the latter. A variation in the magnetic flux induces voltage in the measuring coil, which can be registered, and then processed (Jiles et al., 1994; Theiner et al., 1987; Mitra et al., 1996). Numerous studies have shown that relatively small differences in mechanical properties of ferromagnetic materials can be efficiently detected by the micromagnetic method based on the Barkhausen noise. Experimental setup and testing procedure: For investigations an experimental setup was arranged to capture voltage signals of the magnetic Barkhausen noise (BN). It consisted of a magnetisation unit, a sensor for capturing voltage signals, a signal amplifier with a relevant bandpass filter, and a computer-aided unit for determination of microstructure or microhardness and residual stresses. Fig. 1 shows a block scheme of the experimental setup for micromagnetic testing based on the Barkhausen noise. Before starting experiments, optimum magnetising parameters producing the movement of magnetic domains characteristic of the Barkhausen noise were to be determined. Very important parameters in magnetisation are the magnetising frequency (fe), the magnetising current (I), and the magnetic field strength (H) in the specimen. The magnetic field strength depends on the magnetising current (I), the number of windings (N) of the yoke, and the mean path length of the magnetic flux in the yoke and the specimen (L). UNIT FOR CAPTURING THE BN VOLTAGE SIGNAL MEASURING COIL PREAMPLIFIER BAND PASS FILTER UNIT FOR PROCESSING BN VOLTAGE SIGNAL COMPUTER PS 486 SOFTWARE (LabVIEW) EXPERT SYSTEM FOR DETERMINATION OF RESIDUAL STRESSES ACCORDING TO CALIBRATION CURVES V f , , H I f MAGNETISATION FREQUENCY H MAGNETIC FIELD STRENGTH V VARIANCE OF VOLTAGE SIGNAL I MAGNETIZING CURRENT