Authors of the paper present investigation results concerning properties of ADP, ADC and ADNC descriptors which have been obtained during measurements made within three oil transformers and then tested also by other methods (electric and dissolved gas analysis ones). Methodology which makes easier an analysis and identification of acoustic emission signals generated by partial discharges is presented. Results obtained by acoustic emission method have been referred to results coming from other measuring methods.

In this paper, the properties of AE signals originating from phenomena occurring during magnetization of ferromagnetic materials which are used to construct power transformer cores are presented. The AE signals in a selected power oil transformer were recorded and analyzed. The analysis included, i.e., time, frequency, and time-frequency analyses, calculations of amplitude distributions of the signals and defined AE descriptors, determination of the descriptor map on the side walls of transformers, as well as a detailed analysis of selected part of the signals.The maps of descriptors were analyzed in the frequency bands of 20-70 kHz, 70-100 kHz, and 100-200 kHz. The analysis of the properties of the signals was performed in time and frequency domains. Based on the analysis, there were identified the AE signals originating from the phenomena occurring during the core magnetization of a power oil transformer. To identify those phenomena, the maps of the ADC descriptor calculated in the band of 20-70 kHz when selecting the measurement points in which there were no AE sources from partial discharges were used.An analysis of magnetoacoustic emission signals in the bands of 70-100 kHz and 100-200 kHz was also performed. The analysis of the signal properties in such an extended frequency range allowed determining the properties of the magnetoacoustic signals coming from core sheets of power oil transformers.Keywords: power oil transformers; magnetization of transformers cores; magnetoacoustic emission. Magnetic phenomena in ferromagnetic materialsPhenomena in ferromagnetic materials are one of the sources of acoustic signals in power oil transformers. During the magnetization of ferromagnetic materials which are used to construct transformer core laminations, there occurs the reversible movement of domain boundaries (the so-called Bloch walls), irreversible movement of domain boundaries (Barkhausen effect), and rotation of magnetization vectors in the direction of the external magnetic field. Within these processes, there occur processes of creation of domain boundaries, movement of these boundaries, and their annihilation.The Barkhausen jump results in (i) the generation of an electromagnetic wave (Barkhausen field effect -HEB), (ii) a local change of the magnetization and defects in the crystal lattice and (iii) the generation of an acoustic wave (magnetoacoustic emission effect).The phenomenon of magnetoacoustic emission (MAE) consists in generating a sound wave by a jump motion of a 90• domain wall in a material with nonzero magnetostriction (Augustyniak, 1999). A MAE signal depends not only on the microstructure, stresses or the magnetization frequency, but also on the geometry of the sample and its demagnetization. MAE has a relatively short history. It was first reported by A.E. Lord in the middle 1970 -s (Lord,1975. He called that effect "the acoustic Barkhausen effect" by analogy with "the classic Barkhausen effect". Elastic waves generated during MAE are caused by changes in magnetostrictive stresses during t...

In the paper, there are presented a general description of the constructed measuring system 8AE-PD, the results of calibration of this system by the Hsu-Nielsen method as well as the testing of the measuring system during recording signals generated by Hsu-Nielsen sources in a steel plate and a modeled partial discharge (PD) source.There is also presented the methodology of investigations by the calibrated acoustic emission method. The results of analyses of PD signals coming from the modeled sources are given. In particular, there are described the properties of acoustic emission (AE) signals generated by the PDs as well as the dependencies of the peak-to-peak voltage Umm, the RMS voltage Urms and the descriptors of acronyms ADP and ADC on the apparent charge Q introduced by the modeled PD source. There are determined the limit values of the apparent charge Q introduced by the modeled PD source for which the recorded signal is identified as originating from the PDS.

The original system useful for analysis of signals recorded during investigations of partial discharges within power oil transformers by means of acoustic method is presented. This method includes the basic and advanced analysis of recorded data. In the frame of basic analysis of data recorded signals undergo ltration in chosen frequency bands and next the analysis is made in domain of time, frequency, time-frequency and discrimination threshold. In the frame of advanced analysis of data the amplitude distributions of acoustic emission signals and the acoustic emission descriptors (dened by the authors) are calculated in order to outline maps of acoustic emission descriptors on lateral walls of a transformer; it is a base for location of sources of partial discharges by means of the original method consisted in determination of advance degree of a signal. Results of this analysis, for signals recorded in two chosen transformers with identical construction (partial discharge occurred only within one of them), are presented in the paper. The source of partial discharge, situated within oil near transformer tank, was localized; the revision conrmed this result. Properties of recorded emission acoustic signals at chosen measuring points situated on the tank, in function of distance between the partial discharge source and measuring points, are presented. The original system of analysis of acoustic emission (AE) signals includes the basic and advanced analysis of recorded data, dedicated to location and description of partial discharges (PD) within power oil transformers by means of AE method. The basic analysis of recorded signals is made by means of AE+JFTA program, written in LabView environment. The program contains the main level and detailed ones. In the rst order, data from the set created during registration process should be read in. The next step comes down to determination of data for analysis, by means of selection of measuring channel and length of analyzed signal. Afterwards, one should pass to detailed levels and to dene over there suitable parameters which control calculations. After determination of data and parameters the program is ready to realization of calculations. Calculations enable us to analyze signals in domain of time, frequency, time frequency and discrimination threshold. Example calculation results obtained from the program for one signal in one measuring channel is presented in Fig. 1. These results are composed of: signal after ltration (for ltration there is used a band-pass lter of 5. order) * corresponding author; e-mail: Franciszek.Witos@polsl.pl

No abstract