[Back]


Talks and Poster Presentations (with Proceedings-Entry):

G. Shilyashki, H. Pfützner, P. Hamberger, M. Aigner, A. Kenov, I. Matkovic:
"Spatial Distribution of Magnetostriction, Displacements and Noise Generation of Model Transformer Cores";
Talk: International Conference on Engineering Vibration, Ljubljana, Slovenia; 2015-09-07 - 2015-09-10; in: "ICoEV Book of Abstracts", (2015), ISBN: 978-961-6536-96-7; 106.



English abstract:
Recently, the relevance of audible noise of power transformers tends to increase due to the growing environmental awareness. for Sound assessment, two standardised methods reveal the global noise of the whole system, as resulting from the interaction of core, windings, oil and tank. But for a deeper understanding of noise generation, it would be advantageous to investigate also the individual roles of the single above components in closer ways. This paper summarizes attempts to study the first component, i.e. the soft magnetic core, by means for model cores. For the first time, we analyzed local distributions of all three strain, displacement and audible noise, keeping in mind that these quantities depend on many parameters like material, stacking, clamping, induction, rotational magnetization of additional DC-bias, in complex ways. In-plane strain proved to be dominated by magnetostriction (MS), with maximum intensities in corners and T-Joints. The interpretation of in-plane displacements proves to be complicated by the unknown resting point of the whole system. However,the results reflect contributions of both MS and magneto-static forces (MF). Out-of plane displacements proved to be dominated by effects of MF, in particular at overlaps of corners and T-Joints, due to imperfect clamping. Regional measurements of audible noise were performed in the near-field mode by means of automatic scanning by microphone over free core regions within a noise-isolating scanning chamber. As to be expected, the results showed strongly inhomogeneous distributions with maxima at T-Joints and corners. As a conclusion, model core results have very restricted relevance for full sized cores in quantitative ways. But they favour an understanding for crucial mechanisms and for core regions that play dominating roles.

Created from the Publication Database of the Vienna University of Technology.