Publication detail

Spectral Analysis of EME Signal

Jiri Majzner, Josef Sikula, Petr Sedlak and Yasuhiko Mori

Original Title

Spectral Analysis of EME Signal

English Title

Spectral Analysis of EME Signal

Type

conference paper

Language

en

Original Abstract

The new measuring method for detection of small cracks fine spectra of electromagnetic and acoustic emission (EME and AE) signals is described. It requires wide band ultra-low noise amplifiers, analogy filters, optimization of signal to noise ratio of sensors and application of noise elimination methods. The analyses of noise sources in sensors and preamplifiers are given. They are thermal noise, polarization noise and low frequency 1/f noise. Measuring set-up background noise suppression including electromagnetic shielding allows us to detect signals in the range of 100 nV. This measuring set-up was used to observe crack creation in granite samples. AE and EME signals show different behaviour in the first interval of about 10 s just after crack creation. In frequency domain EME spectra show that in the first stage of crack creation the spectrum is given by crack walls eigen vibration, internal friction and sample electrical conductivity. We observed that crack opening and crack wall vibration create high frequency signal in the frequency band up to 10 MHz. These signals were observed in time period of 10 s. In the second period for time interval 10 to 300 s the frequency spectrum is given by sample eigenvibration or sample boundary conditions and we observed spectra in the frequency range 100 kHz to 2 MHz. In this second period EME signal amplitude depends on sample electrical conductivity and with increasing conductivity decreases.

English abstract

The new measuring method for detection of small cracks fine spectra of electromagnetic and acoustic emission (EME and AE) signals is described. It requires wide band ultra-low noise amplifiers, analogy filters, optimization of signal to noise ratio of sensors and application of noise elimination methods. The analyses of noise sources in sensors and preamplifiers are given. They are thermal noise, polarization noise and low frequency 1/f noise. Measuring set-up background noise suppression including electromagnetic shielding allows us to detect signals in the range of 100 nV. This measuring set-up was used to observe crack creation in granite samples. AE and EME signals show different behaviour in the first interval of about 10 s just after crack creation. In frequency domain EME spectra show that in the first stage of crack creation the spectrum is given by crack walls eigen vibration, internal friction and sample electrical conductivity. We observed that crack opening and crack wall vibration create high frequency signal in the frequency band up to 10 MHz. These signals were observed in time period of 10 s. In the second period for time interval 10 to 300 s the frequency spectrum is given by sample eigenvibration or sample boundary conditions and we observed spectra in the frequency range 100 kHz to 2 MHz. In this second period EME signal amplitude depends on sample electrical conductivity and with increasing conductivity decreases.

Keywords

Acoustic emission, electromagnetic emission, cracks

RIV year

2006

Released

21.05.2006

Publisher

MIDEM

Location

Ljubljana, Slovenia

ISBN

961-91023-4-7

Book

Proceedings / EMPS 2006 - 4th European Microelectronics and Packaging Symposium with Table-Top Exhibition

Pages from

315

Pages to

318

Pages count

4

BibTex


@inproceedings{BUT19063,
  author="Jiří {Majzner} and Josef {Šikula} and Petr {Sedlák} and Yasuhiko {Mori}",
  title="Spectral Analysis of EME Signal",
  annote="The new measuring method for detection of small cracks fine spectra of electromagnetic and acoustic emission (EME and AE) signals is described.  It requires wide band ultra-low noise amplifiers, analogy filters, optimization of signal to noise ratio of sensors and application of noise elimination methods. The analyses of noise sources in sensors and preamplifiers are given. They are thermal noise, polarization noise and low frequency 1/f noise. Measuring set-up background noise suppression including electromagnetic shielding allows us to detect signals in the range of 100 nV. This measuring set-up was used to observe crack creation in granite samples. AE and EME signals show different behaviour in the first interval of about 10 s just after crack creation. In frequency domain EME spectra show that in the first stage of crack creation the spectrum is given by crack walls eigen vibration, internal friction and sample electrical conductivity. We observed that crack opening and crack wall vibration create high frequency signal in the frequency band up to 10 MHz. These signals were observed in time period of 10 s. In the second period for time interval 10 to 300 s the frequency spectrum is given by sample eigenvibration or sample boundary conditions and we observed spectra in the frequency range 100 kHz to 2 MHz. In this second period EME signal amplitude depends on sample electrical conductivity and with increasing conductivity decreases.",
  address="MIDEM",
  booktitle="Proceedings / EMPS 2006 - 4th European Microelectronics and Packaging Symposium with Table-Top Exhibition",
  chapter="19063",
  institution="MIDEM",
  year="2006",
  month="may",
  pages="315",
  publisher="MIDEM",
  type="conference paper"
}