Statistical characterisation of metals from ultrasonic aberrations

J. A. Hernández, M. Clark, I. J. Collison, Michael Geoffrey Somekh, S. D. Sharples

Research output: Chapter in book / Conference proceedingConference article published in proceeding or bookAcademic researchpeer-review


Materials with random microstructure can have an adverse effect on ultrasonic measurements through scattering and aberration of the acoustic field. We have developed a theoretical and experimental technique for characterising the effect of the material microstructure on the propagation of acoustic waves. The theoretical method predicts propagation of the statistical properties in the random medium. Using this model and a novel experimental technique we can extract materials properties from observations of acoustic aberrations. The microstructure has been modelled as having grains randomly orientated and weakly anisotropic. Moreover, individuals grains are treated to be equiaxed. This is summarised by assuming that the correlation function for the wave number has a Gaussian shape. Under this assumption, the approximated power correlation function for the acoustic field has been obtained using the stochastic wave equation for random media, along with numerical simulation using phase screen theory. The experimental evidence that aberrations are frequency dependent is presented. Multiple cscans on titanium were performed using a 10MHz transducer as an ultrasonic source. The output of the transducer gives substantial frequency components between 6MHz and 16MHz. The variation of the frequency on a fixed sample enables us to examine different measurement regimes (higher frequencies correspond to large grains). The statistical analysis and estimated power correlation function from measurements are compared to a modelled power correlation function. The correlation length and standard deviation of the wave number define the power correlation function of the field. This function has been fitted to the estimated power correlation from measurements on titanium 6-4, and values for the variance and correlation length were obtained. The paper demonstrate that the stochastic model is capable of quantitative prediction of the predominant wave scattering effect in granular materials.
Original languageEnglish
Title of host publication2005 IEEE Ultrasonics Symposium
Number of pages4
Publication statusPublished - 1 Dec 2005
Externally publishedYes
Event2005 IEEE Ultrasonics Symposium - Rotterdam, Netherlands
Duration: 18 Sept 200521 Sept 2005


Conference2005 IEEE Ultrasonics Symposium

ASJC Scopus subject areas

  • Acoustics and Ultrasonics


Dive into the research topics of 'Statistical characterisation of metals from ultrasonic aberrations'. Together they form a unique fingerprint.

Cite this