Defectoscopy

Defectoscopy, Elasticity and Acoustic Emission

Our research deals with the exact model of localization of acoustic emission (AE) sources on real complex solid bodies containing the holes/obstacles. The approach is based on numerical computations of precise geodesic curves on 3D vessels which can be composed from several parametrized surfaces with multiple intersections. Instead of solving the geodesic differential equations under specific Christoffel symbols, we compute our geodesics numerically by Finite Difference, Newton–Raphson, and Fixed Point Iteration methods. For faster calculations we require new technical improvements and optimizations. These techniques are then used in real experiments, such as steam reservoir.

We develop real-time algorithms which allow classification of Acoustic Emission (AE) signals into finite number of groups which correspond with the physical nature of AE sources. Classification can be based on spectral parameters and non-metrical distances, both independent on signal amplitudes. We would like to distinguish detected signals emitted by different suppressed acoustic emission sources in materials and also to separate them from a common noise. The models which estimate the underlying distribution of spectral parameters originate from the frame of multimodal probability density estimates, e.g. finite distribution mixtures or nonparametric kernel density estimates.

The Preisach-Mayergoyz model (PM) of hysteresis can be used to evaluate the mechanical properties of hysteretic materials. We develop algorithmic automatic identification of this PM space probability density within a class of distribution mixtures. The main tool is optimization stochastic algorithms (e.g. simulated annealing) using phi-divergence measures. Both, new index of elasticity and index of damage based on identified PM space of hysterons of hysteretic material is to be defined. Experimentally measured data of porcine skin and hysteretic dampers under cyclic loading are evaluated in cooperation with University of Granada.

Cooperation

 Members

Václav Kůs

assistant professor at FNSPE CTU in Prague

(divergence based methods in elastic defectoscopy, acoustic emission source localization, PM elasticity modelling)

Zuzana Dvořáková

Ph.D. student at FNSPE CTU in Prague

(acoustic emission signal classification)

Milan Chlada

IT, CAS

(defectoscopy methods, signal propagation in materials, neural networks)

Radovan Zeman

master student at FNSPE CTU in Prague

(image source method for description of dispersive elastic wave propagation in thin plates)

Erik Dolejš

master student at FNSPE CTU in Prague

(statistical techniques for PM space elasticity index evaluation, application to earthquake dampers)

Martin Kovanda

master student at FNSPE CTU in Prague

(deep learning methods for acoustic emission evaluation (CNN))

Jan Zavadil

bachelor student at FNSPE CTU in Prague

(wise SDDT/MBC/KDE classification of AE signals and other ML methods)

 

Graduates

Petr Gális

(numerical optimization of geodetics and localization maps in acoustic emission)

Colette Kožená

(PM based evaluation of material elasticity/damage under hysteresis)

Michal Záveský

(geodetic iterative acoustic source localization; statistical cluster separations in acoustics)

Jan Tláskal

(statistical classification methods in acoustic emission)

Šárka Černá (Vejvodová)

(nonlinear ultrasonic tomography with time-reversed signals)

Jana Habásková (Papoušková)

(new families of statistical distributions in Preisach-Mayergoyz hysteron space)

Petr Svojítka

(geodetic localization algorithms on compound surfaces)

Bohumil Pospíšil

(micro-crack localization based on acoustic emission signal detection)

Martin Petřík

(micro-crack localization based on acoustic emission signal detection)

Jan Máca

(application of statistical methods on detection of the change point in acoustic signals)

Significant papers

  • Kůs V. and Dolejš E. Elasticity index evaluation based on Le Camdivergence and kernel density estimator in PM space. Journal of Physics: Conf Ser, 2021, Volume 1730, 012008, DOI: 10.1088/1742-6596/1730/1/012008
  • Kožená, C.; Kůs, V.; Galego, A.; Benavent-Climent, A., Damage assessment of earthquake dampers based on Preisach-Mayergoyz model . Biennial Baltic Electronics Conference (BEC). USA: IEEE Computer Society, 2018. p. 1-4. ISBN 978-1-5386-7311-9. DOI: 10.1109/BEC.2018.8600971
  • Gális, P.; Kůs, V.; Chlada, M., Acoustic emission source localization based on exact geodesic curves (with application to watering can). In: Application Of Contemporary Non-Destructive Testing In Engineering – Conf. Proceedings. Ljubljana: The Slovenian Society for Non-Destructive Testing, 2017. p. 313-319. ISBN 978-961-93537-3-8. SCOPUS: 2-s2.0-85049240591
  • Kožená, C., Kůs, V., Dos Santos, S., Hysteresis and memory effects in skin aging using PM space density identification. Biennial Baltic Electronics Conference (BEC), USA: IEEE Computer Society, 2016, p.179-182. ISSN 1736-3705, ISBN 978-1-5090-1393-7. DOI: 10.1109/BEC.2016.7743758
  • Kober, J.; Dvořáková, Z.; Převorovský, Z.; Krofta, J., Time reversal transfer: Exploring the robustness of time reversed acoustics in media with geometry perturbations. JASA, Journal of the Acoustical Society of America, 2015, 138(1), EL49-EL53. ISSN 0001-4966. DOI: 10.1121/1.4922623
  • Dvořáková, Z.; Dos Santos, S.; Převorovský, Z.; Kůs, V., Spatial localization and statistical classification of nonlinear ultrasonic signature in bio-mechanical medium with time reversal based news methods, International Congress on Ultrasonics. Singapore: Research Publishing Services, 2013. pp. 796-801. ISBN 978-981-07-5938-4. DOI: 10.3850/978-981-07-5938-4_P0426

Bachelor and Diploma projects for students