Estimation of uniaxial compressive strength and intrinsic permeability from ultrasounds in sedimentary stones used as heritage building materials

• Main Authors: Benavente, D. (Author), de Jongh, M. (Author), Galiana-Merino, J.J (Author), Garcia-Martinez, N. (Author), Martinez-Martinez, J. (Author), Pla, C. (Author)
• Format: Article
• Language: English
• Published: Elsevier Masson s.r.l. 2022
• Subjects: Elastic waves; Non-destructive test; Petrophysics; Stone conservation; Ultrasonic pulse velocity
• Online Access: View Fulltext in Publisher

 This paper combines ultrasonic parameters recorded from P- and S- waves and effective porosity to estimate uniaxial compressive strength, σC, and intrinsic permeability, k - two of the most crucial physical parameters used in stone conservation studies. The ultrasonic parameters have been chosen because they are sensitive to the microstructural properties of stones which influence both σC and k differently. Velocities of compressional or primary, Vp, and shear or secondary, Vs, elastic waves are related to porosity and the mineralogical composition; the waveform attenuation is related to the presence of vuggy macropores and pore and particle size; and the wavelength is related to the pore and particle size. Fifteen types of stones used in the Spanish built heritage were measured including calcarenites, sandstones, dolostones and travertines. Several simple and multiple predictive expressions for σC and k are fitted, which are quantified through the Pearson correlation coefficient, R. Curve fitting improves as the number of petrophysical parameters taken into account in the multiple linear regressions increases. The obtained Pearson's correlations values for σC are slightly higher than for k. The best correlation for σC (R = 0.921) and k (R = 0.903) is found when the predictive equation incorporates all of the elastic wave parameters obtained non-destructively, and the connected porosity. In general, effective porosity is statistically the most significant variable of the fitted regressions. The σC and k estimation from all the ultrasonic parameters, without considering connected porosity, shows a good correlation (R = 0.852 and 0.864, respectively). The addition of Vs to multiple regressions does not significantly improve the goodness of the fit. Consequently, the measurement of Vs is unnecessary as it is difficult and time-consuming. Finally, results also show that vuggy porosity (present in travertines) influences waveform attenuation and σC and k differently, relative to other microstructural properties; limiting the techniques use for stones with this type of porosity. Estimating σC and k using ultrasonic parameters exclusively is advantageous as the estimations can be obtained non-destructively using portable equipment, while also providing an acceptable goodness of fit. © 2022