11:15 AM - 11:30 AM
[SSS09-06] Influence of a void in a cylindrical cement mortar sample on transmitted elastic waves
Keywords:Void, Scattering, Elastic waves, Coda waves, Experiments, Heterogeneity
In general, rocks constituting the earth's crust are composed of multiple mineral species, hence heterogeneous. Seismic waves propagating through heterogeneous media are strongly scattered and the wave field is perturbated. Nishizawa (2005) performed elastic wave transmission experiments on granite samples with different characteristic scales of heterogeneity and found that the characteristics of the wave field depended on the heterogeneity scale. Porous media such as subsurface ground have stronger heterogeneity due to a large contrast between the acoustic impedance of minerals and pores, which is unable to apply well-known approximate descriptions for the propagation function of elastic waves in weakly heterogeneous media (e.g., Sato and Fehler, 1998). Kawakata (2023, 2024) performed elastic wave transmission experiments on a natural gabbro rock sample (weakly heterogeneous sample) and a cylindrical cement mortar sample including hollow polypropylene balls (PP balls) as voids (strongly heterogeneous sample) to investigate the propagation characteristics of elastic waves in strongly heterogeneous media. However, it was too difficult to understand the propagation characteristics due to the complexity of the behavior of the waves and the insufficient signal intensity of recorded waveforms. In this study, using a broad-band piezoelectric transducer to improve the signal intensity, we investigated the effects of a single void on the propagation characteristics of transmitted waves as a simpler setting. We performed elastic wave transmission experiments on a cylindrical cement mortar sample with a single void (S-PP) and one without a void (S-Intact) and investigated the difference in the propagation characteristics.
We mixed powder cement mortar and water at a mass ratio of 40:7 (hereafter referred to as cement mortar). For S-PP, we poured the cement mortar until about half the height of the mold and placed a hollow PP ball with a diameter of 9.5 mm as a void in the center of the mold, and then left it for about 12 hours. Cement mortar samples tend to be contaminated with air bubbles (Inanishi et al., 2022), which may cause unexpected scattered waves. We applied vibration using a speaker while the cement mortar was hardening. We poured the cement mortar up to the top of the mold and left it for more than 2 days with applying vibration. For S-Intact, we poured the cement mortar into the mold up to the top of it and left it for more than 2 days with applying vibration. The samples were 56 mm in diameter and ~120 mm in height. We attached a piezoelectric transmitter and a broad-band piezoelectric transducer to the opposite ends of the sample each other. We repeatedly applied high-voltage pulses of -100 V with 1 μs width to the transmitter using a pulse generator. We recorded waveforms at a sampling rate of 10 Msps and stacked them together 2500 times.
Figure 1 shows the envelopes and spectrograms for the stacked waveforms. The envelope for S-PP decayed linearly, whereas that of S-Intact decayed exponentially. For the spectrograms, the direct and coda wave attenuation tended to be larger for S-PP at higher frequencies than 0.04 MHz. At frequencies lower than 0.04 MHz, the coda waves were more dominant in S-PP. Considering the P-wave velocity (~3 km/s) estimated from the recorded waveforms, the wavelength became 75 mm at 0.04 MHz, which was about 8 times as large as the diameter of the PP ball. That may cause backscattering and the direct wave and coda wave strongly attenuated at frequencies higher than 0.04 MHz, whereas forward scattering became dominant and generated the coda waves nearly immediately after the arrival of the direct wave at frequencies lower than 0.04 MHz. In the future, we will investigate the causes of the difference in the behavior of scattered waves in detail.
Acknowledgments: This work was supported by JSPS KAKENHI Grant Number JP22H01336.
We mixed powder cement mortar and water at a mass ratio of 40:7 (hereafter referred to as cement mortar). For S-PP, we poured the cement mortar until about half the height of the mold and placed a hollow PP ball with a diameter of 9.5 mm as a void in the center of the mold, and then left it for about 12 hours. Cement mortar samples tend to be contaminated with air bubbles (Inanishi et al., 2022), which may cause unexpected scattered waves. We applied vibration using a speaker while the cement mortar was hardening. We poured the cement mortar up to the top of the mold and left it for more than 2 days with applying vibration. For S-Intact, we poured the cement mortar into the mold up to the top of it and left it for more than 2 days with applying vibration. The samples were 56 mm in diameter and ~120 mm in height. We attached a piezoelectric transmitter and a broad-band piezoelectric transducer to the opposite ends of the sample each other. We repeatedly applied high-voltage pulses of -100 V with 1 μs width to the transmitter using a pulse generator. We recorded waveforms at a sampling rate of 10 Msps and stacked them together 2500 times.
Figure 1 shows the envelopes and spectrograms for the stacked waveforms. The envelope for S-PP decayed linearly, whereas that of S-Intact decayed exponentially. For the spectrograms, the direct and coda wave attenuation tended to be larger for S-PP at higher frequencies than 0.04 MHz. At frequencies lower than 0.04 MHz, the coda waves were more dominant in S-PP. Considering the P-wave velocity (~3 km/s) estimated from the recorded waveforms, the wavelength became 75 mm at 0.04 MHz, which was about 8 times as large as the diameter of the PP ball. That may cause backscattering and the direct wave and coda wave strongly attenuated at frequencies higher than 0.04 MHz, whereas forward scattering became dominant and generated the coda waves nearly immediately after the arrival of the direct wave at frequencies lower than 0.04 MHz. In the future, we will investigate the causes of the difference in the behavior of scattered waves in detail.
Acknowledgments: This work was supported by JSPS KAKENHI Grant Number JP22H01336.