Developing an Ultrasound Probe Array for a High-Pressure, High-Temperature Solid Medium Deformation Apparatus

TitleDeveloping an Ultrasound Probe Array for a High-Pressure, High-Temperature Solid Medium Deformation Apparatus
Publication TypeConference Proceedings
Year of Conference2019
AuthorsGhaffari, HO, Mok, U, Pec, M
Conference NameAGU Fall Meeting 2019
Conference LocationSan Francisco, CA

Deformation of rocks is achieved by collective movement of various defects. Active and passive ultrasound probes are potential tools to examine the nature of these defects in materials under high pressure and temperature conditions. Here, we report on the development of an ultrasound probe array for a solid-medium deformation apparatus capable of deforming samples up to pressures, P ≈ 3 GPa and temperatures, T ≈ 1200˚C. We utilize 4 broad band needle piezo-electric elements; 3 below the sample and one above to allow for event localization. Proper grounding and electrical insulation of the sensors tremendously reduced EMI noises. We demonstrate that the system is capable of recording excitation of acoustic waves from 50 kHz to 5 MHz at sample temperatures up to 750˚C and confining pressure up to 1.5GPa. To validate our setup and further investigations on the transition from flow to fracture, we deformed natural quartzite samples at P = 0.5 - 1GPa and T = 150 – 750˚C and strain rates of 10-7 to 10-5 s-1. We monitored passive excitations of ultrasound waves due to nucleation and propagation of various defects. Collective behavior of recorded acoustic emissions show that increasing temperature from 150˚C to an intermediate temperature of 375˚C results in a decrease of the probability of finding events with larger magnitudes. Statistically, this indicates that increasing temperature correlates with weaker events compatible with a thermal activation model of micro-failures: increasing temperature raises the thermal energy of the system and decreases the yield strength of the material. Furthermore, decreasing the deformation rate to ~10-7 s-1 and 400’c, the number of AEs decreases dramatically in a way that clipped events disappear and the distribution of events shifts to a Gaussian distribution. Recorded excitations can be classified into at least two main categories: the majority of events in higher strain rate experiments carry high frequency components (>500kHz-2MHz), while the majority of events from lower strain rate experiments coincide with low frequency events with relatively long duration and frequency range of 80-300kHz. Interestingly, we find that the rising time of high frequency events are a factor of 4-5 shorter than long period events, indicating a significantly slower dynamic deformation process.