Attenuation of Compressional Waves in Peridotite Measured as a Function of Temperature at 200 MPa

Abstract

Abstract--A technique has been developed to determine attenuation in rocks at high temperature using a gas-media, high-pressure apparatus. A pulse transmission technique and a spectral ratio method are used to study compressional seismic properties of rocks. Seismic waves are transmitted to and from the sample through buffer rods of mullite. The effect of seismic wave reflections within the sample assembly are cancelled out by taking ratios of the spectra measured at different temperatures. In order to obtain good signal-to-noise ratio for resolving the attenuation at high pressure and temperature, special care is taken in the sample assembly and the ultrasonic coupling between the sample, buffer rods and transducers. A very tight connection of the sample-buffer rod-transducer is essential for obtaining high frequency signals ( > 300 kHz) at high temperature. A small mass is attached to each outside end of the transducer to drive low frequency signals ( < 250 kHz) into the sample. Before attenu~:tion measurements, the sample and the buffer rods are tightly compacted in a platinum tube at high pressure and room temperature to ensure pressure seal of the sample assembly. The frequency range of measurement covers 50 to 450 kHz for the sample. Attenuation is very small in the buffer rod compared to the sample for the entire temperature range of the study. Because of the small attenuation, a wide frequency band of 50 kHz to 3.2 MHz can be covered for investigating the attenuation in the buffer rod. The technique has been used to measure attenuation at high confining pressure, and temperatures including sub-and hyper-solidus of upper mantle rocks. Therefore, effects of partial melting on attenuation can be studied.The method is applied to the attenuation measurement in a peridotite as a function of temperature to 1225~ at 200 MPa confining pressure. At high temperature, signal amplitude decays more rapidly at high frequency than at low frequency, from which attenuation (and Q) can be determined using a spectral ratio method. No frequency dependence of Q is resolved for both the sample and the buffer rod over the entire temperature and frequency ranges of the measurement. The results show that Q decreases rapidly with increasing temperature even in the temperature range below the solidus of peridotites. Such temperature sensitivity of Q is probably more useful to probe thermal structure in the upper mantle than that of conductivity at temperatures below the solidus. The results in this study are compared with available seismic velocity, electrical conductivity and solidus data for peridotites, suggesting that there is no discontinuous change in both mechanical and electrical properties of peridotites at the solidus temperature. PAGEOPH,Even at hypersolidus temperatures, it appears that velocity drops and conductivity increases continuously (not abruptly) with increasing melt fraction. This implies that mechanical and electrical properties of the upper mantle will gradually change at the boundary where the geotherm cross...

The seismic attenuation structure of the East Pacific Rise

“…These values seem incompatible with the results of Sato et al [1988. The results obtained for peridotite at 0.2 GPa, a pressure that corresponds very closely to that at the Moho, predict Qp -6 at the solidus and suggest that the values obtained for the longest range paths require temperatures 200-300* below the solidus.…”

“…Obtaining measurements that are valid for both the compositional, pressure, and temperature conditions in the Earth and for the frequency, scale, and strain rates of seismic observations is a formidable task [e.g., Jackson, 1986;Christensen and Wepfer, 1989]. A number of measurements have been carried out on mafic and ultramafic samples at elevated temperatures [Woirguard and Gueguen, 1978;Berckhemer et al, 1979;Sacks and Murase, 1983;Kampfmann and Berckhemer, 1985;Sato and Manghnani, 1985;Manghnani et al, 1986;Sato et al, 1988Gueguen et al, 1989;Jackson et al, 1992]. The majority of these works were motivated by a desire to understand the properties of the asthenosphere and are thus primarily devoted to ultramafic samples.…”

“…Moreover, since the minimum depth of our observations of the axial low-Q zone is about 2 km, it is not clear how applicable room pressure measurements are to our study. Sato et al [1988 present Qp measurements for peridotite as a function of temperature obtained at pressures of 0.20, 0.48 and 0.72 GPa. The values were obtained by applying a spectral ratio method to pulse transmission measurements at 60-880 kHz, frequencies that are several orders of magnitude above the seismic band.…”

Seismic and magnetic constraints on the structure of upper oceanic crust at fast and slow spreading ridges

“…Shear wave velocities outside the sill were assumed to be equal to Vp/V3" (i.e., surrounding rock was modeled as a Poisson solid). Sato et al, 1988Sato et al, , 1989 to 10 [ Murase and McBirney, 1973;Sato and Manghnani, 1985;Manghnani et al, 1986], but these measurements were made at high frequencies (>100 kHz) and may overpredict attenuation of lower-frequency (10 Hz) seismic energy. Wilcock et al…”

Laboratory Seismic Studies: A Search for Mechanical, Thermal and Chemical Structures of the Earth.

“…By means of a self-consistent correction method (Yoneda, 1990) 43, No. 4, 1995 (to 1,280•Ž) using a gas-media high-pressure apparatus (to 0.73GPa) (Sato et al, 1988b(Sato et al, , 1989b (Kjartansson, 1979;Lienert and Manghnani, 1990). The model of grain boundary relaxation suggested that an appropriate choice of grain size may give an ultrasonic Q that is applicable to the earth (Sato et al, 1989b).…”