TRANSDUCER AND BOND PHASE SHIFTS IN ULTRASONICS, AND THEIR EFFECTS ON MEASURED PRESSURE DERIVATIVES OF ELASTIC MODULI

Abstract

Phase shifts introduced into ultrasonic signals by the presence of transducers and bonds at sample surfaces have been measured using an automated variable frequency ultrasonic interferometer. At zero pressure, phase shifts have been resolved due to transducers, bonds between transducers and samples, and bonds between buffer rods and samples. Observed transducer-bond-sample phase shifts are in good accord with theoretical estimates, and bond thicknesses of about 0.3 μ are inferred. Measurements to 7 kbar are consistent with theoretical estimates of the effect of pressure on transducer-bond-phase shifts. Providing the frequency of the ultrasonic signal is within a few percent of the resonance frequency of the transducer, and the effect of pressure on the transducer resonance frequency is accounted for (as recommended by McSkimin, [1961]), the effect of the bond phase shift on the measured pressure derivative of the elastic modulus should amount to less than about 0.02. If the frequency deviates substantially from the transducer resonance frequency, especially at zero pressure, errors of the order of 0.25 could be incurred in the pressure derivatives. The nonlinearity of transducer-bond phase shifts could cause significant errors in second-pressure derivatives, even under favorable conditions. For shear waves at zero pressure, the observed buffer-bond-sample phase shifts are consistent with those estimated theoretically for a bond of about 1 μ thickness. For compressional waves at zero pressure, phase shifts are very sensitive to the buffer-sample contact: large differences in phase are observed between dry lapped, “wetted” immersed, and resin-bonded contacts. The sources of these differences are not fully understood, but they may be due to variations in contact area produced by the ultrasonic wave. “Normal” buffer-sample bonds are estimated to be capable of affecting measured pressure derivatives by about 0.25. The behavior of the anomalous buffer-sample phase shifts under pressure is unknown, but the shifts could easily give rise to substantial errors in measured pressure derivatives.