ABSTRACT
In this paper, the study of frequency-dependent ultrasonic attenuation in strongly heterogeneous cementitious materials is addressed. To accurately determine the attenuation over a wide frequency range, it is necessary to have suitable excitation techniques. We have analysed two kinds of ultrasound techniques: contact ultrasound and airborne non-contact ultrasound. The mathematical formulation for frequency-dependent attenuation has been established and it has been revealed that each technique may achieve similar results but requires specific different calibration processes. In particular, the airborne non-contact technique suffers high attenuation due to energy losses at the air-material interfaces. Thus, its bandwidth is limited to low frequencies but it does not require physical contact between transducer and specimen. In contrast, the classical contact technique can manage higher frequencies but the measurement depends on the pressure between the transducer and the specimen. Cement specimens have been tested with both techniques and frequency attenuation dependence has been estimated. Similar results were achieved at overlapping bandwidth and it has been demonstrated that the airborne non-contact ultrasound technique could be a viable alternative to the classical contact technique.
ABSTRACT
Ultrasonic gas flowmeters typically use narrowband piezoelectric transducer arrangements for interrogating the flow of gas in a pipe. In this work, the suitability of broadband electrostatic transducers operating at frequencies of up to 1 MHz for ultrasonic measurement of gas flow has been investigated. The transit time method of ultrasonic gas flow measurement was adopted and experiments were carried out using a laboratory test rig capable of producing a range of gas flowrates up to 17.5 m/s. The test rig also allowed easy interchange of different prototype flowmetering sections. Times of flight of ultrasonic waves interrogating the gas flow were measured using separate send/receive electrostatic transducer arrangements. Two flowmeter configurations were considered. The first interrogated the flow at 45 degrees in contra-propagating upstream and downstream directions. The second consisted of an up-stream interrogation at 45 degrees to the gas flow and an interrogation made normal to the flow direction. k factors correlating the fluid velocity along the ultrasonic path with the mean fluid velocity in the pipe were calculated using experimental ultrasonic data and anemometer measurements. All transducer configurations were numerically modelled using the computational fluid dynamics software package FLOTRAN (ANSYS Inc.). Theoretical gas flow velocities for both transducer arrangements were subsequently compared with experimental values and found to be in excellent agreement. A flow-dependent frequency shift of the received ultrasonic signals was also observed simultaneously with the transit time measurement.
