Flume experiments: acoustic beam properties effect on the Doppler velocity measured in a moving sand-bed

The remote sensing of bed load by means of Acoustic Doppler Current Profiler (ADCP) in a river was first introduced by Rennie et al. 2002. The bottom tracking capability is applied to measure bed-load velocity and therefore to estimate sediment transport moving over the bed, i.e., the bed-load. The bias introduced in the Doppler Effect by moving sediment near the bed is measured, that is the apparent velocity of the bed sediments. Despite the promising goal of developing a non-intrusive method for bed-load characterization, many instrument dependent parameters with respect to in situ characteristics may affect the relation between the measurable apparent velocity of the bed and bed-load. Aiming to elucidate these effects, a set of experiments were conducted in a laboratory flume to measure the Doppler Effect from a moving sand bed by means of the acoustic device DOP2000 by Signal Processing S.A. The experiments were performed by using three different sets of probes: 4 MHz, 1 MHz and 0.5 MHz, deployed to project out an acoustic beam at an angle of 25°, 20° and 15° from the vertical and 90 mm from the bed, and the profile resolution changing from 1.13 to 7.5 mm which corresponds to a changing pulse length. The 0.8-m-width, 18-m-length and 1-m-depth horizontal flume is situated at the Hydraulic Engineering Laboratory (LIDR) of the University of Bologna. Two pumps recirculated a constant discharge of water and sediments of about 50 l/s each as maximum, by using two recirculation pipes provided by flow meters. The mean velocity measured in the channel by ADV was around 0.3 m/s. The mobile bed, composed by sand with mean diameter of about 110 μm, had a mean thickness of 10 cm and a total length of about 5 m. The bed forms development was monitored from side windows by means of a camera. The probes of the DOP2000 were deployed in correspondence of these windows. Concurrent measurements were performed by using 8 different probes, in two arrays of four aligned with the flume. The first tests used 1 and 4 MHz probe arrays. A second series of tests applied 0.5 MHz and 1 MHz probes. All the tests were performed in two different hydraulics conditions, with one (53-54 l/s) or two (95-97 l/s) pumps running, but maintaining the same water depth of about 0.30 m by means of a downstream weir. A preliminary analysis of the measured echoes near the bed shows the signal dependence form applied frequencies, deployment angles and profile resolutions. Aiming to parameterize the influence of ADCP features (i.e., frequency, beam angle, pulse length) on the apparent velocity with respect to in situ characteristics, the measured Doppler Effect near the flume bed will be correlated to probe parameters change. Moreover, the average velocity of the bed will be assessed by analyzing the images from the applied camera at flume side to be compared to the measured Doppler Effect.