Author(s): Pasupuleti Laxmi Narayana, Praful Vasharambhai Timbadiya, Prem Lai Patel
Linked Author(s): P. V. Timbadiya
Keywords: Nyquist frequency, noises, spikes, low pass butterworth filter, theoretical spectra
Abstract: Post processing of the data, collected using Ultrasonic Ranging Systems (URS) and Acoustic Doppler velocimeter (ADV), is prerequisite before using it for further analysis of bed elevations and flow velocity respectively in alluvial channels. The data acquired from such systems include undesirable high frequency contents (noises), spikes, affecting the measurement accuracy of aforesaid variables. The present study aims to demonstrate the step wise procedure for post processing of URS and ADV data collected for measurement of bed level variation and three dimensional velocities (3D), respectively. The bed level and 3D velocity data are collected in a 15 m long recirculated tilting sediment transport flume at Advanced Hydraulics Laboratory, SVNIT-Surat, India. The channel is having 0. 89 m width, 0. 60 m depth and flow discharge is kept 44 liter per second. Here, low pass Butterworth filter is employed to remove the noises in the above mentioned variables. The low pass filter at Nyquist frequency does not create any effect on the main signal. The results reveal that computed spectra for three dimensional velocities from ADV data is nearer to theoretical spectra following -5/3 law. Also, the employed filter effectively removes all noises below Nyquist frequencies for URS data. In case of URS data, the maximum and minimum spike counts are 30% and 2%, respectively. It is observed that increase in the distance between blanking and maximum range reduces the spike count
Year: 2017