skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Noise correction of turbulent spectra obtained from Acoustic Doppler Velocimeters

Abstract

Accurately estimated auto-spectral density functions are essential for characterization of turbulent flows, and they also have applications in computational fluid dynamics modeling, site and inflow characterization for hydrokinetic turbines, and inflow turbulence generation. The Acoustic Doppler Velocimeter (ADV) provides single-point temporally resolved data, that are used to characterize turbulent flows in rivers, seas, and oceans. However, ADV data are susceptible to contamination from various sources, including instrument noise, which is the intrinsic limit to the accuracy of acoustic velocity measurements. Due to the presence of instrument noise, the spectra obtained are altered at high frequencies. The focus of this study is to develop a robust and effective method for accurately estimating auto-spectral density functions from ADV data by reducing or removing the spectral contribution derived from instrument noise. For this purpose, the “Noise Auto-Correlation” (NAC) approach was developed, which exploits the correlation properties of instrument noise to identify and remove its contribution from spectra. The spectra estimated using the NAC approach exhibit increased fidelity and a slope of -5/3 in the inertial range, which is typically observed for turbulent flows. Finally, this study also compares the effectiveness of low-pass Gaussian filters in removing instrument noise with that of the NACmore » approach. For the data used in this study, both the NAC and Gaussian filter approaches are observed to be capable of removing instrument noise at higher frequencies from the spectra. However, the NAC results are closer to the expected frequency power of -5/3 in the inertial sub-range.« less

Authors:
; ; ;
Publication Date:
Research Org.:
Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
1129300
Report Number(s):
PNNL-SA-79811
WC0100000
DOE Contract Number:
AC05-76RL01830
Resource Type:
Journal Article
Resource Relation:
Journal Name: Flow Measurement and Instrumentation, 37:29-41
Country of Publication:
United States
Language:
English
Subject:
turbulence; noise; Acoustic Doppler Velocimeter; marine hydrokinetic energy; renewable energy

Citation Formats

Durgesh, Vibhav, Thomson, Jim, Richmond, Marshall C., and Polagye, Brian. Noise correction of turbulent spectra obtained from Acoustic Doppler Velocimeters. United States: N. p., 2014. Web. doi:10.1016/j.flowmeasinst.2014.03.001.
Durgesh, Vibhav, Thomson, Jim, Richmond, Marshall C., & Polagye, Brian. Noise correction of turbulent spectra obtained from Acoustic Doppler Velocimeters. United States. doi:10.1016/j.flowmeasinst.2014.03.001.
Durgesh, Vibhav, Thomson, Jim, Richmond, Marshall C., and Polagye, Brian. Sun . "Noise correction of turbulent spectra obtained from Acoustic Doppler Velocimeters". United States. doi:10.1016/j.flowmeasinst.2014.03.001.
@article{osti_1129300,
title = {Noise correction of turbulent spectra obtained from Acoustic Doppler Velocimeters},
author = {Durgesh, Vibhav and Thomson, Jim and Richmond, Marshall C. and Polagye, Brian},
abstractNote = {Accurately estimated auto-spectral density functions are essential for characterization of turbulent flows, and they also have applications in computational fluid dynamics modeling, site and inflow characterization for hydrokinetic turbines, and inflow turbulence generation. The Acoustic Doppler Velocimeter (ADV) provides single-point temporally resolved data, that are used to characterize turbulent flows in rivers, seas, and oceans. However, ADV data are susceptible to contamination from various sources, including instrument noise, which is the intrinsic limit to the accuracy of acoustic velocity measurements. Due to the presence of instrument noise, the spectra obtained are altered at high frequencies. The focus of this study is to develop a robust and effective method for accurately estimating auto-spectral density functions from ADV data by reducing or removing the spectral contribution derived from instrument noise. For this purpose, the “Noise Auto-Correlation” (NAC) approach was developed, which exploits the correlation properties of instrument noise to identify and remove its contribution from spectra. The spectra estimated using the NAC approach exhibit increased fidelity and a slope of -5/3 in the inertial range, which is typically observed for turbulent flows. Finally, this study also compares the effectiveness of low-pass Gaussian filters in removing instrument noise with that of the NAC approach. For the data used in this study, both the NAC and Gaussian filter approaches are observed to be capable of removing instrument noise at higher frequencies from the spectra. However, the NAC results are closer to the expected frequency power of -5/3 in the inertial sub-range.},
doi = {10.1016/j.flowmeasinst.2014.03.001},
journal = {Flow Measurement and Instrumentation, 37:29-41},
number = ,
volume = ,
place = {United States},
year = {Sun Mar 02 00:00:00 EST 2014},
month = {Sun Mar 02 00:00:00 EST 2014}
}
  • This work details a methodology for measuring hub height inflow turbulence using moored acoustic Doppler velocimiters (ADVs). This approach is motivated by the shortcomings of alternatives. For example, remote velocity measurements (i.e., from acoustic Doppler profilers) lack sufficient precision for device simulation, and rigid tower-mounted measurements are very expensive and technically challenging in the tidal environment. Moorings offer a low-cost, site-adaptable and robust deployment platform, and ADVs provide the necessary precision to accurately quantify turbulence.
  • A field test that was conducted at Garner Valley, California, on 11 and 12 September 2013 using distributed acoustic sensing (DAS) to sense ground vibrations provided a continuous overnight record of ambient noise. Furthermore, the energy of ambient noise was concentrated between 5 and 25 Hz, which falls into the typical traffic noise frequency band. A standard procedure (Bensen et al., 2007) was adopted to calculate noise cross-correlation functions (NCFs) for 1-min intervals. The 1-min-long NCFs were stacked using the time–frequency domain phase-weighted-stacking method, which significantly improves signal quality. The obtained NCFs were asymmetrical, which was a result of themore » nonuniform distributed noise sources. A precursor appeared on NCFs along one segment, which was traced to a strong localized noise source or a scatterer at a nearby road intersection. NCF for the radial component of two surface accelerometers along a DAS profile gave similar results to those from DAS channels. Here, we calculated the phase velocity dispersion from DAS NCFs using the multichannel analysis of surface waves technique, and the result agrees with active-source results. We then conclude that ambient noise sources and the high spatial sampling of DAS can provide the same subsurface information as traditional active-source methods.« less
  • The authors describe a new, compact vector-velocimeter using a self-mixing type laser Doppler velocimeter (SM-LDV), which electrically identifies features of the sawtooth-like Doppler beat signal waveform to determine the direction of target motion within a velocity range of 23 mm/s-23 m/s. The range of velocity measured is 0.2 mm/s-34 m/s. The measured vector velocity of a reciprocally moving target as well as a rotating target agrees well with the theoretical value.
  • Doppler-broadened spectra have been measured in well-annealed high-purity polycrystalline silver as a function of temperature between 9 and 1098 K by a two-detector Doppler-broadening positron-annihilation technique. Those positrons which annihilate with high-momentum core electrons exhibit a larger fractional change in intensity than those which annihilate with the low-momentum electrons when positron trapping at vacancies occurs. The probability of annihilation with the high-momentum core electrons remains constant as a function of temperature before positron trapping at vacancies is detected. The vacancy-formation enthalpy obtained by applying the two-state trapping model to the high-momentum spectra is 1.11 +- 0.05 eV. The importance ofmore » utilizing high-momentum spectra for extracting vacancy-formation enthalpies is discussed, and a new structure is observed in the curves of the ratio between the high- and low-temperature momentum spectra.« less