Development of a non-invasive airborne acoustic technique to monitor the dynamics of water systems

Lead Research Organisation: University of Sheffield
Department Name: Mechanical Engineering

Abstract

This project proposes a new theoretical method to measure non-invasively the key characteristics of a two dimensional dynamically rough free surface of a turbulent open channel flow using airborne acoustic waves. The research of this project will cover both numerical simulation and analytical derivation of the approximated solutions based on extending the Kirchhoff approximation technique to the case of rough surface containing multiple scales. This will be used to develop a technique of recovering water surface profile based on the data recorded on array of microphones. The method will provide sub-millimetre accuracy required to identify various scales of gravity-capillary waves present on the surface of shallow water flow. This information can potentially be linked to underwater conditions and flow velocity profile.

The project outcomes will provide a virtual environment to prove the concept of non-invasive measurements as well as a tool to interpret the collected data. The numerical simulations will be adapted to open source software (i.e. GNU Octave and Scilab). For dissemination of the results and further developments, the software will be realised to acoustic and hydraulic communities as open source.

The key novelty of this project is that the proposed method will be able to work reliably with a broad range of flows typically found in rivers, estuaries and partially filled pipes (sewers), which are the key components of water infrastructure and are critical in the water circulation process. This work will enable the development of non-invasive instrumentation which can determine flow rates and sediment transport in natural and man-made channels. This will enhance our capability for flood risk assessment and pollution monitoring.

Planned Impact

Non-invasive monitoring technique of the water systems will be the main impact of this project. The project is directly linked to the research and development of low-cost low-maintenance acoustic wave probes conducted by Pennine Water Group at the University of Sheffield. The results of these developments will shape technology for a follow-on spin-off commercial project that has the potential to revolutionize the way the water infrastructure is managed through the considerable improvement in the quality of data. The technology will benefit multiple stakeholders and will be used in various applications. The list of potential beneficiaries includes business, industry and local authorities. Currently, there is no experimental system which is able to measure with sub-millimetre accuracy the instantaneous variation of the free surface of intermediate/shallow water flow, particularly in-situ. The technology I propose to develop will dramatically improve the resolution and accuracy of these measurements. This project will provide a technique of analysing acoustic data and linking these analytically with temporal and spatial variability of the dynamic water surface. The results of this work can be of use in research and development projects which are sponsored by the industrial partners to improve the accuracy of the flow measuring devices. The results can also be used within university laboratories to investigate various parameters of the water flow (i.e. flow rate, Reynolds number) and their link to underwater conditions (i.e. condition of the flow bed, water pollution).

The monitoring systems based on the airborne acoustic measurement principle can be installed in sewerage pipes and over rivers/channels to monitor the dynamics of the water flow continuously and with high accuracy. This will enable the water industry to know in advance any potentially critical changes in the monitored system. As well as providing the water industry with continuous monitoring, this technology can also be linked to the insurance companies and help to improve the risk analysis associated with infrastructure deterioration rates and floods. It is suggested that the continuous monitoring technique of the water systems may result in reduced insurance cost for households situated within flooding risk area. Local authorities can benefit from the data that this monitoring technology can provide in terms of improved planning of developments and correctly assessing the need in investments in the local water infrastructure.

A main knowledge impact of this project will be realised in the form of a virtual modelling environment which will enable stakeholders to analyse the way their flow metering devices operate under various flow conditions, to improve the accuracy of the currently used experimental and monitoring techniques, and to inform future development of the monitoring techniques for water systems. The existence of this software will also provide people with advanced skills in developing the acoustic probes. Part of the software will be used to develop a mobile app for smartphones that will illustrate the potentials of the technology and enable wide public engagement.

Publications

10 25 50
publication icon
Dolcetti G (2017) Doppler spectra of airborne sound backscattered by the free surface of a shallow turbulent water flow. in The Journal of the Acoustical Society of America

publication icon
Dolcetti G (2016) Reconstruction of the dynamic surface roughness with acoustic imaging technique in The Journal of the Acoustical Society of America

 
Description The initial idea of monitoring technique has been studied in depth and several monitoring techniques have been developed. In the beginning of the project while collecting data on the dynamic water surface it was discovered that Acoustic Doppler effect observed in forward scattering due to moving rough surface can be used to reconstruct shallow water surface spectra. This was compared with traditional approach for Acoustic Doppler in back scatter and it was concluded that in forward scattering measurements of Doppler effect are more reliable compared to the traditional back scattering case. These studies were published in two papers and presented in the international conferences. The reconstruction technique for one-dimensional (2D acoustic problem with surface uniform in the out of plane direction) and two dimensional (full 3D acoustic problem) surfaces have been developed and validated with the numerical simulations based on the Boundary Integral Equation method. Using the exact integral formulation the effect of multiple scattering supported by rough surface was studied and compared with the developed approximations based on a single reflection assumption (Kirchhoff approximation). It was discovered that for a wide range of shallow water conditions the Kirchhoff approximation is accurate and enough to model the scattered acoustic wave-field. It was discovered in the 2D problem with the optimisation method that in order to achieve accurate reconstruction of the surface the receivers must be placed in the specific locations where the error of the employed inversion method is minimised. This discovery makes developed direct surface reconstruction technique stable and less sensitive to potential uncertainties and noise. These results are in process to be submitted to peer-reviewed journal. In addition to the direct surface reconstruction, the possibility to use Fourier image of the surface (in reciprocal domain) was studied and it was discovered that computational efforts and stability of the inversion technique could be improved. This spectral technique was adapted to surfaces with small and large roughness elevations compared to the acoustic wavelength. The results are in process to be submitted to the peer-reviewed journal. Further approximations were developed based on the stationary phase approach. This method gives simple form of the solution defined at the specular point. With the use of two specular points and wavelet transform the new method was developed to reconstruct dispersion of water waves for stationary and non-stationary dynamic shallow water surfaces. This technique can be used to assess the surface flow velocity for rapidly changing conditions. The results are in process to be submitted to the peer-reviewed journal. It was also discovered that the wavelet-based method has a potential to be applied to CCTV video of open channel and river flows, providing a spatial map of the instantaneous surface velocity field. The preliminary results indicate potentials of the method to be developed into further proposals for future funding.

UPD 2019

Further developments of the reconstruction method have been initiated. In particular, the inversion algorithm is made stable with implementation of broadband source and further improvements in regularisation technique. This is tested at Siemens laboratory and the data is being post-processed for validation purposes.

The reconstruction method has also been applied to data collected with electromagnetic microwave sensors within the STREAM PhD project. The scattered signal has been adequately predicted by the model and next step is to explore the possibility to build an antenna array capable of receiving both signal strength and phase.

UPD 2020

The array of phase matched microphones was tested first time over the dynamic water interface. The acoustic data demonstrated accurate measurements of the surface velocity. The measurements obtained with EM sensors (built by the PhD student) were compared with the data collected with wave probes immersed into the wave tank. The measurements demonstrated the accuracy of the EM sensors in application to reconstruction of water surface dynamic characteristics.
Exploitation Route The new techniques developed in this project pave way for a conceptually new approach to the monitoring of water infrastructure characterised as non-invasive, portable, low-cost and requiring reduced amount of data from the acoustic transducers. Based on the developments of this project the PI is currently preparing standard grant EPSRC application in collaboration with the University of Aberdeen, School of Engineering (Prof. Vladimir Nikora). The aim of this project will be to test augmented acoustic monitoring techniques (combined with optical measurements) in laboratory and in-field conditions. Two external projects have been planned for summer 2018 subject to the funding availability for research visits. One project is concerned with the measurement of water discharge profile studied in the Politecnico di Torino (Turin, Italy). This project will be partially funded by the host institution Politecnico di Torino and the research team will also apply to UK Acoustics network (UKAN, https://acoustics.ac.uk/) for the research visit grant. Siemens laboratory in Leuven (Dr. Laurent De Ryck, Laurent.deryck@siemens.com, Simulation & Test Solutions, Siemens Industry Software NV, Interleuvenlaan 68, B-3001 Leuven, Belgium) has recently expressed interest in employing the developed surface reconstruction technique in application to analysing porous surfaces, metasurafaces and acoustic surface impedance. This project offers new application of the developed methods as well as provides opportunity to validate reconstruction technique for the full 3D acoustic problem (Siemens will provide multichannel data acquisition system and array of microphones as an in-kind support). The PI will apply for short term scientific mission funding offered by COST Association network DENORMS (CA15125).

UPD 2019:

Improving reconstruction technique with the help of broadband signal will potentially provide a platform for the recovery of full 3D surface parameters.

Further developments of the inversion technique are planned with the use of machine learning algorithms applied to acoustic scattering problem. The standard EPSRC grant proposal mentioned previously is in progress of writing and its main objective now is to explore potentials of machine learning approach.

UPD 2020:

The results of this project helped to secure IAA EPSRC funds to support the development of high TRL technique in collaboration with Siemens research group from Leuven, Belgium. This research group also supported our visit to Leuven in February 2019 to perform initial experiments over the static surfaces. The new experiments conducted in November 2019 showed the validity of the developed methods within the project in application to the dynamic water surface. The results obtained over static surface are submitted for the presentation at Acoustic Forum 2020, Lion and results for dynamic surface are drafted into the publication being prepared for submission to a peer-reviewed journal.
Sectors Aerospace, Defence and Marine,Environment

 
Description The research supported by the award produced two publications in a major peer-reviewed journal, two papers in conference proceedings, and one book chapter, as well as three additional journal articles currently under preparation. This research output was career-defining for the research associate employed in the project, and greatly contributed to his establishment in the field of research. Support for the attendance of two international congresses and one summer school provided by the award had a significant impact on the exposure of the research results, and allowed the research associate to build a solid interdisciplinary network of contacts. Some of these contacts have already materialised into a collaboration within the duration of the project, while others will support future bids aimed at developing further applications of the method. The award also supported the development of advanced technical skills of the research associate, such as programming, creation of a database, numerical approaches, as well as transferrable skills such as presentation, networking, reporting, project management and monitoring. This project formed a platform for the new technology that can be offered in future to the water companies to enhance current monitoring capabilities. Planned collaboration with Siemens and Politecnico di Torino (see Key Findings for more information) will enable us to take first steps in validating the technology for 3D acoustic problem with array of acoustic transducers. This will take the proposed technique near the prototype level of technological readiness. The project results have had impact on shaping the aims of two work packages in successful EPSRC application (Rapid monitoring of river hydrodynamics and morphology using acoustic holography). The developed theory can be adapted to electro-magnetic waves to enhance post-processing technique of data collected by currently existing EM sensors used in measuring water flow. This idea led to securing industrially supported STREAM PhD project started in October 2017 (PI is co-supervisor) that provides benefits for both PhD students and company involved. UPD 2019: The results of this project are now contributing to the recently awarded EPSRC grant EP/R022275/1. In particular, the project team (including the PI of this project) are building a prototype of the acoustic array capable of reconstructing dynamic surface parameters. The trial tests were conducted on various static surfaces (with 2D and 3D roughness profiles) in collaboration with Siemens lab in Leuven (Contact details: Laurent de Ryck, laurent.deryck@siemens.com, Product Development Manager of Siemens lndustry Software, Researchpark 1237, lnterleuvenlaan 68, 3001 Leuven, Belgium). The project has provided a platform for new PhD research where it is planned to use developed inversion techniques and machine learning algorithms in reconstruction of 3D surfaces. The PhD project will start in September 2019 and it is offered to UK student graduating this year. The student is expected to receive degree in Mathematics and the project will give candidate an opportunity to apply his skills in Mathematics to a real engineering problem. At least two papers are in line for publication and are currently in progress of writing. UPD 2020 The funds obtained with IAA EPSRC grant helped to progress the use of the inversion techniques in developing a higher TRL level measurement tools which are now more close to be tested in field to measure flow conditions in open channels.
Sector Environment
 
Description 2016 ASA Student Council Travel Award
Amount $250 (USD)
Organisation Acoustical Society of America 
Sector Charity/Non Profit
Country United States
Start 11/2016 
End 11/2016
 
Description Rough surface characterisation with acoustic sensor array, EPSRC IAA
Amount £24,999 (GBP)
Funding ID University number R/162064-11 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Public
Country United Kingdom
Start 07/2019 
End 03/2020
 
Title 2D and 3D models of scattered sound from rigid rough water surfaces 
Description Set of numerical models to predict the acoustic pressure generated by multiple sources and reflected by the dynamically rough water surface at multiple locations. Different models based on various approximations have been derived and cross-validated, namely, Small Perturbation model, Kirchhoff model, Boundary Element Method model. The model accepts arbitrary surface shapes including measured time-records of the surface elevation over a significant spatial region, and arbitrary spatial distributions of source and receivers. The model allows solution of the full three-dimensional problem, of a simplified two-dimensional problem, and of a hybrid 2D-3D problem (3D reflections from a surface with zero-gradient in one direction). The model is restricted to slow surface dynamics (compared to the speed of sound), and reflections in the far field from continuous connected surfaces. 
Type Of Material Computer model/algorithm 
Year Produced 2017 
Provided To Others? No  
Impact The validation of the models in terms of its capability to predict the average Doppler spectra of airborne ultrasound backscattered and forward scattered has been presented in the papers `G. Dolcetti, A. Krynkin, and K. V. Horoshenkov, (2017) "Doppler spectra of airborne sound backscattered by the free surface of a shallow turbulent flow", The Journal of the Acoustical Society of America, 142(6), 3387-3401. (https://doi.org/10.1121/1.5015990)' and `G. Dolcetti and A. Krynkin, (2017) \Doppler spectra of airborne ultrasound forward scattered by the rough surface of open channel turbulent water flows", The Journal of the Acoustical Society of America, 142(5), 3122-3134. (https://doi.org/10.1121/1.5011183)'. The Kirchhoff-based model is currently being used for the optimisation of the geometry of the array of sensors used for the surface reconstruction, and the results will be submitted to a major peer-reviewed journal. The model is easily modified to take into account different boundary conditions, therefore it can be used to calculate reflections of underwater sound, or be further modified to describe electromagnetic scattering. 
 
Title database of rough water surface realisations 
Description Collection of random realisations of linear two-dimensional and three-dimensional surfaces, with spectrum and scales representative of the surface of open-channel flows, and Matlab code to generate them. The surfaces are generated assuming a parametrised power-function spatial spectrum of the surface and Gaussian linear statistics. 
Type Of Material Database/Collection of data 
Year Produced 2017 
Provided To Others? No  
Impact The database was employed for the modelling of the acoustic field scattered by the rough dynamic water surface using a Kirchhoff model and a Boundary Elements Method model. The results of the model, and validation of the statistics of the surfaces within the database against experimental measurements in a laboratory flume, are reported in the papers `G. Dolcetti and A. Krynkin, (2017) "Doppler spectra of airborne ultrasound forward scattered by the rough surface of open channel turbulent water flows", The Journal of the Acoustical Society of America, 142(5), 3122-3134. (https://doi.org/10.1121/1.5011183)`, and `G. Dolcetti, A. Krynkin, and K. V. Horoshenkov, (2017) "Doppler spectra of airborne sound backscattered by the free surface of a shallow turbulent flow", The Journal of the Acoustical Society of America, 142(6), 3387-3401. (https://doi.org/10.1121/1.5015990)`. The details of the numerical model used to generate the surface realisations is reported in the same papers. The database is used for the study on the optimisation of the sensor geometry for the surface reconstruction, in a manuscript currently in preparation by the PI and the research associate. It is planned to further validate the database against experimental field measurements of the surface elevation provided by Dr Carl Legleiter and Dr Brandon Overstreet (US Geological Survey) based on optical methods. Such a validation, in conjunction with the framework developed by Dr Legleiter and Dr Overstreet (C. Legleiter, Curtis D Mobley, and Brandon T Overstreet, (2017) "A framework for modeling connections between hydraulics, water surface roughness, and surface reflectance in open channel flows", Journal of Geophysical Research: Earth Surface, 122, doi:10.1002/2017JF004323), will enable the development of novel approaches to monitoring rivers using optical aerial data, and will facilitate flight planning to improve the remote monitoring of alternative river attributes from planes. 
 
Description Validation of the inversion technique and building acoustic array prototype 
Organisation Siemens industry Software
Country United Kingdom 
Sector Private 
PI Contribution Brought rigid rough surfaces for acoustic tests (3D printed and CNC), frame controlled by Arduino and other equipment for recording scattered acoustic signal at various locations over the rough surface.
Collaborator Contribution Provided lab space (industrial anechoic chamber), LMS system and 7mm microphone arrays (64 microphones in total) for 9 working days. Time of the researcher (at least 2 working days).
Impact - Improvements in inversion technique based on the implementation of broadband signal - Proposal of new method to reconstruct key characteristics of the surface - Potential application of rough surface scattering in measurements of surface acoustic properties - Further experiments and Siemens group research visit planned later this year. Planning to apply for the travel grants offered by DENORMS (https://denorms.eu/) and UKAN (https://acoustics.ac.uk/) networks.
Start Year 2018
 
Title Modelling and Optimisation tool for water surface reconstruction 
Description The toolkit developed in Matlab allows the automatic determination of the geometry of sensors that provides the best reconstruction (in average) of the water surface of an open channel flow using the newly developed inversion technique, based on some user defined constraints (maximum number of sensors, acoustic frequency, required resolution and surface size, average properties of the flow, wave height). The software integrates a parametric linear model of the random water surface, a Kirchhoff model of the scattered acoustic pressure field, and a genetic optimal search algorithm for the optimal geometry. The software has been developed for both 2D and 3D problems, but the 3D validation has not been performed due to computational constraints. The software is also able to provide an estimate of the measurement uncertainties. 
Type Of Technology Physical Model/Kit 
Year Produced 2018 
Impact Validation of the toolkit is included in a manuscript currently under preparation for a major peer-reviewed journal. The toolkit is currently being used for the design of an array prototype to test in the laboratory facilities of the Politecnico di Torino (Italy), in collaboration with Prof. Davide Poggi. Results of this test will be available at the end of Summer 2018, and will be included in a new journal article. After validation, the toolkit will inform the optimal design of the experimental setup for specific applications, hence facilitating the widespread application of the technology developed thanks to this award. The results will also inform the development of shape-adaptive arrays able to implement the technique over a wide range of flow conditions. 
 
Description Wider application of the inversion technique: acoustic surface measurements 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Industry/Business
Results and Impact Working group discussion was organised during the research visit to Siemens laboratory in Leuven. Aim of the discussion was to find further application of the method to measurement of surface properties and facilitate development of alternative inverse methods which are more computationally effective.
Year(s) Of Engagement Activity 2019
URL https://twitter.com/dolcetti_g