Quantitative non-destructive imaging with limited data
Lead Research Organisation:
Imperial College London
Department Name: Mechanical Engineering
Abstract
If imaging required less data, it would enable faster throughput, improved performance in restricted access situations and simpler, cheaper hardware. The information from images enables damage to be accurately quantified within engineering components, avoiding the need to choose between excessive conservatism and unpredicted failures. To enable improved reconstructions from limited data sets, a diverse set of approaches have been identified, incorporating knowledge of physical wave interaction with objects, use of external information, image processing and other techniques. The fellowship will address the broad problem by applying these approaches to several example applications which are of great interest to industry, and will ultimately enable the development of the field of limited data imaging. While primarily focused on NDE (non-destructive evaluation), the applications of this spread to areas including medicine, geophysics and security.
Planned Impact
The high profile incidents such as the A380 engine explosion in Singapore in 2010, or the Deepwater Horizon accident, can cause significant damage to the environment and place people's health, or worse lives, at risk. Even small, unpredicted failures can have a significant financial cost associated with them. A solution of being more conservative, e.g. replacing components well before the end of their predicted lives, might reduce this risk, but necessitates the wasting of resources at great cost.
NDE provides crucial information to enable components to be used to their maximum without the risk of failure, and the improvements envisaged in this fellowship, looking at improved characterisation from limited data, promise to deliver this information at reduced cost, in less time, and with reduced access requirements. The ultimate outcome of the research clearly promises to have significant impact on society and the economy and I will ensure that the academic work is managed through to application to maximise this impact.
This research is undertaken to be closely associated with industrial needs. I have identified three companies (Rolls-Royce Aero and Submarine divisions, Tenaris and BP) with whom I have already collaborated to identify important applications of my research, and I will have regular 6-monthly steering meetings with them to ensure that all future work is well aligned to their needs. These companies are all members of the RCNDE, a body led by Imperial College, consisting of six UK universities and 16 industrial members: Airbus, AMEC, BAE Systems, BP, EDF Energy, National Nuclear Laboratory, Defence Science and Technology Laboratory (DSTL), E.ON Engineering Ltd, GKN, Office for Nuclear Regulation (ONR), Hitachi, Petrobras, Rolls-Royce Plc., Shell, SKF and Tenaris, covering the nuclear, oil & gas, power, defence, aerospace and transportation industries. I will look to broaden the impact of my work outside the three identified companies by inviting additional RCNDE industrial members to review meetings to identify additional applications of my research.
Beyond my collaborations and planned meetings with the three identified companies, I will share any breakthroughs with the broader industrial community through a number of avenues. I will take advantage of the general engineering audience at the annual research showcase within the Mechanical Engineering Department to present my work both via formal plenary presentations and through less formal discussions with smaller groups visiting the NDE lab. My involvement in RCNDE meetings also provides an opportunity to share my work with a more focused industrial NDE community through presentations during the regular RCNDE industrial visits to Imperial. Any promising outcomes of these discussions can be followed up with invitations to the more focused review meetings discussed above. The RCNDE also hosts regular technology transfer events and I plan to exploit these to explain the processes and techniques behind my research outcomes to industry.
The Imperial College NDE group has an extremely strong track record for the delivery of technology to industry, with two successful spin-out companies and several license deals. All commercialisation of Imperial College intellectual property must be undertaken via Imperial Innovations Ltd., which provides an important pathway to enable academic advances to be licensed and sold to industry. I will exploit this to deliver the more promising outcomes of my research to industrial application. There are 30 associate members in the RCNDE involved in the supply chain (many of them SMEs), and I will arrange to engage with them to help deliver solutions to industry. I plan to develop connections with manufacturers of acquisition equipment for both ultrasound and radiography through platforms such as the BINDT conference, to discuss implementation options.
NDE provides crucial information to enable components to be used to their maximum without the risk of failure, and the improvements envisaged in this fellowship, looking at improved characterisation from limited data, promise to deliver this information at reduced cost, in less time, and with reduced access requirements. The ultimate outcome of the research clearly promises to have significant impact on society and the economy and I will ensure that the academic work is managed through to application to maximise this impact.
This research is undertaken to be closely associated with industrial needs. I have identified three companies (Rolls-Royce Aero and Submarine divisions, Tenaris and BP) with whom I have already collaborated to identify important applications of my research, and I will have regular 6-monthly steering meetings with them to ensure that all future work is well aligned to their needs. These companies are all members of the RCNDE, a body led by Imperial College, consisting of six UK universities and 16 industrial members: Airbus, AMEC, BAE Systems, BP, EDF Energy, National Nuclear Laboratory, Defence Science and Technology Laboratory (DSTL), E.ON Engineering Ltd, GKN, Office for Nuclear Regulation (ONR), Hitachi, Petrobras, Rolls-Royce Plc., Shell, SKF and Tenaris, covering the nuclear, oil & gas, power, defence, aerospace and transportation industries. I will look to broaden the impact of my work outside the three identified companies by inviting additional RCNDE industrial members to review meetings to identify additional applications of my research.
Beyond my collaborations and planned meetings with the three identified companies, I will share any breakthroughs with the broader industrial community through a number of avenues. I will take advantage of the general engineering audience at the annual research showcase within the Mechanical Engineering Department to present my work both via formal plenary presentations and through less formal discussions with smaller groups visiting the NDE lab. My involvement in RCNDE meetings also provides an opportunity to share my work with a more focused industrial NDE community through presentations during the regular RCNDE industrial visits to Imperial. Any promising outcomes of these discussions can be followed up with invitations to the more focused review meetings discussed above. The RCNDE also hosts regular technology transfer events and I plan to exploit these to explain the processes and techniques behind my research outcomes to industry.
The Imperial College NDE group has an extremely strong track record for the delivery of technology to industry, with two successful spin-out companies and several license deals. All commercialisation of Imperial College intellectual property must be undertaken via Imperial Innovations Ltd., which provides an important pathway to enable academic advances to be licensed and sold to industry. I will exploit this to deliver the more promising outcomes of my research to industrial application. There are 30 associate members in the RCNDE involved in the supply chain (many of them SMEs), and I will arrange to engage with them to help deliver solutions to industry. I plan to develop connections with manufacturers of acquisition equipment for both ultrasound and radiography through platforms such as the BINDT conference, to discuss implementation options.
Organisations
- Imperial College London (Fellow, Lead Research Organisation)
- Rolls-Royce (United Kingdom) (Project Partner)
- RCNDE (Project Partner)
- University of Manchester (Project Partner)
- BP (United States) (Project Partner)
- Federal Institute For Materials Research and Testing (Project Partner)
- Tenaris (United States) (Project Partner)
- Manufacturing Technology Centre (United Kingdom) (Project Partner)
Publications
Hoyle C
(2021)
Limited-angle ultrasonic tomography back-projection imaging
in Insight - Non-Destructive Testing and Condition Monitoring
Huang M
(2020)
Elastic wave velocity dispersion in polycrystals with elongated grains: Theoretical and numerical analysis.
in The Journal of the Acoustical Society of America
Huang M
(2020)
Maximizing the accuracy of finite element simulation of elastic wave propagation in polycrystals.
in The Journal of the Acoustical Society of America
Huang M
(2022)
Finite-element and semi-analytical study of elastic wave propagation in strongly scattering polycrystals.
in Proceedings. Mathematical, physical, and engineering sciences
Huang M
(2021)
Longitudinal wave attenuation in polycrystals with elongated grains: 3D numerical and analytical modeling.
in The Journal of the Acoustical Society of America
Hutchins D
(2020)
Mid Infrared Tomography of Polymer Pipes
in Journal of Nondestructive Evaluation
Huthwaite P
(2016)
Eliminating incident subtraction in diffraction tomography
in Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences
Huthwaite P
(2016)
Guided wave tomography with an improved scattering model.
in Proceedings. Mathematical, physical, and engineering sciences
Description | Understanding scattering behaviour of guided waves. This has enabled imaging approaches which have significantly improved resolution from the same data. Development of new CT techniques for faster turbine blade imaging by acquiring less data. Development of new inversion approaches for NDE applications, enabling geometric information about components to be extracted from ultrasound, and improved defect imaging and sizing as a result. |
Exploitation Route | The findings have been published in papers and at conferences. A follow up EngD has been set up to utilise the outcomes for guided wave tomography. |
Sectors | Aerospace Defence and Marine Energy Transport |
URL | http://www.imperial.ac.uk/people/p.huthwaite/publications.html |
Description | Guided wave tomography has been subsequently developed through an EngD to give resolution sufficient for reliable corrosion mapping. Through discussions with a company developing guided wave solutions we are exploring approaches for industrial application. For the outcomes from radiographic CT, providing significant speed-ups in the scanning time through the use of limited data, there have been discussions with end users and CT manufacturers to take this forwards. |
First Year Of Impact | 2023 |
Sector | Energy |
Impact Types | Economic |
Title | Guided wave tomography test data |
Description | A data set for the research community to test guided wave tomography algorithms and enable reliable comparisons to be drawn between different groups. |
Type Of Material | Database/Collection of data |
Year Produced | 2016 |
Provided To Others? | Yes |
Impact | N/A at present. |
URL | http://dx.doi.org/10.5281/zenodo.44626 |
Title | Pogo models of axisymmetric and realistic corrosion maps |
Description | Pogo model files used in PRSA paper by AAE Zimmermann, P Huthwaite, B Pavlakovic, 2020, 'High-resolution thickness maps of corrosion using SH1 guided wave tomography'. These models utilise frames to excite various transducer types:
Directional rectangular SH transducer array on surface
SH0 point source transducer array
SH1 point source transducer array
SH2 point source transducer array
SH point source transducer array on surface
|
Type Of Material | Database/Collection of data |
Year Produced | 2020 |
Provided To Others? | Yes |
URL | https://zenodo.org/record/3825136 |
Title | Pogo models of axisymmetric and realistic corrosion maps |
Description | Pogo model files used in PRSA paper by AAE Zimmermann, P Huthwaite, B Pavlakovic, 2020, 'High-resolution thickness maps of corrosion using SH1 guided wave tomography'. These models utilise frames to excite various transducer types:
Directional rectangular SH transducer array on surface
SH0 point source transducer array
SH1 point source transducer array
SH2 point source transducer array
SH point source transducer array on surface
|
Type Of Material | Database/Collection of data |
Year Produced | 2020 |
Provided To Others? | Yes |
URL | https://zenodo.org/record/3825135 |