Early detection of contact distress for enhanced performance monitoring and predictive inspection of machines
Lead Research Organisation:
University of Southampton
Department Name: Faculty of Engineering & the Environment
Abstract
Monitoring the health of tribocontacts requires the study of friction, tribofilm integrity, and wear transitions. These challenge experimental tribologists to develop accurate methods for in-situ measurements and ideally continuous monitoring. Indirect measurements such as friction changes, sudden heating, changes in vibration or debris in the oil can detect severe wear transitions but cannot detect the subtle mechanistic changes which occur in unhealthy evolution of the contact. However, surface charge generated by tribocontacts and measured by single macro sensors, has detected tribological features such as tribofilm chemistry, adhesive wear, abrasive wear, phase transformations and wear debris but over large surfaces areas. This proposal, therefore, will miniaturise existing sensing technology, with embedded electronics to overcome signal to noise issues, and use arrayed sensors for augmented sensing, and machine learning. The sensor array /learning system would be trained to detect early evidence of lubricated contact decay from charge maps of the surface and allow better prediction of remaining useful life or, what corrective adjustment is needed in running conditions, to assure operational integrity.
Planned Impact
Our plan to deliver industrial, academic and societal impact includes the following actions.
ECONOMIC IMPACT: We will pro-actively grow our pool of partners through the course of the grant by engaging with relevant networks within Knowledge Transfer Network Ltd, the UK Tribology (including IMechE, IoP, IoM3, IET and RSC) and this project will benefit from a unique University of Southampton initiative called Future Worlds (futureworlds.com) which is a vibrant network for connecting academic researchers with external collaborators and investors. A Stakeholder Steering group will be formed of industrial partners will help with effective dissemination of the latest findings to key industrial sectors. Workshops will be held at Southampton and synergy sought between the two RAEng Visiting industrial chairs to nCATS (Prof Honor Powrie from GE and Prof Walter Holweger from Schaeffler.
SOCIETY IMPACTS: Tribology spans fundamental science through engineering to real-world applications and this project provides an exciting opportunity to showcase the value of such research and to engage the next generation. The material understanding will be promoted via website and promotion brochures etc. Showcase events will be arranged to link to a wide range of conferences covering machine learning, condition monitoring, sensors and tribology. Outcomes will be used at Bringing Research to Life roadshow (reaching some 20,000 people annually) and the Winchester Science Centre. We will also work with the Public Policy@Southampton team to help researchers reach and influence key policy makers to press home the message that understanding tribology enables companies to save of energy costs and reduce emissions as well as enable new product designs based on smart machines. In addition during National Science and Engineering Week public lectures will be given.
KNOWLEDGE IMPACT: We will publish our work in the highest quality academic journals as well as present our work at seminars and at the major international conferences. Our contacts will act as advocates for the research in their various communities. The project would also inform the ISO standard ISO 13374 on Machinery CM & Diagnostics.
PEOPLE IMPACT: Academic and research staff will learn from the multidisciplinary collaborations and joint working embedded within each work package design. Researchers will be offered specific placement opportunities with the industrial collaborators. PDRAs and PhDs will join the vibrant research community at Southampton which includes targeted ECR events.
ECONOMIC IMPACT: We will pro-actively grow our pool of partners through the course of the grant by engaging with relevant networks within Knowledge Transfer Network Ltd, the UK Tribology (including IMechE, IoP, IoM3, IET and RSC) and this project will benefit from a unique University of Southampton initiative called Future Worlds (futureworlds.com) which is a vibrant network for connecting academic researchers with external collaborators and investors. A Stakeholder Steering group will be formed of industrial partners will help with effective dissemination of the latest findings to key industrial sectors. Workshops will be held at Southampton and synergy sought between the two RAEng Visiting industrial chairs to nCATS (Prof Honor Powrie from GE and Prof Walter Holweger from Schaeffler.
SOCIETY IMPACTS: Tribology spans fundamental science through engineering to real-world applications and this project provides an exciting opportunity to showcase the value of such research and to engage the next generation. The material understanding will be promoted via website and promotion brochures etc. Showcase events will be arranged to link to a wide range of conferences covering machine learning, condition monitoring, sensors and tribology. Outcomes will be used at Bringing Research to Life roadshow (reaching some 20,000 people annually) and the Winchester Science Centre. We will also work with the Public Policy@Southampton team to help researchers reach and influence key policy makers to press home the message that understanding tribology enables companies to save of energy costs and reduce emissions as well as enable new product designs based on smart machines. In addition during National Science and Engineering Week public lectures will be given.
KNOWLEDGE IMPACT: We will publish our work in the highest quality academic journals as well as present our work at seminars and at the major international conferences. Our contacts will act as advocates for the research in their various communities. The project would also inform the ISO standard ISO 13374 on Machinery CM & Diagnostics.
PEOPLE IMPACT: Academic and research staff will learn from the multidisciplinary collaborations and joint working embedded within each work package design. Researchers will be offered specific placement opportunities with the industrial collaborators. PDRAs and PhDs will join the vibrant research community at Southampton which includes targeted ECR events.
Organisations
- University of Southampton (Lead Research Organisation)
- SENSEYE LIMITED (Collaboration)
- Afton Chemical (Collaboration)
- Schaeffler (Collaboration)
- General Electric (United Kingdom) (Project Partner)
- Shell (United Kingdom) (Project Partner)
- Senseye (Project Partner)
- Schaeffler (Germany) (Project Partner)
Publications
Kubacki M
(2023)
Quantum annealing-based clustering of single cell RNA-seq data.
in Briefings in bioinformatics
Kumar D
(2021)
The Importance of Feature Processing in Deep-Learning-Based Condition Monitoring of Motors
in Mathematical Problems in Engineering
Lu P
(2021)
Early wear detection and its significance for condition monitoring
in Tribology International
Magarò P
(2023)
Wear Mechanisms of Cold-Sprayed Stellite-6 During Reciprocated Dry Sliding Under Different Sliding Speeds
in Journal of Thermal Spray Technology
Sakhamuri M
(2023)
Wear induced changes in surface topography during running-in of rolling-sliding contacts
in Wear
Shetta O
(2022)
Convex Multi-View Clustering Via Robust Low Rank Approximation With Application to Multi-Omic Data.
in IEEE/ACM transactions on computational biology and bioinformatics
Tian Z
(2024)
Charge pattern detection through electrostatic array sensing
in Sensors and Actuators A: Physical
Wood R
(2024)
Coatings and Surface Modification of Alloys for Tribo-Corrosion Applications
in Coatings
Yule L
(2024)
Temperature Monitoring of Through-Thickness Temperature Gradients in Thermal Barrier Coatings Using Ultrasonic Guided Waves
in Journal of Nondestructive Evaluation
Yule L
(2021)
Modelling and Validation of a Guided Acoustic Wave Temperature Monitoring System.
in Sensors (Basel, Switzerland)
Description | Monitoring the health of tribocontacts requires the study of friction, tribofilm integrity, and wear transitions. These challenge experimental tribologists to develop accurate methods for in-situ measurements and ideally continuous monitoring. Indirect measurements such as friction changes, sudden heating, changes in vibration or debris in the oil can detect severe wear transitions but cannot detect the subtle mechanistic changes which occur in unhealthy evolution of the contact. However, surface charge generated by tribocontacts and measured by single macro sensors, has detected tribological features such as tribofilm chemistry, adhesive wear, abrasive wear, phase transformations and wear debris but over large surfaces areas. This proposal, therefore, will miniaturise existing sensing technology, with embedded electronics to overcome signal to noise issues, and use arrayed sensors for augmented sensing, and machine learning. The sensor array /learning system would be trained to detect early evidence of lubricated contact decay from charge maps of the surface and allow better prediction of remaining useful life or, what corrective adjustment is needed in running conditions, to assure operational integrity. Early wear detection: wear debris, oxidised lubricant filming, oxidational wear, micropitting and running in wear detected senor array resulted in the spatial resolution being improved when compared to a bar sensor as well as the sensitivity improved (denoised, sub pC resolution) This study looked at the detection of contact potential differences (CPD) induced by dissimilar metals and oxidational surface chemistry using an electrostatic bar sensor and an array sensor. The main findings can be summarised as follows: An electrostatic array sensor with enhanced spatial and temporal resolution was developed. This enhanced spatial and temporal resolution was achieved with a smaller sensing area, enabling the detection of localised surface charges and the discernment of complex charge patterns. Furthermore, the sensor's reliability was enhanced by having three sensing elements thereby giving some redundancy. Averaging signals from these three elements effectively eliminated random noise and highlighted signals linked to the features of interest. This noise reduction technique not only enhanced measurement precision but also improved the sensor's effectiveness in accentuating pertinent data. Both the electrostatic bar sensor and array sensor were calibrated by known charge patterns induced by a matrix of CPD caused by dissimilar metal inserts. The calibration demonstrated the reliability of the sensors, enabling precise identification of diameters and positions of the metal inserts. To validate the performance and reliability of the sensors, COMSOL modelling was employed to simulate the electric fields generated by the metal inserts. The results of these simulations exhibited agreement with the calibration outcomes. This agreement between modelled response and calibration data highlighted the effectiveness of the sensor design and calibration, positioning them as valuable tools for charge pattern detection. To identify the surface chemistry and its variation over the surface, both sensors were employed to detect CPD induced by surface oxidational wear. The bar sensor captured the evolution of oxidational wear and the presence of Fe3O4 surface layers. The array sensor captured the charge pattern and presented it in the form of a charge map. This charge map exhibited a notable correlation with the variations observed in localised wear phenomena and patchy oxide film coverage on the surface, thereby providing valuable insights into the distribution of surface chemistry. |
Exploitation Route | Looking ahead, the development of an array sensor with more channels to achieve a higher spatial resolution of the surface charge generated by wear is anticipated. Moreover, these array sensors are expected to be employed to generate in-situ charge maps, enabling online monitoring of the progression of localised wear phenomena. Possible Spin-out for the wear detection using electrostatic array sensors and multimodal sensing, data fusion and machine learning to help machine operators |
Sectors | Aerospace Defence and Marine Digital/Communication/Information Technologies (including Software) Electronics Energy Transport |
Title | outlier detection |
Description | outlier detection were packaged as a freely available documented easy to use python package, available to use via 'pip install odds' and documented at https://pypi.org/project/odds/0.1.3/. |
Type Of Material | Computer model/algorithm |
Year Produced | 2022 |
Provided To Others? | Yes |
Impact | open software |
URL | https://pypi.org/project/odds/0.1.3/. |
Description | Bearing model for Schaeffler |
Organisation | Schaeffler |
Department | Schaeffler Technologies AG & Co. KG – Herzogenaurach |
Country | Germany |
Sector | Private |
PI Contribution | Regression equation for surface roughness changes during running-in which will be used in their BearinX design tool. |
Collaborator Contribution | They funded a PhD to look at running-in and surface roughness evolution |
Impact | I paper in Wear Journal and several conference papers and presentations |
Start Year | 2020 |
Description | Heath monitoring of bearings |
Organisation | Schaeffler |
Department | Schaeffler Technologies AG & Co. KG – Herzogenaurach |
Country | Germany |
Sector | Private |
PI Contribution | supervising two industry funded PhDs |
Collaborator Contribution | access to expertise, special testing facilities and coating, sensor technologies |
Impact | PhD supervision |
Start Year | 2019 |
Description | Machine learning |
Organisation | SENSEYE LIMITED |
Country | United Kingdom |
Sector | Private |
PI Contribution | running different machine learning algorithms on their test sets |
Collaborator Contribution | real engineering sensors data to test ML codes |
Impact | real engineering sensor data supplied to test our machine learning toolbox |
Start Year | 2020 |
Description | Senseye CDT PhD student sponsorship |
Organisation | SENSEYE LIMITED |
Country | United Kingdom |
Sector | Private |
PI Contribution | PhD working alongside the main project on ML and digital twin of tribological contacts |
Collaborator Contribution | industrial supervisor and direction of research |
Impact | computer science and health monitoring |
Start Year | 2021 |
Description | early wear monitoring |
Organisation | Afton Chemical |
Country | United States |
Sector | Private |
PI Contribution | we are monitoring early wear in lubrication rigs |
Collaborator Contribution | access to expertise, wear rigs and industry standard gear testing rigs |
Impact | chemistry and mechanical engineering and electronic engineering |
Start Year | 2022 |
Title | ML toolbox of various algorithms |
Description | ODDS python package is found at https://pypi.org/project/odds/. This is a publicly available package. The code is also available to use and edit on Github. |
Type Of Technology | Webtool/Application |
Year Produced | 2022 |
Impact | use of toolbox and feedback from user |
URL | https://pypi.org/project/odds/ |
Title | Python script for reading '.hpf' files |
Description | Python script for reading '.hpf' files |
Type Of Technology | Webtool/Application |
Year Produced | 2020 |
Impact | eads QuickDAQ hpf files, returns 'filename_info.txt', 'filename_data.csv' files file contains function 'write_info_and_csv_from_hpf' which takes a filename as input, and has optional output filename argument. This reads through the '.hpf' file and returns an info file as txt, has the name 'filename_info.txt' and the data in double format as a csv file, called 'filename_data.csv' which has the data as a single column, and the header is the channel number. Unfortunately at present I only have single channel hpf data files, so I've not quite finished the coding for outputting as multichannel data. I will update if I get a multichannel '.hpf' file. This was originally written by translating the matlab code from https://www.datatranslation.de/Download/Demos/Hpf2Matlab_Rev2_0.m |
URL | https://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fjogrundy%2FRead_Quick... |
Description | 26th World Micromachine Summit 2023 Bucharest, Romania, May 22- 24, 2023 MMS 2023, Bucharest, Romania (imt.ro) "Condition and Environment Monitoring: New ideas for sensing and powering" Harris. N |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Other audiences |
Results and Impact | talk "Condition and Environment Monitoring: New ideas for sensing and powering" Harris. N |
Year(s) Of Engagement Activity | 2023 |
Description | Artificial Intelligence for Scientific Discovery (AI3SD) Winter Seminar Series 2020/21 |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Industry/Business |
Results and Impact | Artificial Intelligence for Scientific Discovery (AI3SD) Winter Seminar Series 2020/21 'Outlier Detection in Scientific Data' Jan 20th 2021, Seminar Centre For Machine Intelligence Show Case event, 2019, 'Outlier Detection on Sensor Data' Poster |
Year(s) Of Engagement Activity | 2020,2021 |
Description | Austria conference: ÖTG-SYMPOSIUM 2019 Title: Monitoring the health of tribo-contacts |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Industry/Business |
Results and Impact | OTG-Symposium is an opportunity to gain insights relating to tribology which is the science and technology of interacting surfaces in relative motion. The event explores the application of knowledge from all areas of tribological research to control friction and wear, in particular through the optimal design of functional elements. |
Year(s) Of Engagement Activity | 2019 |
Description | Opening talk at IoP/IMechE Digital Tribology conference |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Industry/Business |
Results and Impact | This digital tribology workshop will focus on current state-of-the-art data-driven modelling phenomena associated with friction, wear and lubrication applied to a wide range of tribological systems (bearings, gears, engines, etc.). Special emphasis will be placed on multi-sensing and multi- modality data fusion methods and the impact that these approaches will have in the future, particularly for tackling global challenges such as environmental issues and the move towards carbon-free energy. The workshop will consist of a series of invited presentations delivered by industry and academic leaders followed by round table discussions. A key objective of the meeting is to provide a forum for open discussions and for identifying current and future research opportunities/challenges in digital tribology, especially those arising from global societal and technological demands. Following the workshop, participants will be invited to contribute to a perspective paper that will be submitted to a leading tribology journal. |
Year(s) Of Engagement Activity | 2023 |
URL | https://www.iop.org/events/future-digital-tribology#gref |
Description | Oral Presentation at the STLE 2022, Orlando, USA (May 2022) |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Postgraduate students |
Results and Impact | STLE's Annual Meeting & Exhibition showcases some 400 technical presentations, application-based case studies, best practice reports and discussion panels on technical and market trends. |
Year(s) Of Engagement Activity | 2022 |
URL | https://www.tribonet.org/conferences/2022/05/stle-2022-annual-meeting-exhibition/ |
Description | Oral presentation at WTC 2022, Lyon, France (July 2022) |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Postgraduate students |
Results and Impact | The 7th World Tribology Congress aims to highlight recent important progress in all aspects of Tribology, strengthen the links between academy and industry, provide a unique opportunity for discussion concerning the latest developments in Tribology and to promote international collaborations and exchanges. The Congress will consist of 4 plenary + 4 keynote speakers, 41 invited talks, 700+ oral presentations and 300+ posters on topics at the cutting edge of the various Tribology disciplines, a wide exhibition and various events, scientific and nonscientific, some of them dedicated to young tribologists. |
Year(s) Of Engagement Activity | 2022 |
URL | https://www.wtc-2022.org/ |
Description | Project workshop on digital tribology and early wear detection |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Industry/Business |
Results and Impact | progress presented on project to industry and keynote speakers from industry and discussion on future needs and trends |
Year(s) Of Engagement Activity | 2021 |
Description | Sensors and the IoT for the Environment at Symposium on High-Tech Collaborative Innovation, Quingdao City, China |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Other audiences |
Results and Impact | Invited Talk: Sensors and the IoT for the Environment at Symposium on High-Tech Collaborative Innovation, Quingdao City, China |
Year(s) Of Engagement Activity | 2020 |
Description | Talk at IoP event of the Future of digital tribology 23rd Nov 2022, Chilworth Science Park: The Future of Tribology is digital (Wood and Grundy) |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Industry/Business |
Results and Impact | talk The Future of Tribology is digital (Wood and Grundy) |
Year(s) Of Engagement Activity | 2022 |
Description | keynote at wireless and optical technologies international conference |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | keynote speaker |
Year(s) Of Engagement Activity | 2022 |
URL | https://eventos.uma.es/60361/section/28397/global-conference-on-wireless-and-optical-technologies-20... |
Description | talk at the AI3SD conference in Chilworth on the 1st March 2022 on Outlier detection in scientific data. https://www.ai3sd.org/ai3sd-event/1-3-03-2022-ai3sd-network-conference-2022-chillworth-manor-hotel/ |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Industry/Business |
Results and Impact | talk at this event https://www.ai3sd.org/ai3sd-event/1-3-03-2022-ai3sd-network-conference-2022-chillworth-manor-hotel/ |
Year(s) Of Engagement Activity | 2022 |
URL | https://www.ai3sd.org/ai3sd-event/1-3-03-2022-ai3sd-network-conference-2022-chillworth-manor-hotel/ |