1st IRT Saint Exupéry PhD thesis defense on partial electric discharges

 On 3 April 2017, Cédric Abadie became IRT Saint Exupéry’s first PhD student to defend his PhD dissertation. Awarded by University of Toulouse III – Paul Sabatier, his work was supervised by both Laplace Laboratory[1] (CNRS/INP-ENSEEIHT/UT3 Paul Sabatier) and IRT Saint Exupéry. This highly promising research has resulted in the creation of a new digital tool for non-intrusive detection of partial electric discharge to help in the quest for smaller, lighter, more electrical means of transport.

The concept of more electrical aircraft is possible thanks to power electronics, which optimises energy conversion in a machine’s core electromechanical chain. The idea is to progressively replace pneumatic, hydraulic and mechanical power with electricity in an aircraft’s on-board systems. The power required has resulted in an increase of the used voltage, with alternating current rising from 115 to 230 volts and direct current rising from 270 to 540 volts. This translates into an increased stress on the electrical insulation systems resulting in premature degradation of components. Partial discharge, which causes this deterioration, is often described as ‘the silent enemy’. This physical phenomenon may be well known for alternating current being used on the ground, however for aviation things are very different. This sector uses motors that work with continuous current, sectionalised by PWM[2] to control their speed. These generate a lot more electromagnetic noise, however, making it much more difficult to detect partial discharge. Working at altitude means that the components also have to cope with extremes of pressure and temperature not found on the ground. And this control strategy generates a large amount of voltage edge and surges due to the length of the power cables.


The tool developed during Cédric’s Phd research, named ‘Overshot’ (for nOn-intrusiVE paRtial diScHarge detectiOn Tool), is a non-intrusive, in situ, real time sensor. It can be easily used on any test bench for industrial electric motors without disturbing the normal functioning of the equipment, whatever the pressure and speed of the signal edge. The standard methods are both time consuming and costly, requiring either that the motor to be taken out and tested off-line or invasive measuring equipment be put in place. A fundamental part of the research involved studying the effects of pressure on electric discharge, to ensure that the solution being developed would be highly robust. It was found that when the pressure reduces, discharge appears at lower voltages. The nature of the discharge itself is also changed, which alters the tool’s ability to detect it.

The tool itself is very simple and cheap as it attaches to the power cable that connects the motor to its power source, giving it the nickname of the clothes peg. The sensor had previously been developed by Laplace Laboratory as part of IRT Saint Exupéry engineer Thibaut Billard’s[3] own research PhD thesis.

The work carried out for this PhD thesis has improved and optimised a detection method developed at Laplace Laboratory in order to transfer its use to industry. We were able to validate this detection method in an industrial environment thanks to a partnership between the academic and business worlds, put in place by IRT Saint Exupéry in support of this thesis,” says Thierry Lebey, senior CNRS researcher and director of Laplace Laboratory.


It is linked to a method for detecting partial discharge by suppressing the electromagnetic noise in real time using digital processing of the signal based on an innovative application of the wavelet theory pioneered by Frenchman Yves Meyer, recently awarded Abel Prize for mathematics in 2017.IRT’s industrial partners (Airbus, Actia, Liebherr, Safran Electrical & Power, Zodiac Aerospace) and others (Alstom, Leroy-Somer), who may become soon partners, have conducted over ten independent and confidential test campaigns to validate the technique. Additionally, there was a notable collaboration with ITE[4] SuperGrid, an institute of excellence which, like IRT, is part of the French government’s Investments for the Future Programme (PIA).

ACTIA has opened an engine test bench at the IRT in order to carry out tests to evaluate the performance of this tool in the field of electric traction. Consisting of a probe, a whip antenna and a signal analyser, this complex tool represents a significant innovation: it is non-invasive, inexpensive and very simple to implement. Today, several applications of this tool are under consideration: in the field of preventive maintenance and as a tool for validating the performance of dielectric insulation,” explains Pierre Narat, head of Electric Vehicle Solutions at Actia.


The IRT model was able to bring together basic research and the needs of industry. On one hand, an internationally renowned academic research team made good progress in physical phenomenon understanding and measuring partial discharge. On the other hand, it had been identified the needs of industry in terms of sensor and diagnostic software, with testing development.


The tool will potentially now be used in surveillance and health monitoring for electric vehicles, providing regular checks or constant surveillance of systems.

The project is continuing on several fronts. First off, a marketable measurement system could be developed with an industrial partner. Next, this method covers a single phase so it could be adapted to three phases as electric motors are usually three-phases. Another idea is to use multiple sensors to identify where the partial discharge is coming from. This method could indeed become a standard way of detecting partial discharge under aviation-specific conditions. The relevance of the research also forms a good basis for formulating and contributing to the creation of new international standards.

  • Non-Intrusive Partial Discharges Investigations on Aeronautic Motors. T. Billard, C. Abadie, B. Taghia, SAE Technical Paper 2016-01-2058, 2016, doi:10.4271/2016-01-2058.
  • Influence of pressure on partial discharge spectra. C Abadie, T Billard, T Lebey. IEEE Electrical Insulation Conference (EIC), Montréal (Canada), June 2016.
  • Numerical signal processing methods for partial discharge detection in more electrical aircraft. C Abadie, T Billard, T Lebey. Dielectrics (ICD), 2016 IEEE International Conference on Dielectrics. Jul. 2016.

[1] Laboratory on plasma and conversion of energy
[2] Pulse Width Modulation
[3] “Off-line and on-line detection of partial discharge in an electrical vehicle’s low voltage motor using a non-intrusive sensor”, Thibaut Billard, 2014
[4] French institute for energy transition

1st IRT Saint Exupéry PhD thesis defense on partial electric discharges
Scroll to top