COMPutationally empowered Electromagnetic industrial TalEnts

Project Information

COMPETE

Grant agreement ID: 955476

DOI

10.3030/955476

Project closed

EC signature date 2 July 2020

Start date 1 March 2021

End date 28 February 2025

Funded under

• EXCELLENT SCIENCE - Marie Skłodowska-Curie Actions

Total cost

€ 1 334 103,12

EU contribution

€ 1 334 103,12

Coordinated by

POLITECNICO DI TORINO
Italy

Periodic Reporting for period 2 - COMPETE (COMPutationally empowered Electromagnetic industrial TalEnts)

Reporting period: 2023-03-01 to 2025-02-28

Summary of the context and overall objectives of the project

As an impacting discipline, computational electromagnetics is expanding rapidly. The future is looking strong, mainly due to the growing demand for software for the design and analysis of electrical devices. However, standard doctoral programmes in computational electromagnetics are often too theoretical, abstract and far from the on-field industrial needs. The aim of COMPETE (COMPutationally empowered Electromagnetic industrial TalEnts) is to bridge the gap between the mindset and needs of advanced electromagnetics industry and the academic doctoral programs. This training has provided a very pragmatic prospective on the current industrial needs in modelling and has formed an impactful cohort of young scientists and innovators in the field of industrial computational electromagnetics.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

The project has been articulated in five individual research projects, corresponding to the PhD topics of five fellows:

Individual project of Johann Bourhis (ESR1): new paradigms in moderate, low, and extremely low-frequency modelling for EMC and EMI industrial applications.
Activities performed within this project: After a bibliographic analysis and a substantial phase of preliminary implementation and theoretical investigations, the fellow has obtained two new formulations extending the use of the quasi-Helmholtz projectors to the industrially relevant cases of structures containing junctions and to high order problems.
Publications: Two journal papers and four conference papers published. Other two journal papers are in preparation.
Awards: the research of this individual project has received an honorable mention award at to IEEE AP-S/URSI Denver 2022.

Individual project of Pierrick Cordel (ESR2): new paradigms in moderate, high, and extremely high-frequency modelling for large scale scattering and industrial design
Activities performed within this project: After a preliminary set up, the fellow investigated and obtained three new solvers in time domain of electric type and of combined type allowing wideband simulations including signals with substantial high frequency components.
Publications: two journal papers and five conference papers. Other three journal papers are in preparations.
Awards: The research of this individual project has received an honorable mention award at to IEEE AP-S/URSI Portland 2023 and another paper was selected as one of the 10 finalists at IEEE AP-S/URSI 2022.

Individual project of Margaux Bruliard (ESR3): new paradigms in modeling of impedance and metasurfaces for radiation and scattering for next-generation industrial application.
Activities performed within this project: after a background building phase, the fellow investigated several IBC important scenarios, obtaining a new stabilization techniques for the IBC EFIE.
Publications: one journal paper and two conference papers has been published.

Individual project of Leonardo Pollini (ESR4): new paradigms in uncertainty quantification, sensitivity analysis, and tolerance assessment, modelling, specification, and implementation
Activities performed within this project: the project started with a propaedeutic study of the state of the art, then the project focused on the investigation of the behavior of random-based global optimization strategies obtaining results that are absolutely comparable to the human knowledge-based solution but in a completely automatable way.
Publications: the results obtained have been published in 4 conference papers and submitted to one journal paper.

Individual project of Paolo Ricci (ESR5): new paradigms in computationally empowered machine learning for industrial EM design and assessments.
Activities performed within this project: this project obtained a new single current formulation for imaging that leverages dual discretization schemes to avoid approximating the relationships linking electric and magnetic currents. The formulation was then extended to a resonance-free one and the activities also included the investigations of the enhancing and application of this approach within a machine learning-based current imaging setting.
Publications: two journal papers and four conference papers.
Awards: The research of this individual project has received an honorable mention award at IEEE AP-S/URSI Denver 2022.

Disseminaton and Communication
As detailed above for each fellow, several journal and conference papers has been delivered by this project and are available in open access. There are currently 6 journal papers in preparation. Moreover, during this project, all fellows have been involved in communication activities: they participated to science fairs for high schools, university lab open door presentations, and they have been all actively involved in the European Night of Researchers.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

Several impactful results has been obtained by this project. More in detail:

High-Order Quasi-Helmholtz Projectors

This technique enhances the stability and accuracy of electromagnetic simulations, particularly in low-frequency and dense mesh regimes. By enabling stabilization without requiring the detection of global cycles, it simplifies complex simulations in industrial scenarios, such as EMC/EMI modeling. Its compatibility with high-order discretizations positions it as a valuable enhancement for commercial EM solvers used in automotive, aerospace, and telecommunications industries.

Low-Frequency Stabilization for Junctioned Structures

An advanced method was developed to address low-frequency instabilities in structures with metallic and dielectric junctions. Using a graph-based extension of quasi-Helmholtz projectors, the technique efficiently handles the presence of complex junctions, enabling reliable simulation without performance degradation. This is critical and exploitable in modern electronic packaging and aircraft design where intricate junctions are common.

Filtered EFIE-IBC Solver for Metasurfaces

A robust solver for EFIE-IBC problems with inductive metasurfaces was created using spectral filtering to manage instability in inductive impedance values. This regularization technique allows accurate modeling of reflective and transmissive metasurfaces, a cornerstone in next-generation antenna design and stealth applications.

Modular Software Components for EM Simulation Pipelines

Several modules developed during the project, such as high-frequency time-domain solvers, preconditioners, and impedance model handlers, can be integrated into existing commercial or open-source EM software. Their modular design facilitates adoption by third-party platforms looking to improve solver stability and performance.

In addition to the advancements of the state of the art obtained, the project has addressed real case scenarios impacted by the technology developed so far. Realistic structures taken from industrial challenges has been considered both to benchmark the newly obtained techniques and to show their practical impact on modern advanced electromagnetics. We expect a substantial impact on the modelling state of the art and related business in our field as computational electromagnetics is at the very heart of several technological fronts both in industry and academia.

qh-mobius-1e1.png