Waves in Science & Industry

---

Participating Organisations

ASML  ·  TNO  ·  Marin  ·  Alliander  ·  NLR

---

Programme

Time	Activity
12:00	Welcome & lunch
12:30	Harmen van der Ven (NLR) — The selection of algorithms for the large-scale solution of boundary integral equations
13:00	Sander Rieken & Alex Kolmus (Alliander) — Estimating Underground Cable Temperature from Electromagnetic Waves
13:30	Dave Bekers (TNO) — Future Phased-Array Antennas – Challenges on Electromagnetic analysis and synthesis, leveraging novel manufacturing technologies
14:00	Mario Echeverri Bautista (TNO) — When EM Waves Meet Uncertainty: Understanding Stochastic System Responses
14:30	Coffee break
14:45	Arjen Koop (Marin) — High‑Fidelity CFD for Performance Assessment of Maritime Structures in Waves
15:15	Mark van Kraaij (ASML) — Computational electromagnetics in holistic lithography
15:45	Poster session
16:30	Drinks & food

---

Abstracts

Harmen van der Ven (NLR)
The selection of algorithms for the large-scale solution of boundary integral equations

For homogeneous media, the Maxwell equations are best solved when formulated as boundary integral equations. The multi-level fast-multipole algorithm renders the solution of the dense matrix system feasible. In the actual design and implementation, a lot of choices remain; some good, some bad. Making the right choice from the vast scientific literature is no easy task for the mathematical engineer. In this short talk, experiences with different algorithmic components for the large-scale solution of the equations are shared; where the number of unknowns is typically in the range of one hundred million.

Sander Rieken & Alex Kolmus (Alliander)
Estimating Underground Cable Temperature from Electromagnetic Waves

Alliander operates a large portion of the Dutch electricity distribution grid, which is increasingly constrained by congestion due to electrification and decentralized renewable generation. Accurate temperature estimation of grid components is critical, as it directly determines the grid's operational capacity.

We propose a novel inverse method for estimating cable temperature from time-domain reflectometry (TDR) measurements. The method exploits the temperature dependence of the parameters governing electromagnetic wave propagation in transmission lines. By combining a forward model of wave propagation and reflections with measured TDR signals, the spatial temperature distribution is reconstructed through an inverse problem. The approach is validated using both laboratory experiments and live measurements obtained from the electricity grid.

Dave Bekers (TNO)
Future Phased-Array Antennas – Challenges on Electromagnetic analysis and synthesis, leveraging novel manufacturing technologies

Phased-array antennas are large constellations of individual antennas, often (but not necessarily) organized in a regular grid or lattice. They offer electronic steering of radiated electromagnetic (EM) energy by means of Huygens’ principle, as opposed to mechanical steering, and are used for various applications, such as radar, communication, and astronomy. Initial design and optimization of phased-array antennas relies on the infinite-array assumption, by which the computational effort is reduced to a single unit cell and the physics of the EM problem is described by Bloch-Floquet modes.

However, advances in e.g. power and signal generation (both in power level and in flexibility) require more accuracy in the evaluation of the actual finite-array performance, in particular of cross-talk or EM coupling between antenna elements.

Moreover, current research on array antenna synthesis is mostly focused on the layout. The design of the array antenna element resorts to the optimization of canonical structures, which could be efficiently realised by parameterization and interpolation methods, where boundary or edge effects in the overall array can be accounted for by various approximation methods. At the same time, novel manufacturing technologies such as 3D printing make it possible to fabricate ever more complex antenna elements. Exploitation of these technologies requires new approaches to the synthesis of three-dimensional antenna geometries, capable of fully harnessing the additional degrees of freedom they offer, as well as advanced electromagnetic analysis methods.

Furthermore, fast reconfiguration of the antennas is needed to adapt to changes in the surrounding electromagnetic environment, or to compensate for failure at element-level. This poses additional challenges in terms of real time optimisation.

Mario Echeverri Bautista (TNO)
When EM Waves Meet Uncertainty: Understanding Stochastic System Responses

The response of electronic systems to electromagnetic (EM) waves is an increasingly relevant topic in today's technology-driven society. Advances in electronics have enabled lower power consumption, greater computational capabilities, and ever more compact devices, leading to the widespread adoption of technologies such as wearables, smart homes, autonomous systems, and industrial automation.

As electronic systems become more pervasive, they are exposed to increasingly complex and congested electromagnetic environments. This creates conditions that can enhance unintended interactions between EM waves and electronic systems, commonly referred to as Electromagnetic Interference (EMI). Understanding EMI requires identifying the mechanisms through which electromagnetic energy couples into a system and developing models capable of predicting potential vulnerabilities during the design phase.

Traditional EMI analysis relies on deterministic models. A system is described through its geometry, materials, and operating conditions; an excitation source is defined; and numerical or analytical methods are used to predict quantities of interest, such as voltages or currents within the system.

Real-world systems, however, are not deterministic. Manufacturing tolerances, material variability, environmental conditions, and aging introduce uncertainty into both the system and its response. As a result, the quantities of interest are more appropriately described by probability distributions than by single values. Of particular interest are rare but potentially critical responses, whose accurate estimation can become computationally prohibitive when relying on conventional Monte Carlo approaches.

In this talk, we will explore how uncertainty enters electromagnetic wave–system interactions and discuss the challenges associated with estimating low-probability, high-impact events. We will then introduce importance sampling as a strategy to efficiently characterize the tails of response distributions and illustrate its application to electromagnetic interference problems.

Arjen Koop (Marin)
High‑Fidelity CFD for Performance Assessment of Maritime Structures in Waves

High‑fidelity Computational Fluid Dynamics (CFD) has evolved into a robust, accurate and predictive method for resolving the nonlinear hydrodynamics governing maritime structures in operational and extreme wave environments. This presentation highlights the current state‑of‑the‑art of CFD for quantifying motions, wave‑induced loads, and local free‑surface phenomena for typical maritime applications. Three representative case studies are examined: low‑frequency motion response of an offshore drilling semi‑submersible; the global performance of a floating wind turbine platform in irregular waves; and green‑water impact loads on a ship operating in severe wave conditions. The results highlight CFD's capability to accurately capture complex wave–structure interactions that are often beyond the reach of potential‑flow or mid‑fidelity models. These findings underscore the role of high‑fidelity CFD as both a design‑support tool for engineering decision‑making and a source of high‑quality benchmark data for the calibration and validation of reduced‑order simulation methods.

Mark van Kraaij (ASML)
Computational electromagnetics in holistic lithography

Lithography technology is fundamental to mass producing semiconductor chips. At ASML, extreme ultraviolet (EUV) lithography systems provide the highest resolution for high-volume manufacturing. In this talk we will focus on two areas, optical metrology and computational lithography, that enable holistic EUV lithography. Both areas have a strong foundation in computational electromagnetics. Maxwell solvers (or fast surrogate models) are employed to solve various wave propagation problems. We will discuss some of its computational challenges, solutions, and open problems.