2014

Caterina Bigoni – Whitney forms in computational electromagnetism

 About ABB S.p.A – Low Voltage Product Division

 

Description

My internship took place at ABB S.p.A in Vittuone, Italy under the supervision of Dr. Luca Ghezzi, head of the Simulation group in the R&D department of the LP (Low Voltage Product) division.

ABB is a global leader in power and automation technologies that enable customers to improve performance while lowering environmental impact.

Differential forms have significant advantages in electromagnetic field theory as they clarify the relationship between field intensity and flux density. In the discrete world, Whitney (differential) forms represent a family of finite elements that satisfy exactly the relations on fluxes and circulations introduced by Maxwell, specifically Stokes’ theorem. The aim of the internship was to get familiar with differential forms and their discrete counterpart on the theoretical level at first and then use these tools to solve electrostatic and magnetostatic problems in three dimensions. In particular, the electrostatic problem can be seen as a Laplace equation, while the curl-curl operator rules the magnetostatic one. The former aims to find the electric potential using nodal elements, while the latter determines the magnetic potential vector A using the so-called edge elements; finally, by interpolations of A by means of Whitney 2-forms, it is possible to calculate the magnetic flux density.

With the objective of solving with linear Whitney elements these two PDEs, an object-oriented code was written with MATLAB and graphical outputs of the solutions were obtained using both MATLAB and the scientific visualization program ParaView.

 

Fig. 1: Electric potential (0-form), solution of the electrostatic problem for a test case (copper wire in an air box).

 

 

Benjamin Paccaud – Création d’un outil de modélisation et de calcul de structure

About Ingeni Ingénierie structurale

 

Description

L’objectif de ce stage était de créer un petit logiciel de calcul de structure permettant de donner rapidement aux ingénieurs une idée du fonctionnement général de leurs conceptions. Il est en effet parfois difficile d’avoir une bonne intuition du cheminement des efforts dans une structure. Le projet se restreint à un type de matériau précis : le béton armé. Ce matériau à la particularité d’avoir des comportements très différents selon qu’il est sollicité en traction ou en compression, le béton ne résistant presque pas à la traction, on considère que seul l’acier reprend les efforts de tractions. Or, dans la construction traditionnelle, les barres d’armatures dans le béton sont placées pour former une grille avec une trame régulière. Le matériau est modélisé en utilisant un modèle de bielles et tirants : on considère que le béton travaille en compression avec des bielles d’orientation variable et que l’acier reprend la traction selon l’orientation des barres d’armature. On peut donc considérer des éléments finis de type barre en 2D, et résoudre l’équation de l’équilibre des forces pour un cas statique.
Le programme a été développé en C++ avec la librairie Qt pour l’interface graphique et Eigen pour la résolution de système d’équations.

Mur d’une villa avec porte-à-faux, câbles de précontrainte et ouvertures

Thomas Kilian – Simulation of crystalline Si-solar cell

About CSEM


Description

My internship took place at CSEM Neuchâtel in the photovoltaic division under the supervision of Dr. Jacques Levrat, R&D Senior engineer and specialized in advanced characterization of crystalline-Silicon solar cell.

CSEM is an applied research and development center specialized in micro and nanotechnology, microelectronics, communication technologies, system engineering and photovoltaics.

 Photovoltaic is an actual and continuously growing domain of research in the field of renewable energies. In order to make it competitive with other technologies, the highest efficiencies must be reached at the lowest possible costs. This can be obtained by reducing the resistive losses and optimizing the quantity of expensive metals in the solar module. One way to optimize the research process is the numerical simulation of new solar cell and module designs. The goal of this internship was developing a 2D simulation software able to predict the electrical losses at the cell but also at the module level. In the latter case, the interconnection design was one of the key aspects that had to be optimized.

The program developed during the internship relies on NM-SESES, a physical systems solver using Finite Element Methods (see Fig. 1). It allowed the modeling of front and back metallization of a solar cell and calculation of the IV-curve giving the different electrical parameters relevant for the development and characterization of real solar cells.

The obtained results have been compared with other simulation tools and experimental results, assessing the accuracy of the newly developed solution in most of the cases.

The second part of the internship was dedicated to the upgrade of a characterization system (relying on photoluminescence and electroluminescence), aiming at calculating the maps of the cell electrical properties (see Fig. 2).

 Finally, this internship was a great opportunity to work in a very good atmosphere and take part to experimental realizations of solar cells and modules.

Fig. 1: Interface of developed program giving potential mapping, IV curve and cell parameters.

 

 

 

Fig. 2: PL imaging of a solar cell