Quantitative Developer - C++ / Rust

Epam; Paris; Sur place; EUR 60 000 - 80 000; Plein temps

Résumé du poste

Une entreprise innovante recherche un candidat pour un doctorat passionnant en physique computationnelle et science des matériaux. Ce projet de recherche se concentrera sur l'étude des motifs de croissance eutectique à travers des simulations numériques avancées et des expériences en microgravité. Le candidat travaillera sur des modèles de phase-field et portera des codes vers des architectures modernes, tout en analysant des données expérimentales précieuses. Avec un encadrement de chercheurs expérimentés, cette opportunité offre une chance unique de contribuer à des découvertes fondamentales dans le domaine des systèmes dynamiques dissipatifs. Si vous êtes passionné par la recherche et l'innovation, ce poste est fait pour vous.

Qualifications

• Expérience en modélisation numérique et en physique computationnelle.
• Capacité à travailler avec des systèmes de simulation avancés.

Responsabilités

• Réalisation de simulations numériques de solidification eutectique.
• Analyse des images optiques en temps réel des motifs de croissance.

Connaissances

Physique computationnelle

Science des matériaux

Modélisation numérique

Analyse d'images

Formation

Doctorat en physique ou science des matériaux

Outils

Phase-field models

GPU processors

Description du poste

25-081 Numerical simulation and in situ observation of eutectic growth patterns

Mission

Solidification microstructures in alloys are a frozen trace left behind in the bulk material by remarkable shapes and patterns that arise at the moving interface between the solid and the liquid mixture. The dependence of basic solidification microstructures on alloy characteristics and control parameters, and the relevant spatiotemporal scaling behaviour, are well established. However, the sensitivity of the emerging phenomena to initial and boundary conditions makes solidification intrinsically complex – difficult to predict and control – thus motivating an inspiring research with shared interest in materials science and the nonlinear physics of out-of-equilibrium systems.

Eutectic solidification is paradigmatic in this context. A binary eutectic directly solidifies into a two-phase solid with composite microstructures. In steady state, directional solidification at an imposed velocity in a fixed temperature gradient typically delivers lamellar or rod-like microstructures in the bulk. Both of these patterns are prone to morphological instabilities, which have been intensely studied under the light of the phenomenology of symmetry-breaking instability modes. So far, however, the so-called lamellar-rod transition has resisted standard analysis. There is evidence that a transformation from a lamellar morphology to a rod-like one, and vice versa, can occur in a given system upon mild variations of the experimental parameters. Information on this dynamic bistability still remains partial, and key questions on the nature of the underlying instability modes, the role of propagative processes, the possible coexistence of the two types of structures, and the relevant scaling parameters are left pending.

The PhD work will be based on a research methodology that combines time-resolved numerical simulations and in situ experiments. On the numerical side, so-called phase-field models have been developed since two decades that are now capable of reproducing the pattern-formation processes during solidification in extended three-dimensional systems. On the experimental side, the study will be directed to the analysis of several series of real-time observations obtained during recent science-in-microgravity campaigns carried out in the International Space Station (ISS), in the framework of a project led by the European Space Agency (ESA).

Simulations and theory will be compared with real-time optical images of the growth front patterns during directional solidification of model transparent alloys. For this purpose, work on two main axes is needed. First, the existing phase-field codes will be ported to modern computational architectures, in particular GPU processors, and three-dimensional simulations of eutectic solidification will be carried out, for alloy parameters and processing conditions that correspond to the experiments. Second, numerical image processing and analysis methods shall be developed, with the aim to define and extract meaningful geometrical parameters (size distribution, shape correlations) from the images of spatially extended hybrid patterns. The image analysis will serve for both experimental and numerical data. Depending on the progress of the study, additional solidification experiments using different model alloys could be carried out at INSP. On a fundamental level, all of these elements will cast new light on the large-scale evolution of ordered vs disordered structures in dissipative dynamic systems.

The thesis will take place on the sites of Sorbonne University (Paris) and the Ecole Polytechnique (Palaiseau), and will be co-supervised by S. Akamatsu (INSP, Paris) and M. Plapp (LPMC, Palaiseau).

Profil

Computational physics - Materials science