Universität Wien

520026 VO Atomistic materials modelling (2022W)

electronic structure methods, statistical mechanics and machine learning

3.00 ECTS (2.00 SWS), SPL 52 - Doktoratsstudium Physik

Registration/Deregistration

Note: The time of your registration within the registration period has no effect on the allocation of places (no first come, first served).

Details

max. 15 participants
Language: English

Examination dates

Lecturers

Classes (iCal) - next class is marked with N

  • Thursday 13.10. 15:00 - 16:30 Seminarraum 18 Kolingasse 14-16, OG02
  • Thursday 20.10. 15:00 - 16:30 Seminarraum 18 Kolingasse 14-16, OG02
  • Thursday 27.10. 15:00 - 16:30 Seminarraum 18 Kolingasse 14-16, OG02
  • Thursday 03.11. 15:00 - 16:30 Seminarraum 18 Kolingasse 14-16, OG02
  • Thursday 10.11. 15:00 - 16:30 Seminarraum 18 Kolingasse 14-16, OG02
  • Thursday 17.11. 15:00 - 16:30 Seminarraum 18 Kolingasse 14-16, OG02
  • Thursday 24.11. 15:00 - 16:30 Seminarraum 18 Kolingasse 14-16, OG02
  • Thursday 01.12. 15:00 - 16:30 Seminarraum 18 Kolingasse 14-16, OG02
  • Thursday 15.12. 15:00 - 16:30 Seminarraum 18 Kolingasse 14-16, OG02
  • Thursday 12.01. 15:00 - 16:30 Seminarraum 18 Kolingasse 14-16, OG02
  • Thursday 19.01. 15:00 - 16:30 Seminarraum 18 Kolingasse 14-16, OG02

Information

Aims, contents and method of the course

The lecture will discuss state of the art methods in soft and condensed matter simulations using atomistic modelling methods. This includes an introduction to electron structure theory with special focus on Kohn-Sham methods and the projector-augmented-wave method, including algorithms to determine the electronic groundstate. In the second part, machine learning methods are presented. These allow energies and forces to be learned from such first-principles calculations, thereby speeding up calculations by many orders of magnitude. Both kernel methods and neural network methods (perceptrons) are covered. Finally, applications of these methods to the computation of thermodynamic properties, such as the free energy and various autocorrelation functions, will be addressed.

The goal is to cover a comprehensive set of topics that allow one to understand the concepts typically used in large-scale simulations of soft and condensed matter. In particular, the focus is on simulating observables at finite temperature and predicting phase transitions using atomistic methods, as well as bridging the gap between electronic structure theory and thermodynamics, while always retaining first principles accuracy.

Tentative schedule (cF Cesare Franchini, gK Georg Kresse, cD Christoph Dellago)
Donnerstag 13.10.2022 15:00 -16:30 Seminarraum 18 Kolingasse 14-16, OG02 02.20 Basic intro to electronic structure theory cF
Donnerstag 27.10.2022 15:00 -16:30 Seminarraum 18 Kolingasse 14-16, OG02 02.20 Kohn-Sham methods and the variational principle cF
Donnerstag 03.11.2022 15:00 -16:30 Seminarraum 18 Kolingasse 14-16, OG02 02.20 LCAO and PW cF
Donnerstag 10.11.2022 15:00 -16:30 Seminarraum 18 Kolingasse 14-16, OG02 02.20 The PAW method gK
Donnerstag 17.11.2022 15:00 -16:30 Seminarraum 18 Kolingasse 14-16, OG02 02.20 Finding the groundstate efficiently gK
Donnerstag 24.11.2022 15:00 -16:30 Seminarraum 18 Kolingasse 14-16, OG02 02.20 Kernel based methods gK
Donnerstag 01.12.2022 15:00 -16:30 Seminarraum 18 Kolingasse 14-16, OG02 02.20 Neural Network Potentials cD
Donnerstag 15.12.2022 15:00 -16:30 Seminarraum 18 Kolingasse 14-16, OG02 02.20 Molecular dynamics; The free energy: TPD and TI cD
Donnerstag 12.01.2023 15:00 -16:30 Seminarraum 18 Kolingasse 14-16, OG02 02.20 The classical fluctuation dissipation theorem cD
Donnerstag 19.01.2023 15:00 -16:30 Seminarraum 18 Kolingasse 14-16, OG02 02.20 Autocorrelation functions and transport properties cD

Assessment and permitted materials

Oral examination. Duration approximately 30-45 minutes. Students must select as their central examination subject all topics covered by one of the lecturers. In addition, knowledge of the topics covered by the other lecturers must be demonstrated.

Minimum requirements and assessment criteria

The core subject accounts for 70% of the points that can be achieved. 30% of the points to be earned are on topics outside the core subject. 50% of the points are required to pass the exam.

Examination topics

As described above.

Reading list

Literature will be distributed during the lecture. This includes electronic slides used in lecture as well as handwritten notes.

Association in the course directory

M-ERG

Last modified: We 13.12.2023 15:47