Course content

Experimental or numerical investigations of thermo-fluid systems typically lead to large amounts of data with high resolution in space and time. Extracting relevant and useful insight from these data often turns out to be as challenging as generating it. The course introduces classical and modern concepts and methods of data analysis that are particularly relevant to thermo-fluid systems, such as elementary signal and image processing, discrete Fourier analysis, coherent structures, vortex identification, principal component analysis, dynamic mode decomposition. Elements of data visualization and presentation.

Learning outcome

The course provides the participants with access to fundamental and advanced techniques of data analysis relevant to thermo-fluid systems. It enables them to choose and apply suitable methods for the analysis of experimental or numerical data in the context of a given research question. The course participants acquire the theoretical background of the covered methods and the skills for implementing them in a suitable numerical framework. They will be capable of applying the methods and understand their theoretical foundation.

Knowledge:

- signal and image processing

- applied Fourier analysis- structure identification

- orthogonal decompositions

Skills:

- selection of appropriate data analysis methods

- implementation of methods into suitable numerical framework

- accurate and clear representation of experimental and numerical data

General competence:

- understanding of the crucial role data analysis plays in the investigation of thermo-fluid systems

- improved proficiency in using a data processing application

- critical assessment of data representation

- preparation and delivery of a technical presentation

Learning methods and activities

The classes are roughly split 50/50 into theory and application sessions. In the theory sessions, the analytical and numerical background for the various methods is introduced and illustrated based on elementary examples. In the application sessions, the course participants implement and apply the learned techniques to experimental and numerical data from our group (for example, pressure signals, velocity fields, flame images). Problem sets are distributed throughout the course that challenge the participants' skills and creativity in extracting relevant information from synthetic and real data sets. The course participants work on a semester project based on data from their own PhD work. The details are to be discussed at the beginning of the course. The project results will be presented in a conference-style talk.

The course is taught in English.

Compulsory assignments

• Semester project

Specific conditions

Recommended previous knowledge

Solid knowledge of fluid dynamics and engineering mathematics, in particular, linear algebra; basic knowledge of a scientific scripting language such as Python or Matlab.

Required previous knowledge

Requires admission to study programme: Engineering (PHIV)

Course materials

Lecture notes, excerpts and code snippets distributed in class. Course material will be made available on Blackboard. All course material is in English.

Credit reductions