Master and Diploma Projects
For excellent students we also offer projects which should lead to journal publications.
The chair of Fluid Dynamics also supports students who want to arrange their projects in cooperation with industry - provided that there are no non-disclosure agreements.
To top students, we also offer external Master’s projects with friends and colleagues at other universities, for example in Trondheim (Norway), Newcastle (UK) or Berkeley (US). Some prior work as a HiWi with us is a prerequisite, so we can make sure to only send out excellent students.
On request outstanding students can also realize important projects as Research Assistants.
Please contact: project.cfd [at] uni-due.de
Master and Diploma Projects which should lead to publications
200127: Master's Project: Numerical Investigation of Engine Knock
Engine knock remains a little understood phenomenon and represents a major unsolved problem in the context of ICE (internal combustion engines). It prevents the use of an engines optimal state of load, thus lowering the efficiency of the engine. Pre-combustion knock (or super-knock, a local premature ignition) is especially dangerous, as it may result in the complete failure of the engine. The aim of this thesis is to numerically investigate the physics of knock for a detonation wave propagating at Chapman-Jouguet speed. This detonation wave exhibits an unstable wave front, including the stochastic appearance of high momentum jets, which need to be analyzed. The simulations will be carried out with in-house code PsiPhi, whereby the setup of this type of simulation must be implemented and suitable ways to post-process the data must be found.
The project is challenging and requires strong skills in the field of fluid dynamics, thermo dynamics, reaction kinetics and numerics (FORTRAN, Python). It will be only attempted with very good candidates or long term HiWis.
190708: Master’s Project on analysis of simultaneous flame tomography and LIF imaging data
180619: Master’s Project on Camera Calibration
For a generic tomographic reconstruction algorithm highly accurate calibration of the cameras' intrinsic and extrinsic parameters (e.g. focal length and location) is necessary. Having sound parameters will help to decrease various artefacts and improve the reconstruction quality. The student will design a calibration target that is easy to manufacture and handle, but leads to a set of optimal parameters. The influence of the shape of the target on the parameters can be assessed by an existing in-house calibration algorithm. In a first step of the development process the student will become familiar with the code and test different target shapes. In a second step the target, based on the previous findings needs to be manufactured, e.g. with help of a 3D-printer. In the last phase of the project the target will be tested under real world conditions with our tomographic setup.
180523: Bachelor/Master Project: Background-oriented schlieren deflections using different patterns
Background-oriented schlieren is a flow v
isualisation method that detects the effect of variable refractive index on light rays. In a typical setup, a camera is focused onto a background pattern. When the refractive index between the camera and background is constant the light rays travel in straight lines, but when the refractive index varies (for example by introducing a flame between the camera and background), the pattern on the background gets distorted due to the refraction (bending) of light. The deflection magnitudes and directions can be calculated to give in indication about the refractive index field that the light rays have traveled through. The aim of this project is to investigate different background patterns, at different distances from the camera and write a code that calculates the deflections. The candidate must have a strong mathematical and programming background.
For further information please contact firstname.lastname@example.org.
180507: Master Project: 3D Imaging by Background-Oriented Schlieren Tomography
Background-Oriented Schlieren (BOS)
BOS is a diagnostic tool that is used to infer the refractive index n-field. Light travels in straight lines when n is constant within a region of interest. However, the light rays will bend if the refractive index changes along their path. A variable n-field can be generated by many sources, for example a flame. In a BOS experiment, a camera is focused onto a background image that contains a random pattern. A reference image of the background pattern is obtained and compared to an image of the pattern when a variable n-field is placed between the camera and background. Due to bending of the light, specific points on the background image will exhibit deflections which can be calculated from the image pairs (reference and deflection).
We are performing flame BOS measurements using multiple cameras that are focused onto multiple patterns, in order to reconstruct the 3D refractive index field within the flame. The aim of this project is to develop an algorithm that calculates the deflections by using different types of backgrounds. Both the pattern on the background and the plane shape, e.g., flat or curved surfaces, should be investigated. The results of the investigation should be discussed in terms of best sensitivity. This project will combine simple experiments with programming. Therefore, the candidate should have a solid programming knowledge, and any previous experimental work will be of benefit. The possibility of publishing the results in a peer-reviewed journal exists.
For further information please contact email@example.com.
171006: Bachelor/Master project: Flame bullet time movie production
The aim of this project is to produce a bullet time movie of a highly turbulent swirl flame. The bullet time effect is often used in the film and video games industry to elaborate on high-speed movements such as a moving bullet or the famous scene in The Matrix Trilogy movie (bullets flying over the actor as he bends backwards). Several cameras are arranged around a fast-motion target such as a flame in any chosen path (most obvious and simplest being a half or full circle). Controlled triggering of the cameras with a sequential time delay between them allows many images to be obtained, which when put together will generate a movie of the flame’s motion with an effectively super high frame rate. The resulting movie would pan around the flame, showing details in the motion which would be have been impossible to achieve using just one of the cameras in the set used. Interpolation software can be used to make the transition between the images occur smoothly. A total of 32 CCD cameras are available, an example of 24 of them arranged around a swirl burner is shown in Fig. 1. The candidate must have an interest in laboratory work and good programming skills are essential. Experience in technical lab work, image processing, scientific imaging or movie production will also be beneficial.
Please contact firstname.lastname@example.org for further information.
Fig. 1: 24 cameras, with filters, arranged around the swirl burner
170927: DNS and LES of lifted turbulent flames (HiWi/Bachelor/Master)
The correct simulation of the stabilization of lifted turbulent flames is still a massive problem. In the present project, DNS and LES are used to investigate how well-established models describe the flame lift-off - and what needs to be done to improve the prediction. For this purpose, two-dimensional simulations shall be carried out, in which a flame is stabilized on a flow body. Direct numerical simulations represent the "reference", which is then to be followed by simple 2D DNS. The deviation between the simulations corresponds to the model errors, which must be compensated for. One of the CFD codes OpenFOAM or PsiPhi (in-house) will be applied. Knowledge of commercial packages for flow simulation (Ansys, CFX or similar) is therefore not a significant advantage. The work requires a very good student with strong background in flow mechanics and no fear of programming tasks or the use of high-performance calculators. The work will be published in international journals.
170605: Large-Eddy Simulation of non-linear Thermoaccoustics (HiWi/Master)
Low emission (lean premixed combustion) gas turbine engines are prone to thermo-accoustic instabilities that can lead to the destruction of engines, which must be avoided, necessitating a detailed understanding of the phenomenon from simulation. Such Large-Eddy Simulations will be conducted using the in-house code PsiPhi, which is suitable for efficient, massively parallel simulations on super-computers. Recent work by our research group has shown that the normal, linear treatment of wave-propagation may not be sufficient at high frequencies of the instability, and we aim to simulate the phenomenon at high pressure levels, where even (non-linear) shock waves may occur. The present project aims at analyzing the phenomenon and preparing further research into this new topic. The work will involve setting up simulations, processing and interpreting the data, and writing a suitable report. A short version of the thesis shall be turned into a research paper that can be submitted to an international journal. A suitable student will have a strong background in fluid mechanics, mathematics and computer programming, and will be open to learning about new phenomena. The present project can be conducted in collaboration with researchers from the Technical University of Munich (TUM). A HiWi job on related topics can be offered prior to the Master’s project, using a closely related software.
170604: Direct Numerical Simulation of Solid Particle Combustion (HiWi/Master)
Solid fuel particles are burned in coal boilers, solid rocket engines, bio mass combustors and similar systems. A problem with understanding this type of combustion is the poor optical access, so that suitably detailed experiments cannot be performed to improve combustion, to understand the physics, or to reduce the emissions produced in such a process. The present work will aim at shedding light on such cases, by simulating the combustion of particles (coal, which is well characterized) by direct numerical simulation. Such simulations resolve the turbulent flow, chemical reactions, species transport, heat release, radiation and their interactions, producing large datasets that can be mined for understanding the process and for building models that can be used in engineering applications. The DNS will be conducted by the project student, using the inhouse code PsiPhi and high-performance computing facilities. The work will involve setting up simulations, processing and interpreting the data, and writing a suitable report. A short version of the thesis shall be turned into a research paper that can be submitted to an international journal. A suitable student will have a strong background in fluid mechanics, mathematics and computer programming, and will be open to learning about new phenomena. The present project can be conducted in collaboration with researchers from Trondheim University (Norway).
A HiWi job on related topics can be offered prior to the Master’s project, using a closely related software.
170531: Master Project: Large-eddy simulations investigating thermo-acoustic combustion instabilities in gas turbines with different combustion models
To comply with regulatory emission limits, lean premixed combustion is applied in gas turbines, which leads to low flame temperatures and hence low NOx emission. However, lean premixed combustion systems are prone to unsteady heat release, which can trigger thermo-acoustic instabilities. The interaction between pressure waves and unsteady heat release drives the oscillations, resulting in noise and vibrations that can induce severe damage to the engine. For this reason, analyses of the thermo-acoustics are essential for improving gas turbine combustion systems.
An advanced combustion model has been developed to improve the prediction of CO and NOx emissions and of thermo-acoustic instabilites, which must be tested and compared to available models. The candidate must first familiarize himself/herself with the CFD code “OpenFOAM” and the combustion models for reacting compressible flow. The aim of the project is to analyze the sensitivity of the models towards thermo-acoustics. For this purpose, the flame is excited artificially by velocity perturbations, the response of the flame is then examined to determine the flame transfer function (FTF). The prediction of the flame dynamics using different combustion models will be validated against experimental data. The result of the Master’s thesis will be a validated combustion model implemented in “OpenFOAM”, permitting to predict both emissions and thermo-acoustics in gas-turbine combustors.
The project will require a motivated student with a strong background in computational fluid dynamics, combustion and computer programming.
For further information please contact: pascal.gruhlke[at]uni-due.de.
170224: Master Project: Tomographic reconstruction of stratified flames: DNS phantom burner
The aim of this project is to produce a 3D phantom, that is an exactly known field, of a stratified flame using the existing high quality DNS data of the flame, which was generated by our in-house code PsiPhi. Projections of the phantom are to be used with our in-house Computed Tomography of Chemiluminescence (CTC) algorithm to reconstruct the flame. The original and reconstructed phantoms can then be cross-correlated to quantify the accuracy and quality of the reconstructions. The DNS phantom is considered as a ‘numerical equivalent’ of the real flame, and will thus contain a large amount of flame structures. The goal is to assess how much of the structural detail can be recovered with our CTC algorithm, and to what accuracy.
This project is ideal for students of engineering, computer science or related subjects. Very good programming and mathematical skills, as well as English language are required. The project can be undertaken as a masters or HiWi position. The results from this project are to be published in a peer-reviewed journal.
161220: Master Project: Camera calibration method using 3D targets
The calculation of camera location in 3D space is applicable to a range of fields including 3D computer vision and tomographic reconstruction based on multiple camera images. Different methods include obtaining images of a target, which may be 1D, 2D or 3D, with known geometry. An algorithm then calculates the camera intrinsic and extrinsic properties. Existing literature shows that techniques using 3D targets tend to give higher accuracy than using 2D targets for example. In this project, the aim is to produce a camera calibration technique using 3D targets and to quantify the error in the camera location based on controlled tests. The work shall be presented in a journal publication and the method will be used by the group's projects that involve tomographic reconstruction using multi-camera setups. The candidate should have very good numerical skills and knowledge of programming. Some past experience in photogrammetry or related subjects will be of tremendous benefit.
For further information please contact: email@example.com
160811: Numerical Simulation of Primary Jet Breakup
Fuel injectors operate at high pressures and inject fuel into a combustion chamber, where the jet breaks up and eventually forms droplets that break into further droplets. The present project aims at researching methods for the simulation of multi-phase flows with large density differences and surface tension, choosing the most promising technique, implementing it into the in-house code PsiPhi and running test-simulations with it. Applications of the work extend from liquid jets in a gas-phase to bubbles raising in liquids. The project is very challenging and will only be attempted with very good candidates or HiWis.
For further information please contact: andreas.kempf [at] uni-due.de
161108: Optimised Numerical Schemes for the Large-Eddy Simulation of Turbulent Combustion
Large-Eddy Simulation is a modern CFD technique for accurately predicting turbulent reacting flows by affordable computer simulation. The method has evolved over the last 15 years and is finally becoming available in commercial software programs and being used by industry leaders.
However, the method requires numerical schemes that combine high numerical accuracy with low numerical oscillation. Our group has used a hybrid approach with good success, combining accurate central differencing schemes for momentum transport with non-dispersive TVD schemes for scalar transport. Where this hybrid approach combines good accuracy with low dispersion, it can lead to inconsistencies when applied with certain combustion models based on the "Flames Surface Density" approach.
The present project will apply different combinations of available numerical schemes to different test cases, to eventually assess the overall error resulting from the schemes. Based on the findings, further transport schemes (e.g. (W)ENO) shall be implemented and tested, aiming to improve overall accuracy, reliability and robustness of the simulations.
Students interested in this project will require a strong background in fluid-mechanics and should ideally have some background in numerical techniques, programming, combustion and turbulence modelling.
Please contact andreas.kempf [at] uni-due.de for further information.