ALIs

kommt noch

Achtung: Die Navigationslinks auf dieser Seite funktionieren nicht, weil das nicht der Rahmen ist, der zu diesem Skript gehört. Wenn das Skript ordnungsgemäß dort installiert ist, wo es später laufen soll, funktionieren auch die Links in der dortigen Umgebung.

LINUX Cluster Project

Thermodynamik und Wärmeübertragung


Institution

  • Name: Lehrstuhl für Thermodynamik
  • Address: Boltzmannstraße 15, 85748 Garching
  • Project Proposal Date: 2019-12-06 11:29:27

Abstract:

1.Experience shows that flame flashback in the flow boundary layer, one of four identified mechanisms for flame flashback in typical gas turbine burners, can commonly be regarded negligible for the combustion of natural gas. However, the substitution of natural gas by hydrogen significantly changes the flame propagation speed and thereby increases the risk of flashback into the fuel supply gear of lean premixed combustors. In order to adapt existing burner design rules to the demands of hydrogen-rich fuels, a database is desirable which correlates flashback limits, flow Reynolds number and turbulence conditions, mixture composition and the thermodynamic state of the mixture. The generation of such a database as well as more detailed investigations on near-wall flame propagation will be enabled by a dedicated new combustion experiment. The design of the flowpath within the combustion chamber is a complicated task due to specific requirements to the boundary layer development and wall shear stress distribution. Since an adverse pressure gradient as well as boundary layer suction devices are present, the application of simplified correlations is not an option here. Instead, CFD simulations with the solver Ansys CFX 11.0 have been used for the design exploration. The resulting final design comprises a rectangular, asymmetric diffusor geometry with two subsequent backward-facing steps for flame stabilization. 2.Thermoacoustic instabilities can lead to very high levels of pressure fluctuations, resulting in structural damage of the combustor. For this reason, the correct description of the flame dynamics has technological importance nowadays. Experimental investigations on industrial gas turbines with multiburner annular configurations are very costly, and tests are often carried out in a single burner configuration. To analyze the stability of a combustion system, it is necessary to know how the heat release of a flame responds to perturbations of velocity or pressure by the determination of its Flame Transfer Function. The Flame Transfer Function (FTF) describes the acoustic properties in terms of amplitude and phase of a flame. A way to characterize a system is to establish the frequency dependent relationship between the signals before and after the system. This can be obtained performing an unsteady CFD simulation, and then using system identification (SI) methods to reconstruct the flame transfer function. Large Eddy Simulations (LES) has become a powerful tool in the study of flame dynamics, where the flame/acoustics interaction is well predicted. In the project, LES/SI methods will be used to identify the flame transfer function in a single burner and in an annular combustor. LES/SI methods consists of a two step procedure. At first, a LES simulation of the system under consideration is set up. After obtaining a statistical stabilized solution, the system is excited with broadband noise superimposed on the mean flow. Therefore, the variations of the characteristic wave amplitudes at the inlet are chosen as excitation perturbations. These perturbations will propagate to the flame front and create a response in the heat release of the flame. The velocity fluctuations and the total heat release of the flame are exported every time step and then imported in a post-processor based on the correlation based Wiener-Hopf-Inversion technique to determine the flame response in the linear regime. The identification of Flame Transfer Functions in single burner combustors is done using plane waves where the non-axial components of acoustic velocity perturbations are zero, while in annular geometries also higher order with azimuthal velocity components acoustic waves exist which can influence the flame response. Therefore, the flame dynamics in both configurations are analyzed in the project to establish whether single burner results are applicable to annular multiburner configurations. 3.The project is a part of a big DFG project "Aero-thermodynamische Auslegung eines Scramjet - Antriebssystems für zukünftige Raumtransportsysteme". Here a combustion chamber, in which there's a flow in supersonic regime and a supersonic combustion takes place, should be analyzed. By the simulations of supersonic flames are really important the ignition limits and the quenching phenomena. For this reason is a detailled chemistry necessary to take into account all the phenomena that could led to the quenching of the flame. The right ignition behavior cannot be estimated only by all the relevant species, but also in turbulent flows by the fluctuations of temperature and species. In a supersonic flow also the fluctuations of the density should be taken into account. Due to the big influence on the combustion, the fluctuations should be predicted as well as possible. Through a LES simulation a jump in the simulation of the fluctuations can be expected. The first step is to produce an accurate simulation of the cold flow to analyze the mixing phenomena. Later a chemistry module, that takes into account a reduce kinetic mechanism, should be added to simulate the combustion. 4.Im Rahmen eines GRS (Gesellschaft für Reaktorsicherheit) - Projekts soll das Gefährdungspotential von Wasserstoffreisetzung in Kernreaktoren untersucht werden. Während bisherige Untersuchungen auf die Verbrennung homogener Gemische in vollständig abgeschlossenen Geometrien beschränkt waren, sollen in diesem Projekt erstmals die Einflüsse von realistischen, vertikalen Wasserstoff-Gradienten auf die Flammenbeschleunigung und den Übergang in die Detonation untersucht werden. Außerdem sind die mitigierenden Effekte von partiell geöffneten Geometrien untersucht werden. Für eine realistische Beschreibung der Reaktionskinetik sowohl während einer Deflagration als auch während einer Detonation wird ein Verbrennungsmodell mit detaillierter Chemie verwendet. In Abstimmung mit experimentellen Untersuchungen sollen aus den Ergebnissen verlässliche Deflagrations-Detonations-Transitionskriterien ermittelt werden. With the development of lean- premixed combustion technologies thermoacoustic combustion instabilities have become a recurrent problem. In low order modelling of such instabilities [1] the Flame Transfer Matrix (FTM) relating the acoustical properties upstream and downstream of the flame is obtained starting from a Flame Transfer Function (FTF), that relates the acoustic fluctuations of velocity at the burner with the heat release rate of the flame, with the assumption that the flame is an acoustically compact element. This compactness assumption is only valid for Helmholtz numbers He = k.Lflame << 1, where k is the wave number and Lflame the flame length. This work a) extends the applicability of low-order models to non-compact conditions, and b) evaluates the damping potential of distributed heat release on the thermoacoustic stability of a combustion system. The first part of this work a) has been successfully validated with the academic test cases of Lieuwen and Zinn [Journal of Sound and Vibration, 2000], where the applicability of the compact flame assumption was first tested. The new low order model approach should also handle any heat release rates distribution, and the numerical study here proposed will provide realistic flame data, partially validated against experimental results obtained at the University of Cambridge in the framework of the ongoing AETHER/ EU project. The FTF and FTM describe the acoustic properties of a flame, e.g. u' and p', in terms of amplitude and phase. A way to characterize a flame is to establish the frequency dependent relationship between these signals before and after the flame. This can be obtained performing an unsteady CFD simulation, and then using system identification (SI) methods to reconstruct the FTF and the FTM. Large Eddy Simulations (LES) has become a powerful tool in the study of flame dynamics, where the flame/acoustics interaction is well predicted and, in this project, LES/SI methods will be used to identify both the FTF and the FTM. The LES/SI method consists of a two step procedure. First, an LES simulation of the system under consideration is set up. After obtaining a statistical stabilized solution, the system is excited with broadband noise superimposed on the mean flow. These perturbations will a) propagate to the flame front and create a response in the heat release of the flame and b) reach the region downstream of the flame. The velocity and pressure fluctuations at both upstream and downstream of the flame and the heat release fields of the flame are exported at every time step, pre-processed to 1D axial resolved data, and then post-processed with a Wiener-Hopf-Inversion correlation technique. This post- processing allows finding the FTF and FTM for linear regimes and for a chosen frequency range, providing both input data for the extended low- order model, i.e. the FTF, and the FTM used to validate the complete flame response obtained with the extended low order model. The project is focusing on optimization of the CBF (Characteristic Based Filtering) subroutine allowing separation of the acoustical and turbulent fluctuating components from the LES simulation. For testing purposes of the new approach I need an access to data (approx. 1.5 Tbyte) obtained and stored by user t7811aa (see project description for more information). 5. A CFD-code within the open source software OpenFoam should be developed that can predict quenching and re-ignition of a reacting jet. These two effects have a fundamental role in the evaluation of a stable combustion for the JIC and for the ignition process in a prechamber engine. Beyond this the pollutant formation is especially interesting in the context of gas turbines. Therefore the combustion model must capture the chemical process in a certain depth as well as turbulence effects. The code should be evaluated with experimental data produced at the TUM on suitable test rigs so that the final code can be used as a design tool in the development process.