Contact
Engler-Bunte-Ring 7
76131 Karlsruhe
Building number 40.13.I
Tel: +49(0)721 608-42571
Fax: +49(0)721 608-47770
E-Mail: Secretariat
| The majority of the research currently being undertaken in the field of calculating flow processes focuses on the investigation of turbulent flows. Nearly all naturally occurring flows are turbulent, and therefore it is important to be able to accurately describe turbulent flows with models. Many model descriptions have been developed for this purpose in the past, and Large Eddy Simulation is one such model. The three most important types of method for describing turbulent flows are Direct Numerical Simulation (DNS), Large Eddy Simulation (LES) and modelling with equations for (several) statistical moments of flow characteristics such as velocity (e.g. the k-epsilon model). Whilst the method of moments can be very accurate under certain conditions, it is often not suitable for transient flows. This is because the averaging process leaves out most of the important characteristics of a time-dependent solution. Direct Numerical Simulation, on the other hand, attempts to solve all temporal and spatial scales. As a result, the solution is very accurate. Unfortunately, DNS is unrealistic for 99.9% of all CFD problems because it is computationally unattainable at the present time. A compromise between these two methods is Large Eddy Simulation. This technique was originally introduced by scientists in the 1970s to describe atmospheric transport processes. Since then it has been introduced into almost all areas of engineering science and is state of the art. Large Eddy Simulation seeks to directly solve large spatial scales (as with DNS) while modelling the smaller scales (analogous to k-epsilon). The reason for this is two-part: Firstly, the larger scales carry the majority of the flow energy, and hence are always more important. Secondly, the smaller scales have far less of a prominent direction or none at all and thus are usually easier to model. The resulting methodology is a hybrid between these two methods. |
Within this research focus in the past the following research projects were associated: |
PEGASUS will investigate a novel power cycle for renewable electricity production combining a solar centrifugal particle receiver with a sulphur storage system for baseload operation. The proposed process combines streams of solid particles as heat transfer fluid that can also be used for direct thermal energy storage, with indirect thermochemical storage of solar energy in solid sulphur, rendering thus a solar power plant capable of round-the-clock renewable electricity production.
 Process scheme of the solar sulphur cycle / Image source: DLRThe overall objective of PEGASUS is the development and demonstration of an innovative solar tower system based on solid particles combined with a novel thermochemical solar energy storage technology based on elemental sulphur, to achieve dispatchable and firm renewable electricity generation with a significant cost reduction with respect to current state-of-the-art concepts. The technology will be validated under real on- sun concentrated solar irradiation in the Solar Tower Jülich (STJ) thermal plant in Germany owned by the Project Coordinator, DLR.
In this perspective, the project’s specific Technical Objectives of KIT are:
More information is published in a press-release of KIT and on the public website of the project (link below) |
The energy sector accounts for two thirds of the global CO2 emissions and is therefore crucial to ensure future green growth and to achieve the global emission reduction targets. Substantial reduction of CO2 emissions can only be achieved by large scale deployment of renewable energy sources, including in particular the most abundant energy sources, wind and sun. Their intermittent nature however poses significant challenges for the energy system as peak demand from the system and peak production form those intermittent sources do not overlap. As there are no large scale storage solutions available yet, other backup capacities are needed. The installed fossil capacity is large enough to provide this back-up power. However, the plants were designed for baseload operation, which results in increased wear and costs through cyclic operation and unnecessarily high emissions in the start-up phase. Providing technology upgrades to retrofit the installed power plants to enable flexible operation without penalties on life, cost and emissions is an opportunity to quickly provide the necessary backup capacity to keep the energy system stable and resilient and at the same time enabling higher renewable shares. The mission of TURBO-REFLEX is the development and optimisation of technologies, applicable to a selected set of turbomachinery engine components, which can be used to retrofit existing power plants as well as new machines in order to enable more flexible operation, providing the flexible back-up capacity needed for introducing a larger share of renewables in the energy system. TURBO-REFLEX will assess the impact of such technologies at plant level and prepare the transfer of component technology gains into reduction of both (unplanned and planned) outages and maintenance and operation costs.
The Lean Blow Off (LBO) limit is a significant hurdle to further reduce the part load of gas turbines as the operating zone of the combustor is restricted by the LBO limit. Jet stabilized premixed flames will be forecasted with blow off stability down to 1000°C–1200°C combustion temperature with or without using pilot flames. 1000°C–1200°C combustion temperature would equal emission compliant part load operation down to 20%-25%. Better blow-off stability in the combustor is a prerequisite to running higher load gradients. Therefore, it is expected that jet stabilized premixed flames with better LBO limits will allow also gradients faster than 40MW/min.
|
|
EBIvbt at KIT will model the blow-off of jet flames with advanced computational models. These models include basic geometrical parameters like duct diameter and dump ratio, but also the effect of neighboring pilot flames. A 3D simulation model will be developed and experimentally validated at conditions close to the application. The turbulence-chemistry interaction will be captured by two different combustion models. Within both models a transport equation of a reaction progress variable will be solved. The difference between the combustion models is in the source term modelling. In the first model the source term depends on mixture fraction and the progress variable itself. In the second model, which is based on the “turbulent flame speed closure” approach, the source term depends on the laminar flame velocity. So, one can calculate the influence of the stretch and heat loss on the laminar flame speed by simplified 1D modelling. The comparison of the two models to experimental data will show which model is more suitable for the applied boundary conditions.The current project aims to calculate quantitatively the generation mechanism and propagation of noise caused by turbulent combustion. Fully compressible Large Eddy Simulations (LES), hybrid CFD (Computational Fluid Dynamics)/CAA (Computational Aero-Acoustics) and Direct Numerical Simulation (DNS) are chosen as computational tools. The UTFC combustion model based on a TFC (Turbulent Flame Speed Closure) approach and a FGM (Flamelet Generated Manifold) tabulation procedure has been developed during this project which is used together with compressible LES for turbulent flames. This enables the direct computation of combustion noise from the resolved large scale fluctuations. The DNS will then be applied for investigation of noise within the sub grid scale. The project is connected with three other project partners from TU Darmstadt (Prof. J. Janicka), TU AAchen (Prof. W. Schröder) and TU Berlin (Prof. C.O. Paschereit) who make inkompressible LES, CAA-computation and measurement for the same burner configuration. |