Tomographic optical soot diagnostics in unsteady flames

点击翻译按钮获取中文摘要

Although great efforts are being made for the electrification of mobility and decarbonisation of energy conversion, the combustion of fossil and other carbon-based fuels will be a major sector of energy conversion in both China and Germany for many years to come. In this context, the reduction of soot emissions from combustion processes, both in terms of adverse health effects and the significant contribution to global warming, is a major challenge.Despite great progress in modelling, there is still no complete understanding of soot formation in turbulent combustion processes, as is, e.g., relevant for combustion engines. A main challenge in contributing to a better understanding is the necessity of a temporally highly resolved detection of complex and fluctuating soot distributions. This project addresses this challenge by enabling a four-dimensional (3D spatially and temporally resolved) determination of essential soot characteristics. The collaborative approach of the groups in Erlangen and Shanghai, whose expertise ideally complements each other, forms a most promising basis for the development of this advanced diagnostic tool.The overall objective of the proposed research project is the development of a method with which characteristic parameters of soot (concentration, primary particle size, temperature) can be measured with high temporal resolution (kHz) in their full three-dimensional structure. For this purpose, the methods of two-color pyrometry (2CP) and time-resolved laser-induced incandescence (TiRe-LII) with volumetric illumination will be further developed and combined with advanced tomographic reconstruction methods. A specific feature of the approach is the realization of a set-up with minimized experimental expenditure, performing simultaneous measurements of 2CP and TiRe-LII with three or four high-speed cameras in total, which is achieved through the use of customized fiber bundles.The project uses well-defined steps, closely coordinated between experimental and numerical studies. Starting from a reference flame with well-known characteristics, the method can finally be applied to a transient diffusion flame.For a fast and efficient evaluation of the large amounts of data, both the algorithms for LII and tomographic evaluation will be further developed including the use of convolutional neural networks. At the end of the project, a measurement system and evaluation technique will be available with which reference data can be obtained for different flame types to understand soot formation and oxidation and to validate numerical methods.