Measurements and model development for flameless combustion in a lab-scale furnace

The technique called "flameless combustion", also denoted as "MILD" combustion, was developed to reduce the nitrogen oxides (NOx) emission in the combustion process. The term "flameless" refers to the low visibility of the flame. The technique is particularly of interest when hot exhaust gas is used to preheat inlet air to high temperature. The combination of flameless combustion and exhaust gas heat recycling techniques simultaneously reduces the emission and increases the energy efficiency. Over the past few decades, flameless combustion has been successfully applied to industrial furnaces or tested at pilot scale setups in other applications. Nevertheless, despite the successful industrial application, many fundamental issues of flameless combustion are still unresolved. Detailed measurements of flameless combustion have been performed in jet in- hot-coflow (JHC) flames, but it is unclear whether the findings can be related to the flameless combustion in a furnace because only part of the features of flameless combustion are mimicked in JHC flames. Concerning modelling, it is found that the existing combustion models are not suitable for numerical modelling of flameless combustion and new model development is needed. The objective of this research is to characterize the flameless combustion in a labscale furnace that is simple enough to allow detailed measurements while keeping most relevant characteristics found in large scale furnaces. This thesis is divided into two parts, experimental measurements and model development and validation. The goal of experiments is to observe the flame behaviour and obtain detailed velocity and temperature data of flameless combustion in the furnace by means of high speed imaging and laser diagnostic techniques. The goal of the model development is to extend the Flamelet Generated Manifolds (FGM) method to take into account the effects of dilution by recirculated burnt gases. One of the databases of the Delft jet-in-hot-coflow (DJHC) flames and a new database obtained in a new lab-scale furnace are used for the model validation.