Differential diffusion in Multiple Mapping Conditioning (MMC) Model

Abstract

This thesis describes the recent development in the Multiple Mapping Conditioning (MMC) method for non-premixed turbulent combustion focussing on differential diffusion effects. Turbulent combustion is a considerable issue in different engineering fields during the last decades. Nowadays, there are more global concerns on sustainability, environmental impacts and efficiency of fuels. These challenges have motivated additional research to maximize the efficiency and reduce pollutants of future combustion systems.Modelling of turbulent combustion is a complementary approach to experimental analysis of combustors. In general, mixture-fraction based methods and joint Probability Distribution Function (PDF) methods are two main categories for turbulent combustion modelling. MMC, which is the subject of this thesis, combines the useful features of the two aforementioned categories and is applicable to both premixed and non-premixed combustion. One of the great advantages of MMC is localising mixing within an independent reference space, which enforces localness of mixing in the composition space.Differential diffusion effects due to differences in molecular diffusivity of species are neglected in most turbulent combustion models for simplification. However for the cases of fuels containing highly diffusive species (e.g. hydrogen), ignoring these effects leads to errors evidenced by many experimental and numerical works in literature. Hydrogen and hydrogen-enriched fuels are of interest as an alternative to fossil fuels, which can address environmental concerns. Therefore, combustion models need to be improved to include differential diffusion.In the present thesis, two MMC models are suggested and implemented for emulating differential diffusion effects in a homogeneous turbulent non-reacting flow. A side stepping approach is developed to obtain benefits of MMC in conjunction with and theoretical analysis of differential diffusing scaling of physical parameters of flow. This approach accounts for differential decay rate of scalar variances, which is one of the key effects of differential diffusion. It is shown that this novel MMC model can also emulate the decorrelation of scalars, which is a more refined differential diffusion effect. Results indicate good agreement with the previous DNS studies.