Sudipta Saha

Date of Award

Spring 2023

Document Type

Open Access Dissertation


Mechanical Engineering

First Advisor

Tanvir Farouk


Gas-phase chemical reactions coupled with multidimensional fluid flow and heat and mass transport are found in various applications, i.e., from conventional engine applications to novel combustion techniques. With the goal of understanding such complex coupling in reacting flow systems, this dissertation work focuses on developing multi-physics simulation frameworks to investigate the effect of multidimensional transport on flame dynamics. This study primarily focuses on the modeling and simulation of low temperature flame formation in i) a canonical experimental setting with counterflow burners and ii) a supercritical water medium (i.e., hydrothermal flame).

In the first part of the dissertation, simulations of the cool flame formation of dimethyl ether (DME) combustion in a counterflow geometry have been presented. Historically, flame formation in counterflow geometry is studied to understand the strength of the flames under the effect of the stretch. The formation of hot flames in this arrangement has been studied profoundly over the last half-century. The counterflow/opposed-flow diffusion flame configuration allows one to implement a similarity solution approach, which makes the numerical modeling approach tractable and inexpensive [1]. This approach assumes the dependent variables to be a function of the axial distance only. Hence, the problem can be represented by a quasi-one-dimensional (1D) numerical model. This assumption is valid for high Reynolds numbers (i.e., high strain rate) and large activation energy (i.e., thin flame sheets), which are generally the case for a hot flame formation in counterflow geometry. However, this similarity solution approach becomes questionable for a low-temperature cool flame where the strain rate is relatively low, and the flame achieves a finite thickness. Therefore, multidimensional numerical analyses are required to simulate the cool flame formation in a counterflow geometry. In this dissertation, a two-dimensional (2D) axisymmetric multi-physics numerical framework has been proposed to simulate the opposed flow diffusion flame configuration operating in the low-temperature cool flame regime. An OpenFOAM (Open-source Field Operation And Manipulation) based reacting flow solver has been utilized for this purpose. The goal is to understand the effect of multidimensional transport in the formation of cool flame in counterflow geometry and its deviation from ideal conditions. The 2D-axisymmetric model prediction has been found to be in better agreement with the experimental measurements compared to the quasi-1D model prediction. One of the critical findings of this study is identifying the Richardson number as a key parameter to characterize the system. The strain rate required to establish a cool flame is lower in comparison to that of a hot flame. At a lower strain rate, the competition between flow inertia and buoyancy dictates the location of the stagnation plane of the opposing jets and the consequent flame location. The proposed model has been able to capture this perturbation without imposing any radial velocity gradient at the nozzle exits. The extinction limit of the cool flame has been studied in 1, 3, and 5 atmospheric pressure. The extinction limits predicted by the 2D model are found to be higher than the 1D model predictions. The 2D model permits the velocity boundary conditions to be perturbed at the nozzle exits, which allows the flame to sustain higher strain rates.

In the second part of the dissertation, simulation of hydrothermal flame of methanol oxidation in supercritical water is conducted utilizing the proposed mathematical model. The high-pressure combustion is an emerging technique for improving thermodynamic efficiency and reducing harmful contaminants. Additionally, recent studies suggest the presence of low-temperature oxidation in high-pressure conditions, making it worth investigating the effects of multidimensional transport on the reacting flow system in these conditions. The simulation results provide insight into the possible multi-dimensional mixing that can lead to low-temperature flames even under supercritical conditions. In this dissertation, it is found that the hydrothermal flame exhibits different flame structures depending on the fuel loading. For a higher fuel loading (XF=0.12), a classic non-premixed flame was observed with a peak temperature of ~~2000K. Whereas, for the leanest case studied (XF=0.071), a peak temperature of ~1200K was observed with extremely low hydroxyl OH concentration (ranging to a few ppm levels). It is difficult to conclude if it is a hot flame or a reduced-temperature flame phenomenon due to partial oxidation processes. Additionally, it was observed that the flame structure is determined by the balance between convective and diffusive flux, which is dictated by the fuel loading. For high fuel loading, the flame forms on the jet periphery, while as the fuel loading decreases, the flame is observed to be lifted in the radial direction. At lower fuel loading, radial mixing is more prominent, and the multidimensional effect appears to be significant.