The tasks and achievements accomplished during the fellowship in each WP are summarized as follows.
WP1: Development of the kinetic mechanism.
The researcher developed a kinetic mechanism for plasma-assisted ammonia combustion by taking the following tasks. Task 1: Ammonia mechanism improvement and validation. In this initial task, the researcher conducted a thorough evaluation of previously developed chemical mechanisms for ammonia combustion. The goal was to identify the most accurate mechanism capable of predicting the key characteristics of ammonia combustion including ignition delay time, flame speed, flame thickness, and pollutant emissions. Task 2: Combination of the thermal ammonia model with model for plasma process. Following the validation of the selected combustion mechanism against existing experimental data in the literature, the researchers developed a plasma mechanism for NH3/H2/N2/O2/He mixtures. Subsequently, this newly created plasma mechanism was combined with a combustion mechanism for the first time. Task 3: Mechanism reduction. In the final task, the researcher developed a reduced mechanism suitable for implementation in CFD codes. This was accomplished through the implementation of the directed relation graph with error propagation model, ensuring a more computationally efficient and accurate mechanism. The researcher was unable to undergo the secondment to Lund University due to travel restrictions that were in place during the COVID19 outbreak. In response, the researcher proactively acquired the necessary knowledge from the available resources to self-train for this WP.
WP2: Fundamental characteristics of premixed plasma-assisted ammonia flames.
In this WP, first, the researcher developed a numerical code combining Cantera, a high-order open-source chemical software with ZDPlaskin, an open-source software to model plasma chemistry. This code was then used to study plasma-assisted ammonia combustion through two tasks. Task 1: Code evaluation. The developed code was rigorously validated against previously obtained experimental data for plasma-assisted methane-air flames as a benchmark. Task 2: Generating a comprehensive database. In the second task, the researcher created an extensive database encompassed a wide array of plasma-assisted ammonia flame characteristics under various operating conditions and plasma settings. These characteristics included flame speed, flame thickness, extinction strain rate, flammability limits, and pollutant emissions.
WP3: Novel swirl burner experiments.
The researcher was unable to proceed with this WP and the associated secondment due to an unforeseen delay in the delivery of the required plasma generator to Cardiff University.
WP4: Large-eddy simulations (LES) of plasma/ammonia flames in the swirl burner
Relying on LES to investigate complex fluid flows like plasma-assisted ammonia combustion requires extensive validation of LES results against experimental data. Unfortunately, the cancellation of WP3 and the lack of experimental data compelled the researcher to adopt his approach within this WP. As a result, instead of using LES, direct numerical simulations (DNS) were used to study flame-plasma-turbulence interactions. DNS offers the most precise predictions of fluid flows where there is no need for specific validations, as it captures all turbulent scales within the simulations. However, DNS is a computationally intensive approach. Considering such limitations, the researcher shifted his focus from a swirl burner to investigating freely propagating turbulent flames in ammonia/air mixtures stimulated by plasma discharges. These adjustments, while demanding and resource-intensive, have yielded valuable insights into fundamental aspects of plasma-assisted ammonia combustion that were previously unexplored.
The main findings of the project in WP1 and 2 have been published in the top flag journals of the combustion field, namely “Combustion and Flame” and “Fuel”. Additionally, the researcher is currently preparing two papers to present the key research findings from WP4.
