|M.Sc Student||Shalumov Vitaly|
|Subject||Investigation of Jet-A and Methanol Spray Combustion Using|
Large Eddy Simulation
|Department||Department of Aerospace Engineering||Supervisor||Professor Emeritus Yeshayahou Levy|
|Full Thesis text|
As the world’s oil reserves are depleting, mankind is looking for other sources of energy. The alcohol fuel Methanol, which can be produced either from fossil or renewable resources, in particular natural gas and biomass or simply from carbon dioxide and water was considered in this study as an alternative to kerosene type jet fuel Jet-A.
The purpose of this thesis was to compare the performance of a generic combustion chamber with simplified geometry using Jet-A and Methanol fuels. One of the key elements investigated are the pollutants CO and NOx.
The solution was mainly conducted using numerical simulations in OpenFOAM with computational mesh produced in Salome - both open-source codes. The simulation consists of non-premixed multiphase combustion and uses the Lagrangian-Eulerian approach to solve the multi-phase flow.
LES (Large Eddy Simulation) level of modeling was used to solve the filtered Navier-Stokes equations. Large Eddy Simulation, unlike RANS, resolves the large scales of the flow (Grid Scales) which contain most of the energy and models the smaller (Sub Grid Scales) scales.
One-equation eddy modeling was used for the Sub Grid Scale (SGS) closure.
The chemistry kinetics applies a global reaction scheme with 4 reactions both for Jet-A and Methanol fuels.
The experimental burner setup simulates an industrial burner with heating power of 50 [kW]. The setup includes an electrical compressor for air flow control and a swirler with a fuel atomizer at the swirler center.
Jet-A simulation results fit the experimental results with exception of NOx percentage in the burned gas. The amount of NOx obtained by the simulation is smaller by several hundred percent than the experimental results, a difference attributable to incompatible combustion rate mechanism of the global reaction (too fast in fuel-rich regions and too slow in fuel lean regions) which produces smaller combustion zone than the experiment.
Methanol simulation results showed a good fit to experiment for both the temperature and burned gas composition.
Both simulation and experimental results showed that the amount of CO produced in Methanol combustion is higher by several hundred percent in comparison to Jet-A case. The experimental results showed that the percentage of NOx in Jet-A combustion is much higher than in Methanol combustion experiment (analytical simplified case suggests that the result is attributed to a slightly higher Jet-A adiabatic temperature with addition to small but highly energetic combustion area near the nozzle).