A detailed chemistry solver on adaptive curvilinear meshes and its application to rotating detonation simulations
A detailed chemistry solver on adaptive curvilinear meshes and its application to rotating detonation simulations
The rotating detonation engine (RDE) has drawn increasing interest in recent years due to its potential for high thermal efficiency and pressure-gain properties. The simulations of RDE have primarily focused on premixed injection and its wave structure, as the premixing assumption allows for the use of simplified chemistry models. However, a fully premixed RDE may lead to potential flashback issues during actual experiments. Another common simplification present in the simulation of RDEs is the adiabatic wall boundary condition. As the duration of operation increases to hundreds of seconds, the need of a cooling system has become more urgent. Simulations addressing these challenges require a three-dimensional solver capable of accurately and efficiently handling detailed chemistry and boundary flows.
In this work, a three-dimensional solver is developed based on the Adaptive Mesh Refinement in Object-oriented C++ (AMROC) framework. The adaptive mesh refinement technique enables dynamic mesh refinement based on a user-defined threshold. Initially proposed for Cartesian meshes, the adaptive mesh technique is extended and implemented on curvilinear meshes in this study. This novel combination of a stretched body-fitted mesh and the dynamic adaptive mesh reduces the total number of meshes required for an RDE simulation. The developed solver is verified and validated with different benchmark tests. The results show that the present solver achieves second-order accuracy and ensures conservation across the multi-level hierarchy. In addition, the solver demonstrates the capability for robust and accurate simulation of high-speed reacting flows, including unsteady shock-induced combustion and curved cellular detonation.
Finally, the solver is applied to investigate the effects of partial premixing on RDE performance. Hydrogen fuel is blended into the air stream at different bypass flow rates. An increase in bypass flow rate results in improved RDE performance, as indicated by higher detonation velocity, thrust, and specific impulse. The effects of cooling walls are also studied, and the results confirm that introducing a cooling wall during the operation process still preserves the number of detonation heads and the macro structure on the middle plane. The adiabatic case overestimates the detonation velocity without considering the heat loss on detonation. The cooling walls play a crucial role in suppressing parasitic and commensal combustion waves near the walls, leading to a reduction in low-pressure heat release. Moreover, the walls experienced unequal heat loads, primarily influenced by the internal combustion zones and the channel width. These findings enhance the understanding of partial premixing and cooling walls on RDE from the numerical aspect.
University of Southampton
Peng, Han
62906b46-9628-43fc-921d-b6257b1fec6f
17 January 2024
Peng, Han
62906b46-9628-43fc-921d-b6257b1fec6f
Deiterding, Ralf
ce02244b-6651-47e3-8325-2c0a0c9c6314
Peng, Han
(2024)
A detailed chemistry solver on adaptive curvilinear meshes and its application to rotating detonation simulations.
University of Southampton, Doctoral Thesis, 232pp.
Record type:
Thesis
(Doctoral)
Abstract
The rotating detonation engine (RDE) has drawn increasing interest in recent years due to its potential for high thermal efficiency and pressure-gain properties. The simulations of RDE have primarily focused on premixed injection and its wave structure, as the premixing assumption allows for the use of simplified chemistry models. However, a fully premixed RDE may lead to potential flashback issues during actual experiments. Another common simplification present in the simulation of RDEs is the adiabatic wall boundary condition. As the duration of operation increases to hundreds of seconds, the need of a cooling system has become more urgent. Simulations addressing these challenges require a three-dimensional solver capable of accurately and efficiently handling detailed chemistry and boundary flows.
In this work, a three-dimensional solver is developed based on the Adaptive Mesh Refinement in Object-oriented C++ (AMROC) framework. The adaptive mesh refinement technique enables dynamic mesh refinement based on a user-defined threshold. Initially proposed for Cartesian meshes, the adaptive mesh technique is extended and implemented on curvilinear meshes in this study. This novel combination of a stretched body-fitted mesh and the dynamic adaptive mesh reduces the total number of meshes required for an RDE simulation. The developed solver is verified and validated with different benchmark tests. The results show that the present solver achieves second-order accuracy and ensures conservation across the multi-level hierarchy. In addition, the solver demonstrates the capability for robust and accurate simulation of high-speed reacting flows, including unsteady shock-induced combustion and curved cellular detonation.
Finally, the solver is applied to investigate the effects of partial premixing on RDE performance. Hydrogen fuel is blended into the air stream at different bypass flow rates. An increase in bypass flow rate results in improved RDE performance, as indicated by higher detonation velocity, thrust, and specific impulse. The effects of cooling walls are also studied, and the results confirm that introducing a cooling wall during the operation process still preserves the number of detonation heads and the macro structure on the middle plane. The adiabatic case overestimates the detonation velocity without considering the heat loss on detonation. The cooling walls play a crucial role in suppressing parasitic and commensal combustion waves near the walls, leading to a reduction in low-pressure heat release. Moreover, the walls experienced unequal heat loads, primarily influenced by the internal combustion zones and the channel width. These findings enhance the understanding of partial premixing and cooling walls on RDE from the numerical aspect.
Text
Han_Peng_PhD_Thesis_Final
- Version of Record
Text
Final-thesis-submission-Examination-Mr-Han-Peng
Restricted to Repository staff only
More information
Published date: 17 January 2024
Identifiers
Local EPrints ID: 486406
URI: http://eprints.soton.ac.uk/id/eprint/486406
PURE UUID: 31623ac4-f613-46b3-b535-85945127aeb4
Catalogue record
Date deposited: 19 Jan 2024 18:24
Last modified: 20 Mar 2024 02:58
Export record
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics