A DIRECT NUMERICAL SIMULATION STUDY ON THE PHYSICOCHEMICAL ASPECTS OF TURBULENT JET IGNITION

Open access
Author
Date
2020Type
- Doctoral Thesis
ETH Bibliography
yes
Altmetrics
Abstract
The next generation of internal combustion engines requires innovative technologies aiming at high efficiency and low emissions. Natural gas engines operating under lean conditions can lead to low fuel consumption and low emissions. However, the low ignitability of natural gas lean mixtures requires new ignition system technologies. Apart from conventional spark ignition or other ignition systems (e.g. plasma igniters, laser-induced ignition), the turbulent jet ignition (TJI) system has come to the fore as a promising system to sustain stable and successful combustion with the potential to decrease unburned hydrocarbons and CO emissions.
In such a system an auxiliary chamber (prechamber) is located at the top or side of the main engine combustion chamber, where a richer or same mixture composition with the main chamber is ignited by a spark plug. The hot gases produced by combustion in the prechamber (PC) exits into the main chamber (MC) through one or more orifices as turbulent reacting jet(s), providing favorable local mixing and rapid combustion throughout the main chamber. In such a system the flow and chemical underlying processes and interactions are tightly coupled and are affected by geometric design parameters and operating conditions.
Multiple numerical and experimental studies have been conducted over the last decades to understand the different physical aspects of TJI process. Numerical experiments, such as Direct Numerical Simulations (DNS) can provide the whole information of the thermochemical and flow system free from modelling assumptions, although in a very narrow range of initial thermochemical conditions due to their high computational cost.
The objective of this study is to investigate and characterize the TJI pro- cess under different initial thermochemical and flow conditions, geometric and other parameters (e.g. spark location) aiming to make advances in understanding for the design of an efficient process. The combustion phe- nomenology, flow and reacting regions interactions, combustion mode and the underlying chemical and transport processes are investigated. From modelling perspective, a DNS investigation can provide useful data and insights regarding the physicochemical aspects of the process.
A series of DNS simulations were conducted. Firstly, two-dimensional simulations were conducted in order to apply a parametric study under reasonable computational cost. The effects of initial temperature, main chamber composition as well as the effects of initial flow field (quiescent or turbulent), geometric factors (orifice corners’ shape and diameter), prechamber spark location and wall boundary conditions on the pro- cess were investigated. In addition, two-dimensional DNS simulations of smaller size were further carried out in order to investigate the effect of prechamber composition on ignition and the influence of orifice on react- ing flow.
The two-dimensional simulations show that the TJI process can be divided in three phases. Flame front propagation takes place in the prechamber during the first phase while unburned PC mixture is pushed into the MC generating a non-reactive jet. During the second phase the PC burned gases issue into the MC flow field. During the third phase an inverse flow from MC to PC takes place due to increase of MC pressure. The flame surface growth rate in the MC depends on the vortices size of the non-reactive jet and on the vortical structures created at the lower chamber wall due to jet impingement. The initial temperature affects the dissipation rate of the vortices and therefore the surface growth and combustion phenomenology. Small orifice diameter leads to small vortices (small contribution on surface growth) but at the same time it leads to long jet penetration which enhances the flame surface growth due to the effect of wall vortical structures and strong entrainment of the unburned mixture into the jet. The prechamber spark location and the initial flow field affect the jet penetration and the temperature reduction of the PC burned gases. The nozzle can affect the reacting flow especially when its size is on the order of the flame thickness and can also lead to significant temperature reduction. PC rich mixtures were found to lead to higher consumption rate in the MC in com- parison to stoichiometric PC mixtures. The equilibrium state of PC burned gases and the radicals contained are crucial for the ignition. The thermo- chemical evolution of reaction zones and their interaction were further statistically investigated highlighting phenomena which can be significant for the modelling of the process.
The study proceeds with three-dimensional DNS simulations. The effect of main chamber composition is investigated by two simulations in a simplified domain in which the prechamber is excluded and replaced by time-varying boundary conditions. Similar analysis with the 2-D DNS leads to similar observations regarding the thermochemical evolution of flame surface. The third DNS considers a laboratory-scale setup consisting of both chambers in order to compare with the findings of the previous 2-D and 3-D simulations. Show more
Permanent link
https://doi.org/10.3929/ethz-b-000457251Publication status
publishedExternal links
Search print copy at ETH Library
Contributors
Examiner: Boulouchos, Konstantinos
Examiner: Mastorakos, Epaminondas
Examiner: Frouzakis, Christos
Publisher
ETH ZurichSubject
turbulent jet ignition; Prechamber ignition; Natural gas combustion; Direct numerical simulationOrganisational unit
03611 - Boulouchos, Konstantinos (emeritus) / Boulouchos, Konstantinos (emeritus)
More
Show all metadata
ETH Bibliography
yes
Altmetrics