Repository logo
 

A computational study of auto ignition, spark ignition and dual fuel droplet ignition in a rapid compression machine

Date

2017

Authors

Bhoite, Siddhesh, author
Marchese, Anthony J., advisor
Olsen, Daniel B., committee member
Mahmoud, Hussam N., committee member

Journal Title

Journal ISSN

Volume Title

Abstract

A series of computational modeling studies were performed using the CONVERGETM computational fluid dynamics (CFD) platform to gain in-depth understanding of the chemically reacting flow field, ignition and combustion phenomena in a various rapid compression machine (RCM) experiments conducted at CSU including homogeneous autoignition, laser ignition and droplet ignition experiments. A three-dimensional, transient computational modeling study was initially performed to examine premixed, homogeneous autoignition of isooctane/air and methane/air mixtures. A reduced chemical kinetic mechanism for isooctane comprising of 159 species and 805 reactions was developed using direct relation graph error propagation and sensitivity analysis (DRGEPSA) method. Computational results showed good agreement with experimental results capturing the negative temperature coefficient (NTC) behavior of isooctane. The premixed computations also revealed the importance of the piston crevice design for maintaining a homogenous flow field inside a RCM. The result showed that, as the volume of the piston crevices is increased, the roll up vortices are eliminated, which reduces the mixing of the lower temperature boundary layer gases with the higher temperature core gases, thereby maintaining the homogeneity of the flow field. Next, three-dimensional computational modeling laser-ignited premixed fuel/air mixtures at elevated temperatures and pressures in the RCM was performed with detailed chemical kinetics (86 species, 393 reactions). For methane/air mixtures, the computational results were compared against previously reported RCM experiments. Computations were also performed on laser-ignited n-heptane/isooctane/air mixtures under similar simulated conditions in the RCM. In the computations, a simulated spark modeled as a localized hotspot was introduced in the center of the combustion chamber resulting in an outwardly propagating flame, which, depending on the fuel reactivity, produced ignition in the end gas upstream of the flame. Methane/air computations were performed at equivalence ratio of 0.4 ≤ Ф ≤ 1.0 for direct comparisons with experimental measurements of instantaneous pressure, flame propagation rate, and lean limit. For compressed temperature of 782 K, a methane/air lean limit of Ф = 0.38 was predicted computationally (combustion efficiency, χ = 0.8), which was in good agreement with the experimental measurement of Ф = 0.43. For n-heptane/isooctane/air computations, auto-ignition of the end gas was predicted depending on the compressed temperature and Octane Number, which suggests the use of the laser ignition/RCM system as a means to quantify fuel reactivity for spark ignited engines. Lastly, RCM experiments in which single n-heptane droplets are suspended and ignited via compression-ignition in a quiescent, high-pressure, high-temperature, lean methane/air environments were simulated using the 86-species dual-fuel chemical kinetic mechanism developed previously. The simulations capture the ignition event in the vicinity of a spherical n-heptane droplet, which bifurcates into a propagating, premixed methane/air flame and stationary n-heptane/air diffusion flame. Comparisons against experimental measurements of droplet gasification rate, premixed flame propagation speed, and non-premixed flame position will be used to develop revised dual-fuel chemical kinetic mechanisms.

Description

Rights Access

Subject

Citation

Associated Publications