Choo, Edwin Jia Chiet
(2023)
Development of a new n-dodecane-n-butanol-PAH reduced mechanism and the applications under diesel spray combustion.
PhD thesis, University of Nottingham.
Abstract
Despite exhibiting low throttling losses and generating high thermal efficiency, it is inevitable that the heterogeneous combustion nature of a diesel engine will cause large amount of soot to be produced. Lately, n-butanol has emerged as a very competitive second-generation biofuel in diesel engine because of its similarities in physiochemical properties to diesel fuel. As an oxygenated hydrocarbon, blending n-butanol with diesel is able to reduce the harmful exhaust emissions. According to the literature, most studies related to diesel-n-butanol are conducted experimentally whereas only a handful of them are done numerically. However, numerical modelling is as important as conducting experiments as it allows researchers to gain further insights on the chemical kinetics. In order to support accurate combustion modelling results, the selection of a chemical kinetic mechanism that could describe the combustion chemistry precisely is crucial but it is found that the developments of diesel-n-butanol mechanisms in particular those using long carbon chain surrogates are still scarce. Since there is also a lack in effort on the combustion modelling of diesel-n-butanol, this study aims to address the limitations related to the chemical kinetic mechanisms for diesel-n-butanol and the applications in multi-dimensional simulations concerning the spray combustion.
First, a reduced chemical kinetic mechanism for n-dodecane-n-butanol-polycyclic aromatic hydrocarbon (PAH) is developed using the direct relation graph with error propagation (DRGEP) and isomer lumping methods. In the mechanism, n-dodecane is being used as the diesel surrogate fuel. Consequently, the n-dodecane-n-butanol-PAH mechanism consists of 105 species and 584 reactions (DB105) and it is validated under a wide range of engine operating conditions such as shock tube (ST) ignition delay (ID) times, jet-stirred reactor (JSR) and premixed laminar flame species concentrations and laminar flame speed. The DB105 mechanism shows good agreements to the experimental measurements and predictions by the detailed mechanism for majority of the tested conditions. Subsequently, a set of Computational Fluid Dynamics (CFD) sub-models is formulated to describe the spray, ignition, combustion and soot of n-dodecane-n-butanol in a diesel engine-like constant volume combustion chamber. The fidelity of the DB105 mechanism is further assessed in CFD simulations by comparing the predicted liquid penetration length (LPL), vapour penetration length (VPL), ID, flame lift-off length (FLOL) and soot to the experimental measurements. Nine cases at various n-butanol blending ratios and ambient temperatures are tested and results show that the predicted LPL and VPL deviated with a maximum of 4% whereas the ID and FLOL deviated by 20% and 12%, respectively. For soot validation, the soot volume fraction (SVF) of pure n-dodecane (Bu0) at 850 K and 900 K are compared to the experimental measurements where the maximum errors are recorded to be 1.0 ppm (100%) and 0.3 ppm (10%), respectively. However, the location and size of the soot cloud are also well emulated. Overall, the good validation results of the DB105 mechanism indicates that improvements have been made to the present mechanism and it is more superior than the existing n-dodecane-n-butanol mechanisms in terms of its size and accuracy.
The fundamental ignition, combustion and soot characteristics of n-dodecane-n-butanol spray flames are then numerically investigated using the formulated CFD sub-models and the DB105 mechanism. Results show that n-dodecane-n-butanol blends undergo a two-stage ignition regardless of n-butanol blending ratios and ambient temperatures. The first-stage ignition site is located at the spray periphery for all test cases except for 80% n-dodecane – 20% n-butanol (Bu20) at 800 K. However, the second-stage ignition site moves to the spray head for all test cases but it remains at the spray periphery for Bu0. At quasi-steady state, key species that pertains to ignition are shifted to the fuel-lean region of the spray at higher n-butanol blending ratio and at lower ambient temperature. Moreover, combustion mode analysis shows that the low temperature combustion (M-LTC) mode is dominant at the first-stage ignition but gradually diminishes as the high temperature combustion (M-HTC) and high temperature diffusion combustion (M-HTC-diff) modes take over from the second-stage ignition onwards. Meanwhile, simulation results also show that as the n-butanol blending ratio increases and as the ambient temperature decreases, the SVF and soot particle size in the n-dodecane-n-butanol spray flames decrease. Comparing to the case of Bu0 (900 K), both peak SVF and soot particle size of Bu40 (900 K) decreases by 93.11% and 56.58%, respectively. Similarly, Bu20 (800 K) shows a decrease of 88% and 50.62% in the peak SVF and soot particle size when comparing to Bu20 (900 K). Apart from the SVF and soot particle size, the soot precursor species also decrease at higher n-butanol blending ratio and at lower ambient temperature. Besides, the spatial distributions of soot and its relevant species shrink due to the lower local equivalence ratio. The soot formation and oxidation mechanisms are suppressed at higher n-butanol blending ratio and at lower ambient temperature. However, the soot formation mechanism is more influential to the resulting soot. Nonetheless, the high oxygen concentration in high n-butanol blends have shown to be able to compensate the deteriorated soot oxidation mechanism while suppressing the soot formation mechanism simultaneously. As a result, the SVF and soot particle size decrease more significantly with the addition of n-butanol blending ratio than the decrease in ambient temperature.
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
