LIST OF FIGURES Figure 5-4 Temporal developments of mean pressure and mean HRR for BL. During the first stage of ignition, the temporal evolutions of the mean HRR for cases 6 and 7 are identical to case 1.
Principles and fundamentals of HCCI combustion
Therefore, prototypes of HCCI internal combustion engines have been developed as an alternative to conventional gasoline SI and diesel CI engines. Among the derivatives of HCCI combustion concepts, thermal stratification (TS), spark assisted compression ignition (SACI), stratified charge compression ignition (SCCI) and reactivity controlled compression ignition (RCCI) HCCI engines have received considerable attention. by the motoring community.
HCCI with thermal stratification
Stratified charge compression ignition (SCCI)
1-6, flexible adjustment of the late direct injection (DI) timing and/or the fraction of PFI/DI generates desired levels of mixture stratifications or appropriate equivalence ratio distributions. For φ-sensitive fuels, the use of DI fraction and direct injection timing can be adopted as control parameters for adjusting the degrees of fuel stratification and for adjusting the HRR under certain operating conditions.
Reactivity controlled compression ignition (RCCI)
As the load increases, the overall reactivity of the charge decreases, leading to too advanced combustion stages. At this stage, there is no way to effectively control the combustion phasing and process of the RCCI combustion.
Direct dual fuel stratification (DDFS)
Scope and objective
First, Chapter 2 is devoted to the discussion of the chemical reaction flow formulation and the numerical methods used in the present study. Next, the ignition characteristics of lean primary reference fuel/air mixtures with temperature inhomogeneities are investigated in Chapter 3.
Turbulent flow fields and scalar fields
The S3D DNS Code
Bansal and Im [52] investigated the effects of compositional inhomogeneities along with temperature fluctuations on HCCI combustion of the same lean hydrogen/air mixture. 53] investigated the ignition characteristics of a lean heptane/air mixture with different averages and root-mean-squares (RMSs) of temperature and the effect of the negative-temperature coefficient (NTC) regime on overall HCCI combustion.
PRF reduced mechanism
Skeletal and reduced mechanisms were validated for various PRF compositions with an octane range of 0 to 100 through autoignition delay and extinction residence time over a wide parameter range of pressures from 1 to 50 atm, equivalence ratios from 0.3 to 0.7, and onset temperatures from 750 to 1800 K for self-ignition and inlet temperature of 300 K for PSR. It is easy to observe that both skeleton and reduced mechanisms accurately reproduce the results of the detailed mechanism.
Initial conditions
Details of the physical and numerical parameters for the fifteen cases are presented in Table 3-1. Typical profiles of the initial temperature and vorticity in the DNS cases are shown in Figs.
Effects of fuel composition and initial temperature RMS
Overall characteristics of PRF HCCI combustion
For the purpose of comparison, the temporal evolutions of 0-D homogeneous ignition of the corresponding PNS/air mixtures are also shown in the figure. These results are qualitatively similar to those of the ignition of a lean hydrogen/air mixture [50, 51] and of lean hydrocarbon/air mixtures with high initial mean temperatures [53, 54].
Chemical explosive mode analysis
It is also important to note that for the cases with the same T0, the effect of the fuel composition disappears with increasing T0. It can also be observed that Dac upstream of the reaction fronts is much greater than unity (yellow area), indicating that the chemical reaction (λe) is also much faster than the mixing process (χ).
Front speed and burning rate
In cases with a large T0, however, the deflagration mode of combustion occurs at the reaction fronts, while the spontaneous combustion mode also occurs in front of the reaction fronts. First, the fraction ˙q from the deflagration mode increases with increasing T0 regardless of fuel composition, as expected from the above explanation.
Effect of turbulence
3-6 that the ignition delay of PRF50 is slightly longer than that of the other PRFs when the initial temperature is higher than 1080 K. As a result, the overall combustion of PRF50 is delayed slightly more by high u0 than that of the other two PRFs.
Conclusions
Therefore, these expansion and quench characteristics result in retardation of the overall combustion of the PRF50/air mixture with a large u0. The remainder of the fuel (up to 20% of the total fuel volume) is then directly injected during the late compression stroke or near top dead center (TDC) to introduce a certain amount of variation in the equivalent ratio (φ) [17].
Initial conditions
Extensive validation of the reduced mechanism relevant to the present DNAs is shown in Fig. Furthermore, the initial conditions are representative of the TDC under high-load conditions in air-diluted HCCI combustion [90, 116].
Effect of T 0 and φ 0
Overall combustion characteristics
- Baseline cases
- Cases with uncorrelated T − φ
- Cases with negatively-correlated T − φ
- Instantaneous HRR fields
As found in a previous study [119], the overall combustion of HCCI is generally accelerated by the compression heating of the piston motion in practical IC engines. Second, a large φ0 (case 10) can overcome the mutual canceling effect of the negative correlation of the T−φ fields on the overall combustion.
Chemical explosive mode analysis
These results imply that the reaction fronts of Case 8 are deflagrations rather than spontaneous ignition fronts. In particular, Dac upstream of the reaction fronts has a magnitude much larger than unity (yellow area), indicating that the CEM (λe) is also much faster than the mixing process (χ).
Combustion mode
Third, for cases 9, 11 and 12 with negatively correlated T−φ fields, the fraction of HRR from the deflagration mode disappears much sooner than the appearance of the peak ˙q, and as such the total heat release from the deflagration mode is about 3 , 6 and 9% respectively; in contrast, for large T0 orφ0 (i.e., Cases 4 and 6–8), a significant fraction of HRR still occurs through deflagration mode at each peak ˙q, resulting in a relatively large total heat release from the deflagration mode (> . 30% ). A small fraction of HRR from the deflagration mode verifies that spontaneous autoignition occurs mainly for cases with small T0 and/or φ0 or cases with negatively correlated.
Effect of turbulence
Finally, the temporal evolutions of the fraction of HRR from the deflagration mode are investigated to quantify the combustion mode as shown in Fig. This result indicates that large u0 can effectively homogenize the mixture so that the overall combustion occurs mainly through spontaneous ignition, which leads to the decrease of the fraction of HRR through the deflagration mode.
Discussion
In general, uncorrelated T−φ fields induce much shorter ignition delays in mixtures regardless of fuel types, and as such the average HRR is more distributed and the overall combustion is more advanced with increasing T0 and φ. However, for two-stage ignition fuels, it is interesting to note that large φ0 with small T0 can promote overall combustion and benefits.
Conclusions
These results suggest that stratification of a fuel/air mixture in both temperature and composition can control the ignition timing of HCCI combustion and ensure smooth operation of HCCI under high load conditions. However, the ignition of a lean biodiesel/air mixture under HCCI conditions exhibits relatively weak NTC behavior and as such, the effects of the strong NTC behavior of the fuel/air mixture on HCCI combustion were not fully evaluated.
Initial conditions
It is interesting to note that the time evolution of 3-D turbulence is qualitatively different from 2-D turbulence due to the effect of 3-D eddy stretching. However, it is still important to investigate HCCI combustion with 2-D turbulence because the effect of turbulent mixing on HCCI combustion is less important than the effect of mixture stratification.
Effects of T 0 and φ 0 at different T 0
- Effects of T 0 or φ 0 at different T 0 : BL cases
- First-stage ignition: BL cases
- Second-stage ignition: BL cases
- Effects of negatively-correlated T − φ fields: NC cases
- Effects of uncorrelated T − φ fields: UC cases
- Combustion mode
In addition, the overall combustion ends much later than that of the corresponding BL case. However, in reality, the overall combustion of both cases proceeds quite similarly to each other.
Chemical aspects of the ignition
Characteristics of temporal species evolution
Similar to the cases with T0 = 833 K, the time distribution of the average HRR during the second stage ignition shows a progressive increase of OH and decrease of H2O2. Third, for cases with T0 = 1025 K (see Figure 5-14), n-heptane is first gradually consumed, and then the consumption rate increases until the end of complete combustion.
Chemical explosive mode analysis
However, in the unburned region upstream of the deflagrations (see Fig. 5-16e and f), it is easily observed that the chain branching of H2O2 (R48) and the generation of alkyl radical and H2O2 (R292 ~ R300) are the most important reactions to the CEM. Figures 5-17 and 5-18 show the isocontours of EI and PI values of the most important species and reactions.
Discussion
This expansion effect on HCCI combustion can be enhanced more by negatively correlated T−φ fields in the low-temperature regime. On the contrary, a large T0 and/or φ0 can still accelerate the overall combustion of HCCI with a higher PRR peak and shorter burn duration, even if combustion occurs after TDC.
Conclusions
For deflagrations, temperature, CO, and OH are found to be the most important variables, while for CEM, the CO to CO2 conversion reaction and the high-temperature branching reaction of the H + O2 → O + OH chain are important. However, the dominant role of reactivity stratification may not remain the same when T0 lies in the low- and high-temperature regimes, so the effect of T0 relative to the NTC regime on RCCI combustion is still needed.
Initial conditions
The same characteristic length scale as le is used for the concentration and temperature fields. The domain is discretized with 1280 grid points in each direction, and the corresponding grid resolution is 2.5 µm.
Results and discussion
Therefore, the overall RCCI/SCCI combustion in the high temperature regime is more likely to occur similarly to the corresponding 0-D ignition. Furthermore, in our previous DNS study [7], it was found that in the high temperature regime, only T0 is more effective than negatively correlated T0 and φ0 in reducing peak HRR and advancing overall combustion.
Combustion mode analysis
The deflagration mode of combustion can be quantitatively measured by analyzing the Damk¨ohler number in which Dais is defined as. Note that, in general, the deflagration mode of combustion is more suitable for propagation.
Effect of turbulence
Note also that the chemical aspects of the ignition of the present PRF/air mixture under RCCI/SCCI mode are further elucidated in [145].
Conclusions
Therefore, the objective of this study is to provide insight into the chemical ignition process of PRF/air mixture under RCCI/SCCI conditions by analyzing 2-D DNS data with CEMA. To this end, we use the dataset generated from 2-D DNSs for the ignition of a lean PRF/air mixture under RCCI/SCCI conditions by Luong et al.
Methodology
However, both RCCI and SCCI have similar overall combustion characteristics in the high temperature regime because the ignition of the PRF/air mixture becomes less sensitive to reactivity stratification and/or equivalence ratio. Since the Jacobian of the chemical source term, Jω (≡ ∂ω/∂y), preserves the chemical information of the local mixture, the chemical property of the mixture can be determined based on the Jacobian.
Results and discussion
Overall RCCI/SCCI combustion
The results revealed that during the first ignition phase and the start of the main combustion of the RCCI case, a significant amount of heat is released from mixtures with large Z, which corresponds to local mixtures with high reactivity (high heptane concentration) and φ. For the SCCI case, the start of the main combustion also originates from mixtures with large Z.
CEMA: 0-D ignition
Second, temperature, n-heptane, and CH2O near first-stage ignition (∼t/τig0 = 0.27) are identified as the most important variables for CEM, characterized by the peak PIs for. As such, temperature, CO and OH are identified as the key variables for CEM in second stage ignition.
CEMA: 2-D DNS of RCCI/SCCI combustion
In contrast to the early stage of combustion, most heat is released from high-temperature region (T >1500 K) where HTC represented by CO oxidation (R24) and hydrogen chemistry (R1, R9 and R13) is dominant for both RCCI and SCCI combustion. As such, temperature, CO and OH are identified as the most important variables for the CEM and HTHR determines temperature increase at this stage.
Conclusions
DDFS combustion combines the high thermal efficiency of RCCI combustion with the high load capability of Partial Premixed Combustion (PPC) while reducing Exhaust Gas Recirculation (EGR) rates, noise and combustion instabilities. The effect of compression heating and expansion cooling due to piston motion is also taken into account by adopting a new compression heating model.
Model development
Compression heating model
The compression heating model adopted in the present study has an improved feature compared to that used in the previous studies. The previous models used a mass source term that is uniform over all grid points, regardless of the presence of temperature and equivalence ratio inhomogeneities.
Pseudo-iso-octane model
The effective injection duration or the conversion time from PC8H18 toi-C8H18,tdur, of 0.8 ms, is chosen as in [4] for all DDFS cases in the current study.
Initial conditions
The temperature of EGR is assumed to be the same as that of the reactants. Six DDFS cases (cases 3-8 in Table 8-3) are simulated by varying the injection timing.
Results and discussion
Overall combustion characteristics
Second, the overall DDFS combustion was delayed compared to the RCCI combustion regardless of injection timing, tinj. Furthermore, the DDFS combustion can precisely control the occurrence of the main combustion, especially when the direct injection of fuel starts during the HTC regime.
Combustion mode analysis
Therefore, the predominance of the spontaneous ignition mode can be expected for all DDFS cases. These results suggest that the spread of the HRR in the DDFS combustion is not attributed to the deflagration but to the successive injection of i-C8H18.
Misfire
In summary, budget analysis and Damk¨ohler number analysis verify that although there is a mixed combustion mode of both ignition and spontaneous ignition during the early stage of DDFS combustion (Cases 5 and 7), spontaneous ignition is the dominant mode of combustion. combustion. for both DDFS and RCCI combustion.
Conclusions
It was also found that the ignition delay of the DDFS combustion has a non-monotonic behavior with increasing fuel injection time. Furthermore, the rate and duration of the primary DDFS combustion can be effectively controlled when the direct injection of i-C8H18 occurs under the HTC regime.
Future research directions from DNS towards LES
Chen, Direct numerical simulation of flame front propagation in a constant volume with temperature inhomogeneities: II. Yoo, Direct numerical simulations of the ignition of reference lean fuel/air mixtures with temperature inhomogeneities, Combust.
