Title: Effect of dilution strategies and direct injection ratios on Stratified Flame Ignition (SFI) hybrid combustion in a PFI/DI gasoline engine Author names and affiliations: Xinyan Wang a , Hua Zhao a, b , Hui Xie a, * a State Key Laboratory of Engines, Tianjin University, Weijin Road 92, Nankai District, Tianjin 300072, PR China. b Centre for Advanced Powertrain and Fuels, Brunel University London, Uxbridge UB8 3PH, United Kingdom. * Corresponding author: Tel/Fax: +86 22 27406842 8009, Email: [email protected]. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
81
Embed
bura.brunel.ac.uk · Web viewspark ignition, flame propagation and auto-ignition in SFI hybrid combustion, a set of models were adopted and validated as follows. 2.1.1 Spray modeling
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Title:
Effect of dilution strategies and direct injection ratios on Stratified Flame Ignition (SFI)
hybrid combustion in a PFI/DI gasoline engine
Author names and affiliations:
Xinyan Wang a, Hua Zhao a, b, Hui Xie a,*
a State Key Laboratory of Engines, Tianjin University, Weijin Road 92, Nankai District,
Tianjin 300072, PR China.
b Centre for Advanced Powertrain and Fuels, Brunel University London, Uxbridge UB8 3PH,
dominated combustion duration D2 and the ratio of the accumulated heat released (RCAT) at
CAT. CAT is defined as the Crank Angle corresponding to the mode Transition from SI to
CAI. D1 is the duration between CA10 and CAT, and D2 is the duration between CAT and
CA90. In Case 3, the slightly fuel rich mixture (ϕair =1~1.1) gradually moves to the central
region, as shown in Fig. 5, and the corresponding auto-ignition delay can be significantly
shortened with the heating effect by the flame front. As a consequence, the mode transition
from SI to CAI occurs quickly after the flame propagation. As shown in Fig. 8, the flame
propagation dominated combustion duration D1 is shortest and RCAT is also the lowest.
Therefore, the dilution strategy adopted in Case 3 can enhance both the early flame
propagation and the later auto-ignition because of the appropriate fuel/air equivalence ratio
distribution. In Case 4, the flame propagation process occurs mostly in the leaner stratified
charge region. In addition, the reduced hot internal residual gas in Case 4 further lowers the
thermal condition, inhibiting the occurrence and development of auto-ignition. As a result,
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
the auto-ignition shows great reliance on the early flame propagation, leading to increased D1
and RCAT. The over-lean mixture of the mixture in outer region and poor thermal condition in
Case 4 result in the longest combustion duration of the auto-ignition stage (D2).
Fig. 8. The flame propagation dominated combustion duration (D1), the auto-ignition dominated combustion duration (D2) and the ratio of the accumulated heat released (RCAT) at
CAT.
Fig. 9 shows the average auto-ignition tendency of the mixture in the whole combustion
chamber and its variation among different zones. It can be seen that the profiles of the
average auto-ignition tendency in the whole combustion chamber almost overlap in Case 2
and Case 3 at the early stage of the combustion process and gradually deviate from each other
after 5 ºCA aTDC. It can be inferred that the in-cylinder dilution conditions in Case 3 are
quite beneficial for promoting the auto-ignition because of the less heat release from flame
propagation in Case 3, as shown in Fig. 8. With the development of the combustion process,
the leaner mixture in the outer region shows longer auto-ignition delay and reduces the
average auto-ignition tendency to a slight extent in Case 3. This can be verified by the
dramatically increased difference of the average auto-ignition tendency in Zone 7 between
387
388
389
390
391
392393394
395
396
397
398
399
400
401
402
403
404
405
Case 2 and Case 3 with the development of combustion process. Although the early flame
propagation in Case 4 is enhanced because of the higher fuel/air equivalence ratio distribution
in the central region, its positive impact on the later auto-ignition is not sufficient to
compensate for the negative impact on auto-ignition brought by the over-lean mixture and
lower thermal conditions, leading to lowest auto-ignition tendency traces in Fig. 9.
Therefore, both in-cylinder thermal and component conditions show essential impact on SFI
hybrid combustion. In order to optimize SFI hybrid combustion, the adopted dilution
strategies should not only improve the early flame propagation, but also benefit the later auto-
ignition process because auto-ignition is more sensitive to the dilution and thermal
conditions.
Fig.9. The average auto-ignition tendency of the mixture in the whole combustion chamber and its variation among zones for the SFI combustion with different dilution strategies.
3.2 Effect of dilution strategies on SFI hybrid combustion with a CR of 14:1
As discussed in Section 3.1, the proposed hybrid SFI combustion concept could effectively
control the PPRmax with appropriate dilution strategies, which indicates the potential to
accommodate a higher compression ratio to further improve thermal efficiency. In this
section, the effect of dilution strategies on the SFI hybrid combustion with a higher
compression ratio of 14:1 is investigated.
406
407
408
409
410
411
412
413
414
415
416
417418
419
420
421
422
423
424
Fig. 10 shows the effect of dilution strategies on CA50, IMEP and PRRmax of the SFI hybrid
combustion at the higher compression ratio. Case 5 with the same dilution strategy as Case 2
produces greater output and hence leads to higher efficiency than Case 2 because of the
increased compression ratio. However, the PRRmax in Case 5 increases more significantly,
which exceeds 5 bar/ ºCA at all ignition timings. The air dilution strategy adopted in Case 6
enhances the early flame propagation process and elevates both IMEP and PRRmax. The
replacement of the in-cylinder residual gas with the fresh intake air in Case 7 reduces the
PRRmax at the expense of lower IMEP. Therefore, the trade-off between IMEP and PRRmax
still exists with the above dilution strategies in a high compression ratio engine.
In Case 8, a new dilution strategy with increased intake fresh air at a constant concentration
of internal residual gas was studied. The ϕair in Case 8 was kept the same as Case 7, and
correspondingly the ϕdilution was decreased to 0.55 because of the increased total dilution mass.
In this case, the central flame propagation would be enhanced as that in Case 7, while the
subsequent auto-ignition would not be dramatically inhibited because of the maintained
thermal conditions brought by sufficient internal residual gas. In general, the combustion
phasing in Case 8 is delayed compared to that in Case 6 and comparable to that in Case 5.
Compared to Case 7, the combustion phasing in Case 8 is more advanced at the retarded
spark timings because of the improved thermal conditions that can guarantee the stable auto-
ignition even with late spark ignition. As a result, the IMEP values in Case 8 are slightly
lower than those in Case 5 and 6 and relatively higher than Case 7. In the meantime, the
PRRmax values in Case 8 are reduced below 5 bar/ºCA at all spark timings. However, it is
noted that further dilution with the intake fresh air in Case 9 would significantly inhibit the
combustion process and reduce IMEP dramatically as shown in Fig. 10.
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
Fig. 10. Effect of ϕair and ϕdilution on CA50, IMEP and PRRmax of SFI hybrid combustion with a
CR of 14:1.
Fig. 11 (a) compares the in-cylinder pressure traces with different ϕair and ϕdilution with CA50
around 2.6 ºCA bTDC. Under such condition, Case 5 produces the maximum IMEP. In Case
6, the spark timing is delayed in order to maintain the combustion phasing, leading to
relatively weak flame propagation. However, the higher compression ratio increases the
charge pressure and temperature, leading to dramatically increased PRRmax during the auto-
ignition combustion process, as shown in Fig. 11 (b). In Case 7, the additional air facilitates
the early flame propagation but results in lower IMEP, similar to that in Case 4. In Case 8,
the dilution strategy adopted can effectively lower the PRRmax with marginal decrease in
IMEP compared to that of Case 5. As indicated by the pressure traces, the optimized fuel/air
equivalence ratio distribution in Case 8 enhances the early flame propagation process.
However, the auto-ignition combustion in Case 8 is less affected by the increased air dilution
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
when the amount of residual gas remains constant. Therefore, the IMEP only shows slight
decrease compared to that of Case 5. In Case 9, more intake fresh air is introduced and leads
to higher in-cylinder pressure even before the spark ignition because of the increased total in-
cylinder charge. However, the over-diluted condition leads to weak auto-ignition process and
it is hard to observe the transition from SI to CAI combustion from the pressure trace.
Correspondingly, the IMEP is significantly reduced, as shown in Fig. 11 (b).
The above results have shown that the dilution strategies can have significant impact on SFI
hybrid combustion under high compression ratio operations. The dilution strategy in Case 8
with additional air charge enhances SFI hybrid combustion performance with acceptable
PRRmax. However, too much intake air would lead to over-diluted mixture and deteriorate the
SFI hybrid combustion performance, as shown in Case 9.
464
465
466
467
468
469
470
471
472
473
474
475
476
Fig. 11. (a) in-cylinder pressure traces and (b) IMEP and PRRmax for the SFI combustion with
different dilution strategies. The CA50 is fixed around 2.6 ºCA bTDC.
In order to demonstrate the SFI hybrid combustion with different dilution strategies, the in-
cylinder dilution and thermal conditions are analysed. Fig. 12 compares the average fuel/air
equivalence ratio (ϕair) and temperature in different zones. As expected, the local fuel/air
equivalence ratio in each zone shows decreasing trend with the overall ϕair decreasing from
Case 5 to Case 7. The temperature in Case 6 is slightly higher than that in Case 5 because of
the increased specific heat ratio of the in-cylinder charge. In Case 7, the average temperature
in each zone is significantly reduced because of the reduction of the internal residual gas. The
addition of intake fresh air without sacrificing internal residual gas in Case 8 can lead to
higher in-cylinder temperature because of the heating effect from hot residual gas and
increased total dilution mass. The overall fuel/air equivalence ratio in Case 8 is kept the same
as Case 7. This leads to similar local fuel/air equivalence ratio in different zones between
Case 8 and Case 7. The addition of further intake air in Case 9 lowers the overall ϕair and
local ϕair in different zones and meanwhile increases the in-cylinder temperature slightly
because of the increased dilution mass.
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
Fig. 12. In-cylinder local fuel/air equivalence ratio (ϕair) and temperature in different zones at
40 ºCA bTDC.
Fig. 13 and Fig. 14 compare the mass burned fraction (MFB) traces of the SFI hybrid
combustion and the auto-ignition tendency of the mixture in the representative outer region
(Zone 7) with different dilution strategies, respectively. The auto-ignition in Case 6 is
significantly enhanced because of appropriate fuel/air equivalent ratio in central region
although the early flame propagation is weakened because of the delayed spark timing. In
Case 7, the local fuel/air equivalence ratio in central region is closer to 1.1, which
significantly enhances the early flame propagation process. However, the significantly
decreased in-cylinder temperature, as shown in Fig. 12, leads to slower subsequent auto-
ignition combustion.
The addition of intake air in Case 8 leads to a similar distribution of fuel/air equivalence ratio
to Case 7 and results in a stronger flame propagation than Case 5. Although the early flame
propagation in Case 8 is weakened slightly compared to that in Case 7 because of the
increased total dilution mass, the subsequent auto-ignition process is enhanced because of the
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
elevated charge temperature by the presence of the hot residual gas. In Case 9, the fuel/air
equivalence ratio of the mixture in the central region is around 1.1 because of the further
dilution by additional intake air, which further enhances the early flame propagation.
However, the additional intake air also leads to over-diluted mixture in the outer region,
which significantly deteriorates the subsequent auto-ignition process, as indicated in Fig. 13
and 14.
Therefore, both the in-cylinder thermal and dilution condition are vital to achieve better SFI
hybrid combustion performance. The thermal condition in Case 7 is not sufficient while the
dilution condition is not appropriate in Case 9, which both led to poor combustion
performance. Comparatively, the thermal and dilution conditions in Case 8 are suitable to
achieve better combustion performance.
Fig. 13. The mass fraction burned (MFB) traces of the SFI hybrid combustion with different
dilution strategies, and the CA50 is fixed around 2.6 ºCA bTDC.
513
514
515
516
517
518
519
520
521
522
523
524
525
526
Fig.14. The average auto-ignition tendency of the mixture in Zone 7 for the SFI hybrid combustion with different dilution strategies.
Fig. 15 shows the flame propagation dominated combustion duration (D1), the auto-ignition
dominated combustion duration (D2) and the ratio of the accumulated heat released (RCAT) at
the transition point CAT. The comparison between Fig. 8 and 15 indicates that the effect of
dilution strategies in Case 2/5, Case 3/6 and Case 4/7 on the combustion duration and RCAT
shows similar trends under different compression ratio operations. However, the increased
compression ratio leads to shorter combustion duration under different dilution strategies.
Compared to Case 7, the lean boosted dilution in Case 8 elevates thermal condition and
decreases the dependency of auto-ignition on the early flame propagation, leading to lower
RCAT (19.76%) and shorter flame propagation dominated combustion duration (D1).
Meanwhile, the later auto-ignition process is also enhanced, leading to shorter auto-ignition
dominated combustion duration (D2). But it should be noted that the combustion duration in
Case 8 is still longer than that in Case 5, which is responsible for the lower PRRmax. The
additional air dilution in Case 9 leads to increased dependency of subsequent auto-ignition on
early flame propagation because of the over-diluted condition in the outer region. This in turn
increases the RCAT and flame propagation dominated combustion duration (D1) and auto-
527
528529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
ignition dominated combustion duration (D2). The prolonged combustion duration in Case 9
leads to incomplete combustion and deteriorate IMEP dramatically.
Fig. 15. The flame propagation dominated combustion duration (D1), the auto-ignition dominated combustion duration (D2) and the ratio of the accumulated heat released (RCAT) at
CAT.
Fig. 16 compares the peak IMEP and the corresponding PRRmax of the SFI hybrid combustion
with different dilution strategies and compression ratios. The SFI hybrid combustion with a
lower compression ratio (CR=10.66:1) shows lower IMEP and PRRmax than the baseline Case
1 in which the stoichiometric homogenous charge is combusted by traditional SI-CAI hybrid
combustion. Both IMEP and PRRmax become higher with the increased compression ratio. By
replacing a part of residual gas by fresh air in Case 7, both PRRmax and IMEP are lowered
notably. In comparison, by adding the air to the cylinder charge with the same amount of
residual gas in Case 8, there is a 17.5% decrease in PRRmax and slight decrease (3.66%) in
IMEP compared to the baseline Case 1. The over-diluted mixture in Case 9 leads to the
lowest IMEP although the PRRmax is significantly reduced. The above results indicate that the
SFI hybrid combustion with the proposed dilution strategy in Case 8 shows better combustion
546
547
548
549
550551552
553
554
555
556
557
558
559
560
561
562
563
564
performance. The significantly reduced PRRmax in Case 8 also indicates the potential to
optimize the hybrid combustion performance through adjusting the direct injection ratio.
Fig.16. The peak IMEP and the corresponding PRRmax for the SFI hybrid combustion with
different dilution conditions and compression ratios.
3.3. Optimization of SFI hybrid combustion by the direct injection ratio
In this section, the effect of direct injection ratio (rDI) is analysed for the higher compression
ratio operations as the percentage of direct injection varied from 50% to 0%. As shown in
Fig. 17, the homogeneous hybrid combustion (Case 12) is characterised with both the highest
IMEP and PRRmax. The SFI hybrid combustion with direct injection reduces PRRmax. With rDI
=0.16, the PRRmax of the SFI hybrid combustion is significantly reduced to around 2.2
bar/ºCA and the IMEP values show slight reduction. As the direct injection ratio is increased
further to 28% and 50%, the enriched central mixture around spark plug advances the
combustion phasing but slows down the auto-ignition combustion of the leaner premixed
mixture, leading to reduced IMEP. The above results indicate the existence of the optimal rDI
to achieve the air-diluted SFI hybrid combustion with both higher IMEP and lower PRRmax. It
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
is found in this study that a small quantity of direct injection (i.e. rDI =0.16) is preferred to
obtain the optimal SFI hybrid combustion.
Fig. 17. Effect of direct injection ratio (rDI) on CA50, IMEP and PRRmax.
Fig. 18 directly compares the peak IMEP and the corresponding PRRmax of the SFI hybrid
combustion with different rDI. The homogeneous hybrid combustion can obtain highest IMEP
of 3.66 bar, which is 11.59% higher than that of the baseline Case 1. However, the PRRmax of
the homogeneous hybrid combustion is 10.39 bar/ºCA, which is much higher than the
acceptable limit of 5 bar/ºCA for a practical engine. With rDI of 0.16, the peak IMEP can
achieve 3.41 bar, which is 3.96% higher than that of baseline Case 1. Meanwhile, the
corresponding PRRmax is as low as 2.11 bar/ ºCA. The relative low PRRmax indicates the
potential to further elevate IMEP with a lower rDI (<0.16).
582
583
584
585
586
587
588
589
590
591
592
593
594
Fig. 18. Peak IMEP and the corresponding PRRmax with different rDI.
Fig. 19 shows the fuel/air equivalence ratio distributions at 36 ºCA bTDC for a fixed spark
timing at 35 ºCA bTDC. As shown in the figure, the local fuel/air equivalence ratio of the
mixture in the central region gradually increases with the rDI. The local ϕair in Zone 1 has
exceeded 1.5 in Case 10, and consequently leads to slower flame propagation process, as
shown by the MFB traces in Fig. 20. The local ϕair in central region is closest to 1.1 in Case 8
and leads to the fastest flame propagation process. In Case 11, the mixture in the central
region is a little leaner for the flame propagation and leads to moderate flame propagation
process among three cases.
On the other hand, the local ϕair in the outer region gradually decreases with rDI. The local ϕair
in the outer region in Case 10 with highest rDI is as lean as 0.4 and leads to highest in-cylinder
fuel stratification from central to outer region. As a consequence, the over-lean condition in
outer region deteriorates the auto-ignition and leads to incomplete combustion and lowest
IMEP in Case 10, as indicated in Fig. 17 and 18.
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
Fig. 19. In-cylinder fuel/air equivalence ratio (ϕair) at 36 ºCA bTDC. The spark timing is
fixed at 35 ºCA bTDC.
Fig. 20. The mass fraction burned (MFB) traces of the SFI hybrid combustion with different
dilution strategies. The spark timing is fixed at 35 ºCA bTDC.
The previous study on stoichiometric SFI combustion [19] has shown that the higher IMEP is
always accompanied with higher PRRmax when rDI is reduced to obtain a more homogeneous
SFI combustion. Although the trade-off between higher IMEP and lower PRRmax can also be
observed when rDI is reduced from 0.5 to 0.28 with the lean boosted dilution strategy, both
611
612
613
614
615
616
617
618
619
620
621
622
higher IMEP and lower PRRmax can be obtained with an rDI of 0.16 as shown in Fig. 18. In
order to explain the inherent reason for the higher IMEP and lower PRRmax with rDI of 0.16,
the detailed analysis of the SFI hybrid combustion with different rDI is performed. The
combustion phasing of all cases analysed is fixed around 0.8 ºCA bTDC where Case 8
obtains peak IMEP.
Fig. 21 shows the MFB traces of the SFI hybrid combustion with different rDI. The crank
angles with mode transitions (CAT) and maximum PRR (CAPRRmax) are also marked in the
figure. In order to maintain the same combustion phasing, the spark timing has to be delayed
to 20 ºCA bTDC in Case 10 because of the relatively higher heat release rate during the early
stage of the auto-ignition combustion. However, the auto-ignition is gradually weakened at
the later stage of the auto-ignition in Case 10 because of the gradually diluted mixture in the
outer region. On the contrary, the heat release rate of the auto-ignition combustion in Case 11
is moderate and the spark timing has to be advanced to 42 ºCA bTDC in Case 11, which can
be observed in Fig. 21.
Fig. 21. The mass fraction burned (MFB) traces of the SFI hybrid combustion with different
dilution strategies. The CA50 is fixed around 0.8 ºCA bTDC.
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
It is noted in Fig. 21 that the peak of PRR normally occurs just after CAT, i.e. at the early
stage of the auto-ignition stage. Therefore, the control of the early stage of the auto-ignition
combustion is essential to control the PRR in SFI hybrid combustion. Fig. 22 shows the iso-
surface with the local fuel/air equivalence ratio (ϕair) of 1 and the early auto-ignited sites after
CAT. In Case 10 (rDI =0.5), the diameter of the iso-surface with local ϕair of 1 is significantly
larger, which can also be inferred from Fig. 19. As shown in Fig. 22, the early auto-ignition
sites in Case 10 are surrounded and far from the iso-surface with local ϕair of 1, indicating the
early auto-ignition takes place in the region with richer mixture. With the rDI decreasing, the
iso-surface with local ϕair of 1 gradually shrinks, and the auto-ignition sites are closer to the
iso-surface with local ϕair of 1.
Fig. 22. Iso-surface with the fuel/air equivalence ratio of 1 and the auto-ignition sites after
mode transition.
The relationship between the auto-ignition sites and the iso-surface with local ϕair of 1, as
shown in Fig. 22, indicates the early stage auto-ignition behaviour. In Case 10, the early auto-
ignition mainly occurs in the fuel-rich region with larger charge cooling and stratification,
leading to slower auto-ignition process, reflected by the lower auto-ignition tendency in Fig.
23. On the other hand, the auto-ignition in the outer region, e.g. Zone 7, is also slowed down
in Case 10 because of the leaner mixture in these regions. Actually, the over-lean mixture is
hard to auto-ignite, leading to incomplete combustion and lower IMEP in Case 10. In Case
11, the smallest rDI leads to more homogeneous mixture with least charge cooling effect,
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
leading to faster development of auto-ignition process, reflected by the highest auto-ignition
tendency in Fig. 23. In theory, the auto-ignition tendency in Case 8 (rDI = 0.28) should locate
between that in Case 10 and Case 11. However, it is interesting to find that in Case 8 the
auto-ignition process in Zone 2 is comparable to that in Case 11. As indicated in Fig. 19, the
average ϕair of the mixture in Zone 2 in Case 8 is around 0.87 and slightly higher than that in
Case 11, which in turn leads to higher auto-ignition tendency.
Fig. 23. The traces of the average auto-ignition tendency in different zones.
In addition to the evolution of the auto-ignition tendency, the available fuel/air mixture in
these auto-ignited cells also plays an important role on the heat release process of auto-
ignition in the SFI hybrid combustion. The fuel/air equivalence ratio distribution brought by
different rDI actually changes the balance of the competition between the flame propagation
and early auto-ignition process in the central region. Fig. 24 shows the distribution of the
ratio of the fuel consumed by flame propagation (rSI) in these earliest auto-ignited cells (5%
of total cell number at TDC). In Case 10, the over-rich mixture in the central region leads to
slower flame propagation process, leading to lower rSI in these auto-ignited cells. Therefore,
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
the accumulated heat release rate of the early auto-ignition in the fuel-rich mixture would be
very significant once the auto-ignition occurs although the auto-ignition tendency is lowest
(Fig. 23). This explains the delayed spark timing in Case 10 to maintain combustion phasing.
In Case 11, the fuel/air equivalence ratio in central region is around 1.05, which enhances the
early flame propagation process. This leads to significantly higher rSI in these early auto-
ignited cells, indicating lower heat release from auto-ignition. The overwhelming flame
propagation over auto-ignition in central region resulted from the slightly rich mixture
explains the slower heat release process in Case 10, as shown in Fig. 21, although the
corresponding auto-ignition tendency is highest in Fig. 23.
The rSI of the early auto-ignited cells in Case 8 is obvious lower than that in Case 11 because
the mixture is a little richer (ϕair =1.2 in Zone 1) for fast flame propagation. This would leads
to increased heat release from auto-ignition. On the other hand, the increased auto-ignition
tendency in Zone 2, as shown in Fig. 23, also contributes to the increased heat release rate in
Case 8. This finally leads to the highest instantaneous heat release rate as shown in Fig. 21,
and hence the highest PRRmax in Fig. 17.
Fig. 24. Distribution of the ratio of the fuel consumed by flame propagation (rSI) in the early
auto-ignited cells (5% of total cell number at TDC).
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
3.4. Discussion of the effect of thermal and dilution conditions on controlling SFI hybrid
combustion
The SI-CAI hybrid combustion is characterized with early spark ignited flame propagation
and subsequent auto-ignition process. The competition between flame propagation and auto-
ignition dominates the behaviour of SI-CAI hybrid combustion. The introduction of a
stratified mixture through the direct injection enables the control of subsequent auto-ignition
by the stratified flame. This combustion mode was termed as stratified flame ignition (SFI)
hybrid combustion. In this study, it is found the in-cylinder thermal and dilution conditions
show significant impact on SFI hybrid combustion.
First, the auto-ignition process in SFI hybrid combustion shows high sensitivity to the in-
cylinder thermal conditions. The key issue in SFI hybrid combustion is the adjustment of the
quantity and the thermal condition of the dilution components simultaneously. However,
different dilution components, i.e. fresh air, external exhaust gas and internal residual gas,
show different thermal properties and dilution effects on combustion process. The fresh air
can be used to optimize the fuel/air equivalence ratio distribution, which is very effective for
the improvement of the flame propagation process, as shown in Case 3 and 6. The internal
residual gas is favorable to enhance the auto-ignition because of its heating effect. Therefore,
the combustion process would be deteriorated with lower internal residual gas in Case 4 and
7. The external exhaust gas is a pure dilution medium which shows no direct impact on
air/fuel equivalence ratio and thermal conditions. As a result, the SFI hybrid combustion with
constant dilution mass shows high sensitivity to these dilution strategies that it obtains either
high PRRmax or low IMEP.
Therefore, the optimal dilution strategy should meet the basic demand of the thermal
condition to achieve stable auto-ignition combustion, especially for the SFI hybrid
combustion with leaner mixture in the outer region. In this study, the NVO strategy was used
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
to achieve SFI hybrid combustion. Therefore, the internal residual gas fraction is responsible
to maintain appropriate thermal conditions and achieve efficient SFI hybrid combustion.
Otherwise, the incomplete combustion would be occurred and IMEP is deteriorated
dramatically, although the early flame propagation is enhanced with the dilution strategy, as
shown in Case 4 and 7.
Secondly, the SFI hybrid combustion process shows high sensitivity to the in-cylinder
dilution condition, especially to the fuel/air equivalence ratio distribution. As indicated
above, the later combustion process in SFI hybrid combustion, mainly characterized by the
auto-ignition process, is slowed down in the leaner mixture (local ϕair <1) in the outer region.
Therefore, the regulation of early combustion process is essential to control the PRRmax in SFI
hybrid combustion. It can be inferred from this study that the dilution conditions of the
central mixture controls the early heat release of SFI hybrid combustion through the
adjustment of the balance between flame propagation and auto-ignition. Once the mixture in
central region is too rich, the auto-ignition overwhelms the flame propagation, leading to
higher accumulated heat release from the auto-ignition of the fuel-rich mixture. However, the
increased fuel stratification would slightly slow down the heat release rate. When the mixture
in central region is slightly richer than the stoichiometry, the flame propagation overwhelms
the auto-ignition, leading to less contribution of fast auto-ignition to the heat release rate.
These results reveal the inherent mechanism of lower PRRmax for the SFI hybrid combustion
with rDI of 0.5 and 0.16. That is to say, both the degree of fuel stratification (or homogeneity)
and the specific local fuel/air equivalence ratio distribution dominate the heat release process
of SFI hybrid combustion. The former mainly controls the auto-ignition process itself, and
the latter mainly controls the competition between flame propagation and auto-ignition.
However, the higher rDI (e.g. 0.5) would inevitably leads to over-lean mixture in outer region
and deteriorates later auto-ignition process and IMEP. Therefore, the optimal rDI of the air-
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
diluted SFI combustion for the selected engine operation condition is 0.16 in this study,
which can achieve both higher IMEP and lower PRRmax.
Table 4 has been compiled to summarise the findings based on four typical fuel/air
equivalence ratio distribution patterns. In order to facilitate the description, three different
combustion regimes are proposed, including pure flame propagation zone, hybrid combustion
zone and pure auto-ignition zone. The symbols “+”, “○”, “–” and their combinations are used
to qualitatively indicate local fuel/air equivalence ratio and their impact on heat release rates
and combustion performances. It should be noted that “○” represents the stoichiometric
condition of the fuel/air equivalence ratio distribution characteristics in the first part of the
table. With this method, Pattern 1 indicates a strong stratification with over-rich mixture in
flame propagation zone while over-lean mixture in the auto-ignition zone. Pattern 2 indicates
a moderate stratification while Pattern 3 indicates a slight stratification. At last, Pattern 4
indicates the homogeneous lean mixture.
As shown in Table 4, Pattern 1 slows down both flame propagation and auto-ignition process
and deteriorates IMEP, as in Case 10. The moderate stratification, as revealed by Pattern 2 in
the table, leads to slightly weaker flame propagation and auto-ignition process. The slight
stratification in Pattern 3 with slightly richer mixture in flame propagation zone and slightly
leaner mixture in auto-ignition zone ensure a relatively stronger flame propagation and auto-
ignition process, maintaining the IMEP. On the other hand, the heat release rate in the hybrid
combustion zone is suppressed because a larger amount of mixture is consumed by strong
flame propagation, which ensures a lower PRRmax. The homogeneous lean mixture in Pattern
4 would enhance the auto-ignition process because of the lack of stratification although the
flame propagation is slightly weakened, which finally leads to higher PRRmax, as indicated in
Case 12.
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
Table 4 Typical fuel/air equivalence ratio distribution patterns and their impact on the air-
diluted SFI hybrid combustion.
1. Fuel/air equivalence ratio distribution characteristics
Pattern 1 Pattern 2 Pattern 3 Pattern 4
flame propagation zone + + + + + + –
hybrid combustion zone + + + ○ –
auto-ignition zone – – – – – – –
2. Combustion characteristics (heat release rate)
Flame propagation – – – + –
Hybrid combustion + + + ○ ○
Auto-ignition – – – ○ +
3. Combustion performances
IMEP – – – ○ +
PRRmax – + ○ +
Because of the competition between flame propagation and auto-ignition combustion under
stratified conditions, Pattern 3 shows promising potential to achieve optimal performance of
air-diluted SFI hybrid combustion. In this case, the in-cylinder stratified mixture avoids over-
rich mixture in the central region around spark plug to achieve both higher IMEP and lower
PRRmax. Meanwhile, the mixture in outer region is not too lean to achieve complete auto-
ignition combustion at outer region.
4. Summary and conclusions
In this paper, results by the validated 3D CFD simulations are presented and discussed of
the effect of dilution strategies and direct injection ratios on the stratified flame ignition (SFI)
hybrid combustion. The combination of port fuel injection (PFI) and direct injection (DI) was
used to form the premixed lean/diluted mixture and a stratified charge, respectively. Effects
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
of dilution strategies with different combinations of fuel/air equivalence ratio (ϕair) and
fuel/dilution equivalence ratio (ϕdilution) were studied at two engine compression ratios. Then
the effect of direct injection ratio (rDI) was investigated to optimize the fuel/air equivalence
ratio distribution as well as the air-diluted SFI hybrid combustion performance. The main
findings can be summarized as follows:
(1) The dilution strategy shows significant impact on in-cylinder fuel/air equivalence ratio
distribute and thermal condition. Compared to the stoichiometric SFI hybrid combustion, the
air-diluted SFI hybrid combustion optimizes the early flame propagation process because of
the avoidance of over-rich mixture around spark plug. However, the hybrid combustion with
fixed dilution mass can hardly achieve both higher IMEP and lower PRRmax simultaneously
when replacing a part of external exhaust gas (Case 3/6) or internal residual gas (Case 4/7) to
achieve the air-diluted SFI hybrid combustion, which is more apparent under high
compression ratio operation.
(2) The lean boosted dilution strategy with additional intake air and sufficient internal
exhaust gas recirculation (iEGR) was proposed in Case 8 to address the trade-off between
IMEP and PRRmax in air-diluted SFI hybrid combustion. In this strategy, the slightly richer
mixture around spark plug enhances the early flame propagation, and the sufficient hot
residual gas ensures the auto-ignition of end-gas, which leads to relatively higher IMEP.
Meanwhile, the increased dilution mass and fuel stratification from central region to the outer
region effectively suppress the PRRmax. However, the quantity of the additional intake air
mass needs to be controlled as too much intake air would lead to over-diluted mixture and
deteriorate the SFI hybrid combustion performance, as shown in Case 9.
(3) The direct injection ratio (rDI) can directly regulate the in-cylinder fuel/air equivalence
ratio distribution and in turn affect the air-diluted SFI hybrid combustion. It is found that the
optimal SFI hybrid combustion with rDI of 0.16 in Case 11 can lead to simultaneous high
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
IMEP and low PRRmax. Too much stratified fuel charge (Case 10) would leads to deteriorated
IMEP because of the over-lean mixture at outer region, while the further decrease of rDI (Case
12) would leads to unacceptable PRRmax because of more homogeneous mixture.
(4) The auto-ignition combustion in the air-diluted SFI hybrid combustion shows high
sensitivity to the in-cylinder thermal conditions. In order to achieve efficient air-diluted SFI
hybrid combustion, the internal residual gas fraction in the dilution strategy should be
carefully managed to maintain sufficient thermal conditions and ensure stable auto-ignition
combustion.
(5) The in-cylinder fuel/air equivalence ratio distribution pattern dominates the balance of the
competition between flame propagation and auto-ignition in the air-diluted SFI hybrid
combustion. Three different combustion regimes, including pure flame propagation zone,
hybrid combustion zone and pure auto-ignition zone, are proposed to understand effect of
typical fuel/air equivalence ratio distribution patterns on the air-diluted SFI hybrid
combustion characteristics and performances. In order to obtain optimal hybrid combustion
with high IMEP and low PRRmax, the in-cylinder stratified mixture should avoid over-rich
mixture around spark plug. Meanwhile, the mixture in outer region should avoid over-lean
conditions to reduce the deterioration of auto-ignition combustion at outer region.
Funding
The study is a part of the State Key Project of Fundamental Research Plan (Grant
2013CB228403) supported by the Ministry of Science and Technology of China.