A numerical investigation of valve timing and variable compression ratio (vcr) strategies in a stroke volume modulated (svm) spark ignition (si) engine

Abstract

The efficiency of conventional spark-ignition (SI) engines decreases significantly, particularly under part-load conditions, due to the high pumping losses associated with throttle-controlled load management. The Atkinson cycle is fundamentally a thermodynamic cycle in which the effective expansion stroke (while the valves are closed) is longer than the compression stroke. In the conventional Otto cycle, these two strokes are generally equal; however, in the Atkinson cycle, the intake valve is closed later than that in normal (FSIVP, Forward Shifted Intake Valve Profile: shifting the intake valve opening and closing timings forward without changing the total valve opening duration), allowing part of the working gas to be pushed back into the intake manifold during a portion of the upward piston motion. FSIVP reduces throttle dependency by controlling the engine load through backflow rather than throttle restriction, thereby decreasing pumping losses. At the same time, while reducing the effective compression ratio (ECR), it maintains relatively a higher Effective Expansion Ratio/Effective Compression Ratio (EER/ECR), enabling more useful work to be extracted from the exhaust energy. FSIVP allows the excess charge to be pushed back into the intake manifold during the early stage of the compression stroke. This push-back mechanism provided by FSIVP, reduces ECR (since the actual compression process starts only after the intake valve closes), thereby improving the knock limit at high-load conditions. In addition, under part-load conditions, it enables load control without requiring excessive throttle restriction, leading to a reduction in brake specific fuel consumption (BSFC). Forward Shifted Exhaust Valve Profile (FSEVP) reduces the dependence of SI engines on throttle-based load control while also providing an internal exhaust gas recirculation (IEGR) effect. FSEVP allows a portion of the burned gases to remain inside the cylinder or to be reinducted from the exhaust manifold at the beginning of the next intake stroke, thereby reducing pumping losses (PMEP). When FSEVP is optimized together with FSIVP, it provides a dual advantage on engine operating characteristics: under high-load conditions, while the geometric compression ratio (GCR) remains constant, ECR is reduced through FSIVP, suppressing knock tendency. Stroke Volume Modulation (SVM) strategies enable the active cycles to operate at higher Indicated Mean Effective Pressure (IMEP) levels by flexibly varying the total piston displacement according to instantaneous engine torque demand and by reducing the firing frequency (CDA: a system in which only two selected cylinders operate instead of four cylinders; or Skip Cycle operation, where consecutive cycles are alternatively fired and skipped; or dynamic skip fire (DSF), where different numbers of cycles are flexibly deactivated depending on instantaneous engine torque demand). first objective of this study is building up a 1-D engine model in order to validate against normal and SVM mode experimental data. Experimental results obtained from a four-cylinder, water-cooled, port-fuel injected (PFI) gasoline engine with a total displacement of 1.8 L operating under stoichiometric mixture conditions were used to develop a 1-D engine simulation model. At a constant engine load of 2 bar brake mean effective pressure (BMEP) and five different engine speeds (1200–1350–1500–1650- 1800 rpm) ranges, the experimental data were separately validated for both the normal and SVM modes at equivalent power levels. Subsequently, two different modes approach were proposed and implemented into the validated model. Within this scope, the validated SVM model was optimized by integrating variable valve timing (VVT) including Forward Shifted Exhaust Valve Profile (FSEVP) and Forward Shifted Intake Valve Profile (FSIVP) methods; Variable Compression Ratio (VCR) corresponding to Atkinson cycle (FSIVP + VCR). Through the combined implementation of FSEVP and Atkinson-based optimization modes, this study aimed to achieve the highest possible brake thermal efficiency together with the lowest emission levels obtainable from the engine. In order to build up the base engine model, predictive engine parameters such as throttle valve position, burning duration (BDUR) and 50% fuel mass fraction burned (MFB50) are used to estimate the engine performance at 2 bar BMEP and five different engine speed ranges. BDUR is the rapid burning angle as a crankshaft interval required to burn the bulk of mixture charge and MFB50 corresponds to crankshaft angle where half of the fuel mass is burned due to the combustion. In this study, a broad screening analysis was conducted over wide parameter ranges to identify the most influential variables affecting engine performance, combustion characteristics, and emission formation. In the second stage, a refined optimization process was performed within the sensitive parameter regions identified during the initial screening stage, allowing a more precise calibration of the simulation model. Within this optimization framework, the effects of FSEVP, FSIVP, and VCR were systematically investigated together with ignition timing variations. The primary objective of the optimization process was to determine the most suitable operating configurations capable of simultaneously reducing Nitrogen Oxide (NOx) and Unburned Hydrocarbon (UHC) emissions while also improving BSFC. Through the combined evaluation of these flexible valve timing and compression ratio (CR) configurations, this study aimed to identify optimum calibration regions obtaining improved combustion efficiency, lower pumping losses (PMEP), and enhanced overall engine efficiency (ηb). Although SVM mode provides considerable fuel economy benefits, it also leads to higher regulated emissions, particularly NOx and UHC, compared to conventional four-cycle engine operation. The second objective of this study is to implement two VVT modes and flexible GCRs into the available SVM mode by optimization of intake&exhaust valve activation and GCR adjustment in order to increase EER/ECR greater than 1, resulting a significant reduction in increased exhaust pollutants and also further fuel economy in comparison to the standalone SVM. In SVM+FSEVP+FSIVP+VCR configuration, increasing GCR is balanced by retarding the Intake Valve Closing (IVC) to maintain an equivalent power range. While forward shifted IVC generally lowers the compression pressure; at higher GCR levels, the geometric compression begins to dominate the cycle, eventually leading to engine knock. In order for investigating two proposed VVT strategies and flexible GCRs, a gas dynamics 1-D numerical model based on mean value model (MVM) approach, net power cycle definition, time-dependent fluid governing equations and separated twozones gas approach is applied for a four-cylinder, water-cooled and port-fuel injected SI engine with SVM mode at steady state engine simulation cases including a constant low-load (BMEP: 2 bar) and five different engine speed (1200–1350-1500-1650-1800 rpm) ranges. After the validation process of both normal and SVM modes against the experimental results, valve timing strategies associated different GCRs integrated into the validated standalone SVM model. The numerical findings under 2 bar BMEP load are summarised as follows: Using standalone SVM mode result as the reference point, the benefits of SVM+FSEVP+FSIVP+VCR configuration became more evident with increasing GCR. At 1500 rpm, BSFC decreased to 466.5, 452, 426, and 408.7 g/kWh for GCR 10:1, 11:1, 12:1, and 13:1, corresponding to reductions of 8.82%, 11.65%, 16.73%, and 20.1% compared to the standalone SVM, respectively. These findings confirmed that the Atkinson cycle effect is highly effective at mid-range speeds, providing a total fuel consumption reduction of 27.7% compared to normal mode at GCR 13:1. SVM+FSEVP+FSIVP+VCR configuration also pointed out a dramatic reduction in NOx formation. At 1200 rpm, NOx emissions decreased to 130.9 ppm at GCR 10:1, corresponding to a 77.9% reduction compared to the standalone SVM. The reduction is mainly associated with the lower ECR and reduced end-of-compression temperature caused by the Atkinson cycle effect. As GCR increased to 11:1 and 12:1, NOx emissions increased slightly to 150.4 ppm and 181.5 ppm, but still maintained reductions of 74.6% and 69.3%, respectively. At GCR 13:1, NOx decreased again to 130.7 ppm, maintaining an overall 77.9% reduction due to the significant spark retardation (closer to TDC) required for knock suppression. At 1500 rpm, the normal mode at GCR 9:1 produced 3164.9 ppm UHC emissions, while the standalone SVM increased this value to 4921.9 ppm, corresponding to a 55.5% increase due to severe vacuum conditions and oil suction into the combustion chamber. Compared to the standalone SVM, SVM+FSEVP+FSIVP+VCR configuration reduced UHC emissions to 3088.9, 3321.2, 3161.7, and 3040.5 ppm at GCR 10:1, 11:1, 12:1, and 13:1, corresponding to improvements of 37.2%, 32.5%, 35.8%, and 38.2%, respectively. At 1500 rpm, the normal mode at GCR 9:1 exhibited the widest pumping loop and deepest intake vacuum region, whereas standalone SVM substantially suppressed the pumping loop area and reduces PMEP/GMEP ratio from 0.174 to 0.025. In SVM+FSEVP+FSIVP+VCR configurations, increasing GCR from 10:1 to 13:1 slightly increased PMEP/GMEP ratio from 0.145 to 0.161 althoguh the combined effect of advanced expansion characteristics and modified valve timing (an increase in EER/ECR). At 1500 rpm, the normal mode exhibited the lowest in-cylinder pressure range due to the throttled low-load operation and limited trapped fresh charge mass. In comparison, standalone SVM mode increaseed the maximum in-cylinder pressure (pmax) from 10.25 bar to 30.21 bar and moved crank angle at maximum in-cylinder pressure (tpmax) from 30 CAD aTDC to nearly 11 CAD aTDC. The introduction of SVM+FSEVP+FSIVP+VCR configuration further intensified combustion characteristics, increasing pmax from 30.5 bar to 32.84 bar between GCR 10:1 and 11:1 while tpmax toward closer TDC crank angles.