Mapping K factor variations and its causes in a modern, spark-ignition engine

Highlights

• New method developed to determine the K factor in the Octane Index model.

• K maps obtained for a 2L, 4-cylinder, gasoline turbocharged direct-injection engine.

• K mostly negative (positive) with normal (high) operation temperatures.

• K correlated with the unburned gas temperature at the later stages of combustion.

Abstract

The Research Octane Number (RON) and the Motor Octane Number (MON) are primary fuel properties that characterize gasoline’s knock resistance in spark ignition engines. The utility of these two metrics, however, has been questioned with advancements in engine technologies that significantly change the thermochemical environment inside the cylinder. The Octane Index, OI = (1-K)·RON + K·MON, has therefore been proposed to characterize the knock resistance in modern engines where K weights the relative contribution of the RON and MON and is primarily determined by engine design and operating conditions. Quantifying the K factor is central to understanding a fuel’s anti-knock performance in a modern spark-ignition (SI) engine.

This work therefore determines the map of K over engine operating conditions for a 2-litre, 4-cylinder turbocharged, gasoline direct-injection engine. To achieve this, a novel blending system for primary reference fuels (PRFs) was developed to determine K by matching the knock resistance of a 91.6-RON certification gasoline with a PRF at each operating condition. The K values are determined over the engine map with normal and high intake air and coolant temperatures. At normal operating temperature, K is negative at most knock-limited conditions, consistent with previous findings, whereas at high operating temperatures, K is mostly positive and, indeed, exceeds 1 near 8 bar BMEP and 3000 RPM, demonstrating the relevance of the MON at some conditions of practical significance. Further analysis is conducted via engine simulations to examine the relationship between K and end gas conditions, and a strong correlation is observed between K and the unburned gas temperature at the later stages of combustion. This correlation is argued to have a sound physical basis in the engine’s thermochemistry, supporting the utility of this K factor method.

Introduction

Fuel economy regulations have been consistently driving the development of more efficient spark-ignition (SI) engines [1], [2], [3]. The application of modern technologies, such as high compression ratio, turbocharging, engine downsizing and downspeeding, promotes the tendency of engine knock. Various measures have been implemented to control knock, principally spark timing retard and fuel enrichment at some conditions. However, these measures have negative effects on engine efficiency and exhaust emissions.

The knock resistance of a fuel is characterized by two octane numbers – the Research Octane Number (RON) [4] and the Motor Octane Number (MON) [5]. These are determined with the Cooperative Fuel Research (CFR) engine using standard methods (Table 1). The RON and MON are determined by referencing the knock resistance of a test fuel to that of Primary Reference Fuels (PRFs) at corresponding conditions. PRFs are binary mixtures of iso-octane and n-heptane with the former defining an octane number of 100 and the latter defining an octane number of 0. The octane number of a PRF mixture is defined as the volumetric fraction of iso-octane in the mixture. As shown in Table 1, the RON is tested at an overall lower intake temperature than MON, and fuel vaporization cooling further reduces the mixture temperature past the carburetor. For most practical fuels, the RON is higher than the MON, but for PRFs their RON and MON are defined to be the same. The difference between RON and MON is known as octane sensitivity (S = RON – MON). The origin of S lies in the different autoignition chemistry between PRFs and octane sensitive fuels [6], [7], [8] as well as the different physical properties such as heat of vaporization (HoV) [9], [10].

Historically, the two octane tests were designed to bracket the operating conditions of typical engines in the 1930s, where higher RON and MON generally meant better anti-knock quality. However, modern engines with direct injection (DI), turbocharging and variable valve timing (VVT) have vastly different cylinder conditions from those in the octane rating tests. As a result, the relevance of RON and MON to modern engines has received significant attention. Various studies in gasoline turbocharged direct-injection (GTDI) engines found that fuels of higher RON and lower MON (i.e. higher S) showed higher knock resistance and therefore allowed engines to operate with higher efficiency at high-load conditions [13], [14], [15], [16], [17], [18]. Vehicle tests reported similar trends, where the acceleration performance improved with fuels of high RON and low MON [15], [19], [20]. These findings indicate the relevance of RON and MON to knock resistance in modern SI engines should be re-evaluated.

An alternative metric, Octane Index (OI), was proposed by Kalghatgi in 2001 [20], [21] to characterize a fuel’s actual anti-knock quality in modern engines. As given by Eq. (1), OI is a linear combination of the RON and MON. K is a weighting factor that is primarily dependent on engine design and engine operating conditions. For example, the anti-knock index (AKI) used in North America approximates the OI by evenly weighting the RON and MON to a fuel’s actual anti-knock quality (K = 0.5). The OI concept therefore generally reflects a fuel’s knock resistance relative to its RON and MON ratings. A more commonly used expression for OI is given by Eq. (2), where S is the octane sensitivity. For an octane sensitive fuel (S greater than 0), its OI can be larger than the RON (when K less than 0), or smaller than the MON (when K greater than 1), or in between, depending on the engine operation conditions. For PRFs, the OI always equals the RON and MON because S = 0. Although K is irrelevant in this case, PRFs provide a valuable reference for K determination, as shown later.

$$\mathit{OI}=\left(1-K\right)\bullet RON+K\bullet MON$$

$$\mathit{OI}=RON-K\bullet S$$

In principle, K can be used to quantify the relative significance of the RON and MON to SI engine combustion. For example, when K is greater than 0.5, the MON is more important than the RON, and the higher the MON, the stronger the knock resistance is (at that operating condition in that engine); however, when K is negative, a higher MON causes a lower OI. Indeed, previous studies in modern SI engines showed that K was negative at most high-load conditions and therefore concluded that fuels of lower MON were advantageous [15], [16], [17], [21], [22], [23]. However, in practical driving, engines primarily operate at part loads, and recent studies found that K could be positive at part-load, knock-limited conditions [24], [25]. Therefore, further study is needed to determine K over a wide range of conditions in modern SI engines.

The objective of this work, therefore, is to determine K values in the Octane Index model over the full operation map of a modern GTDI engine, including severe operating conditions that involve high intake air and/or high coolant temperatures. A novel fuel blending system is developed to continuously prepare and supply PRF mixtures of the needed composition in real time to match the anti-knock performance of a 92 RON/83 MON gasoline, from which the K of each condition is determined. The obtained K dataset is used to generate a contour map of K overlaid on the engine operation map. The comprehensive dataset also enables statistical analysis for potential correlations between K and the end gas states, of which fundamental understanding is currently lacking.