Search

Transient engine operation of direct-injection spark ignition engines can result in high particulate number emissions. To investigate the causes of soot formation, an engine test rig was developed to perform detailed measurements of real transient operation. For this purpose, a single-cylinder full-metal engine with a real combustion chamber geometry was equipped with minimally invasive optical accesses. Simultaneous high-speed endoscopic PIV, spray visualization, and combustion imaging were applied to investigate the in-cylinder processes in detail. Endoscopic PIV was first compared in the central symmetry plane with classical PIV performed at the equivalent optical engine at steady-state operation for verification. Then the engine parameters of a tip-in performed by the corresponding four-cylinder engine, which led to high particle number emissions, were applied to the single-cylinder engine. The engine parameters were in a good agreement and particle number emissions due to the maneuver were within the same range of several 10⁶ #/cm³. In total, 19 repetitions of the tip-in maneuver were analyzed with respect to the in-cylinder processes and repeatability of engine parameters. Furthermore, the in-cylinder flow field during the late compression stroke, flame propagation, and soot luminosity of single cycles during the tip-in indicated cause-and-effect chains for the formation of pool fire and soot at the injector tip. The direction of the flow below the spark plug influenced the direction of flame propagation. An early arrival of the flame enhanced the formation of soot from fuel films formed on the piston surface or at the injector. In engine applications, counter measures can be applied to reduce the particle number emissions when accounting for these indicated cause-and-effect chains.

An investigation of the interaction between the in-cylinder flow and the spray topology in two spray-guided direct injection optical engines is reported. The bulk flow field in the combustion chamber is characterized using particle image velocimetry. Geometrical parameters such as the axial penetration and the spray angle of the liquid spray are measured using Mie scatter imaging and/or diffuse back-illumination. The measured parameters are compared with data from a constant volume chamber available in the literature. For a late injection strategy, the so-called ECN Spray G standard condition, the mean values of the spray penetration do not seem to be significantly perturbed by the in-cylinder flow motion until the plumes approach the piston surface. However, spray probability maps reveal that cycle-to-cycle fluctuations of the spatial distribution of the liquid spray are affected by the magnitude of the in-cylinder flow. Particle image velocimetry during injection shows that the flow field in the vicinity of the spray plumes is heavily influenced by air entrainment, and that an upward flow in-between spray plumes develops. Consistent with previous research that demonstrated the importance of the latter flow structure for the prevention of spray collapse, it is found that increased in-cylinder flow magnitudes due to increased intake valve lifts or engine speeds enhance the spray-shape stability. Compared with cases without injection, the influence of the spray on the in-cylinder flow field is still noticeable approximately 2.5 ms after the start of injection.

Mid-level ethanol/gasoline blends can provide knock resistance benefits for stoichiometric spark-ignition engine operation, but previous studies have identified challenges associated with spray impingement and wall wetting, leading to excessive particulate matter emissions. At the same time, stratified-charge spark-ignition operation can provide increased thermal efficiency, but care has to be exercised to avoid excessive in-cylinder soot formation. In support of the use of mid-level ethanol/gasoline blends in advanced spark-ignition engines, this study presents spray and fuel-film measurements in a direct-injection spark-ignition engine operated with a 30 vol.%/70 vol.% ethanol/gasoline blend (E30). Crank-angle resolved fuel-film measurements at the piston surface are conducted using two different implementations of the refractive index matching technique. A small-angle refractive index matching implementation allows quantification of the wetted area, while a large-angle refractive index matching implementation enables semi-quantitative measurements of fuel-film thickness and volume, in addition to fuel-film area. The fuel-film measurements show that both the amount of fuel deposited on the piston and the shape of the fuel-film patterns are strongly influenced by the injection timing, duration, intake pressure, and coolant temperature. For combinations of high in-cylinder gas density and long injection duration, merging of the individual spray plumes, commonly referred to as spray collapse, can cause a dramatic change to the shape and thickness of the wall fuel films. Overall, the study provides guidance to engine designers aiming at minimizing wall wetting through tailored combinations of injection timings and durations.

In the present paper, direct numerical simulation (DNS) and particle image velocimetry (PIV) have been applied complementarily in order to generate a database of near-wall turbulence properties of a highly turbulent jet impinging on a solid surface under different inclination angles. Thereby, the main focus is placed on an impingement angle of 45°, since it represents a good generic benchmark test case for a wide range of technical fluid flow applications. This specific configuration features very complex flow properties including the presence of a stagnation point, development of the shear boundary layer and strong streamline curvature. In particular, this database includes near-wall turbulence statistics along with mean and rms velocities, budget terms in the turbulent kinetic energy equation, anisotropy invariant maps, turbulent length/time scales and near-wall shear stresses. These properties are useful for the validation of near-wall modeling approaches in the context of Reynolds-averaged Navier–Stokes (RANS) and large-eddy simulations (LES). From this study, in which further impingement angles (0°, 90°) have been considered in the experiments only, it turns out that (1) the production of turbulent kinetic energy appears negative at the stagnation point for an impingement angle other than 0° and is balanced predominantly by pressure-related diffusion, (2) quasi-coherent thin streaks with large characteristic time scales appear at the stagnation region, while the organization of the flow is predominantly toroidal further downstream, and (3) near-wall shear stresses are low at the stagnation region and intense in regions where the direction of the flow changes suddenly.

Machine learning (ML) models based on a large data set of in-cylinder flow fields of an IC engine obtained by high-speed particle image velocimetry allow the identification of relevant flow structures underlying cycle-to-cycle variations of engine performance. To this end, deep feature learning is employed to train ML models that predict cycles of high and low in-cylinder maximum pressure. Deep convolutional autoencoders are self-supervised-trained to encode flow field features in low dimensional latent space. Without the limitations ascribable to manual feature engineering, ML models based on these learned features are able to classify high energy cycles already from the flow field during late intake and the compression stroke as early as 290 crank angle degrees before top dead center (-290° CA) with a mean accuracy above chance level. The prediction accuracy from -290° CA to -10° CA is comparable to baseline ML approaches utilizing an extensive set of engineered features. Relevant flow structures in the compression stroke are revealed by feature analysis of ML models and are interpreted using conditional averaged flow quantities. This analysis unveils the importance of the horizontal velocity component of in-cylinder flows in predicting engine performance. Combining deep learning and conventional flow analysis techniques promises to be a powerful tool for ultimately revealing high-level flow features relevant to the prediction of cycle-to-cycle variations and further engine optimization.

We continue the investigation on the relations of QCDCL and QBF resolution systems. In particular, we introduce QCDCL versions that tightly characterise QU-Resolution and (a slight variant of) long-distance Q-Resolution. We show that most QCDCL variants - parameterised by different policies for decisions, unit propagations and reductions - lead to incomparable systems for almost all choices of these policies.

Quantified conflict-driven clause learning (QCDCL) is one of the main solving approaches for quantified Boolean formulas (QBF). One of the differences between QCDCL and propositional CDCL is that QCDCL typically follows the prefix order of the QBF for making decisions. We investigate an alternative model for QCDCL solving where decisions can be made in arbitrary order. The resulting system QCDCL^ANY is still sound and terminating, but does not necessarily allow to always learn asserting clauses or cubes. To address this potential drawback, we additionally introduce two subsystems that guarantee to always learn asserting clauses (QCDCL^UNI-ANI) and asserting cubes (QCDCL^EXI-ANY), respectively. We model all four approaches by formal proof systems and show that QCDCL^UNI-ANY is exponentially better than QCDCL on false formulas, whereas QCDCL^EXI-ANY is exponentially better than QCDCL on true QBFs. Technically, this involves constructing specific QBF families and showing lower and upper bounds in the respective proof systems. We complement our theoretical study with some initial experiments that confirm our theoretical findings.