Cycle-to-cycle Variations Research Articles

Dynamic Mode Decomposition for the Comparison of Engine In-Cylinder Flow Fields from Particle Image Velocimetry (PIV) and Reynolds-Averaged Navier–Stokes (RANS) Simulations

Validation of Reynolds-averaged Navier–Stokes (RANS) simulation results against experimental data such as flow measurements from particle image velocimetry (PIV) remains a challenge for the development of thermal propulsion systems. This is partly due to cycle-to-cycle variations (CCVs) in the air motion and partly due to uncertainties in the PIV measurement technique, complicating the question of what constitutes a fair validation target for the RANS model. Indeed, an inappropriate validation target can misguide subsequent adjustments of a RANS model. In this work, the ensemble-averaged PIV field is first investigated for its suitability as a validation target for RANS simulations. The relevance index and the velocity histogram distance are used as quantitative metrics to assess the similarity of the ensemble-averaged field to the full dataset of individual PIV cycles. While a high similarity is seen between the average PIV flow field and the individual cycles on the tumble plane, the similarity is lower and more variable on the cross-tumble plane, where there are significant CCVs. Standard (space-only, phase-dependent) proper orthogonal decomposition (POD) is employed as an alternative method of data processing with the aim of providing a fairer comparison to RANS simulations. The cycle-dependence of the standard POD modes is shown to be an aspect that results in many validation targets and an excessively broad validation range, limiting its utility in this context. Dynamic mode decomposition (DMD) and sparsity-promoting dynamic mode decomposition (SPDMD) are then proposed as alternative solutions, capable of extracting flow structures at specific frequencies. The background 0 Hz SPDMD modes exhibit an ability to produce more realistic flow fields with velocity magnitudes that are significantly closer to the individual cycles.

Investigating the Origins of Cyclic Variability in Internal Combustion Engines Using Wall-Resolved Large Eddy Simulations

Abstract Modern internal combustion engines (ICE) operate at the ragged edge of stable operation characterized by high cycle-to-cycle variations (CCV). A key scientific challenge for ICE is the understanding, modeling, and control of CCV in engine performance, which can contribute to partial burns, misfire, and knock. The objective of this study is to use high-fidelity numerical simulations to improve the understanding of the causes of CCV. Nek5000, a leading high-order spectral element, open source code, is used to simulate the turbulent flow in the engine combustion chamber. Multicycle, wall-resolved large-eddy simulations (LESs) are performed for the General Motors (GM), Transparent Combustion Chamber (TCC-III) optical engine under motored operating conditions. The mean and root-mean-square (rms) of the in-cylinder flow fields at various piston positions are validated using particle image velocimetry (PIV) measurements during the intake and compression strokes. The large-scale flow structures, including the swirl and tumble flow patterns, are analyzed in detail and the causes for cyclic variabilities in these flow features are explained. The energy distribution across the different scales of the flow are quantified using one-dimensional (1D) energy spectra, and the effect of the tumble breakdown process on the energy distribution is examined. The insights from this study can help us develop improved engine designs with reduced cyclic variabilities in the in-cylinder flow leading to enhanced engine performance.

CRUS: A Hardware-Efficient Algorithm Mitigating Highly Nonlinear Weight Update in CIM Crossbar Arrays for Artificial Neural Networks

Mitigating the nonlinear weight update of synaptic devices is one of the main challenges in designing compute-in-memory (CIM) crossbar arrays for artificial neural networks (ANNs). While various nonlinearity mitigation schemes have been proposed so far, only a few of them have dealt with high-weight update nonlinearity. This article presents a hardware-efficient on-chip weight update scheme named the conditional reverse update scheme (CRUS), which algorithmically mitigates highly nonlinear weight change in synaptic devices. For hardware efficiency, CRUS is implemented on-chip using low precision (1-bit) and infrequent circuit operations. To utilize algorithmic insights, the impact of the nonlinear weight update on training is investigated. We first introduce a metric called update noise (UN), which quantifies the deviation of the actual weight update in synaptic devices from the expected weight update calculated from the stochastic gradient descent (SGD) algorithm. Based on UN analysis, we aim to reduce AUN, the UN average over the entire training process. The key principle to reducing average UN (AUN) is to conditionally skip long-term depression (LTD) pulses during training. The trends of AUN and accuracy under various LTD skip conditions are investigated to find maximum accuracy conditions. By properly tuning LTD skip conditions, CRUS achieves >90% accuracy on the Modified National Institute of Standards and Technology (MNIST) dataset even under high-weight update nonlinearity. Furthermore, it shows better accuracy than previous nonlinearity mitigation techniques under similar hardware conditions. It also exhibits robustness to cycle-to-cycle variations (CCVs) in conductance updates. The results suggest that CRUS can be an effective solution to relieve the algorithm-hardware tradeoff in CIM crossbar array design.

Cyclic Variations in the Flame Propagation in an Spark-Ignited Engine: Multi Cycle Large Eddy Simulation Supported by Imaging Diagnostics

Experimental measurements and multi-cycle large eddy simulation (LES) are performed in an optically accessible four-stroke spark-ignition engine to investigate cycle-to-cycle variations (CCV). High-speed combustion imaging is used to measure the early flame propagation and obtain the flame radius and centroids. Large Eddy Simulation generates data-bases for the flame propagation as well as the kinetic energy in the cylinder and confirms the observations from the two-dimensional fields by three-dimensional simulation results. Experiment and simulation are compared with respect to the strength and distribution of CCV. Both approaches reveal CCV causing similar statistics of maximum pressures and combustion speeds. The cycles are categorized as slow and fast cycles using the crank angle of ten percent burnt fuel-mixture. Analysis of the flame centroids shows that slow cycles move further towards the intake-side of the engine compared to fast cycles. The kinetic energy during combustion is averaged for the slow and fast cycles based on the samples being in unburnt and burnt mixture. Studying the kinetic energy level in the unburnt and burnt mixture reveals higher turbulent kinetic energy for the fast cycles as well as larger separation between the global kinetic and the turbulent kinetic energy for the slow cycles, providing evidence for a source of the CCV variations observed in this engine.

A Combined Numerical and Experimental Investigation of Cycle-to-Cycle Variations in an Optically Accessible Spark-Ignition Engine

A combined numerical and experimental investigation is carried out to analyze the cycle-to-cycle variations (CCV) in an optically accessible spark-ignition engine with port fuel injection. A stable and an unstable operating condition is considered. Well-established turbulence, combustion, and ignition models are employed in the large-eddy simulations (LES). High-speed measurements of the velocity field via particle image velocimetry and flame imaging in the tumble plane are conducted in the experiments. A detailed comparison between LES and experiments is carried out, including the in-cylinder pressure, the flow fields, the spatial flame distribution, and the fields conditioned on fast and slow cycles. Good agreement is achieved for the variables considering all cycles; yet, some discrepancies are observed for the conditionally averaged quantities. A systematic quantitative correlation analysis between the selected influencing variables and the CCV is presented, in which the influencing variables are extracted from different length scales (r = 3 mm, 12 mm, and 43 mm) and the CCV are distinguished between the early flame kernel development and later flame propagation. Even though the most relevant influencing parameters are different for the two operating conditions, the location of the coherent vortex structure is found to be important for the CCV of both cases.

Analysis of the in-cylinder flow of a DISI engine using high-speed particle image velocimetry

In addition to exploring electric drives, it is crucial to minimize emissions and improve the efficiency of transport based on internal combustion engines (ICE). For these, in particular, a significant reduction of emissions as well as a considerable efficiency of ICE are required. A reliable formation of an effective mixture prior to ignition and a fast, steady combustion process are two key elements that influence emissions and efficiency. While on the one hand, the air-fuel mixture is formed by strongly three-dimensional, time-dependent flow phenomena during the intake and compression stroke, the turbulence level of the flow near the approaching flame front has a direct impact on turbulent flame propagation. The flow field in the combustion chamber, on the other hand, is highly turbulent and sensitive to cycle-to-cycle variations (CCV). In order to understand the fundamental processes and interactions of the flow field, flow measurements with high spatial and temporal resolution are required. To analyze CCV, the resulting flow field data must be decomposed into its scales, since CCVs are different from random fluctuations due to turbulent flow motions. Besides velocity decomposition based on traditional Reynolds decomposition, which provides an average flow velocity and a fluctuation velocity, most studies include statistical and linear approaches such as proper orthogonal decomposition (POD) and Fast Fourier transformation (FFT). In this paper, time-resolved high-speed stereoscopic PIV (HS-SPIV) measurements on an optical single-cylinder engine are processed using a new nonlinear and transient decomposition method, namely the multivariate empirical mode decomposition (NA-MEMD) based on a data-driven decomposition algorithm. The decomposition is based on the inherent scales of the data and generates physically valid modal representations known as intrinsic mode functions (IMFs), which are arranged with respect to the contained scale values. The approach and the IMFs generated are described in detail using one engine cycle as an example. According to the results, three modes are obtained, the measurement noise and extremely tiny short-term scales, the turbulent scales, and the large-scale tumble vortex.

Modelling Study of Cycle-To-Cycle Variations (CCV) in Spark Ignition (SI)-Controlled Auto-Ignition (CAI) Hybrid Combustion Engine by Using Reynolds-Averaged Navier–Stokes (RANS) and Large Eddy Simulation (LES)

The spark ignition (SI)-controlled auto-ignition (CAI) hybrid combustion is characterized by early flame propagation combustion and subsequent auto-ignition combustion. The application of combined SI–CAI hybrid combustion can be used to effectively extend the operating range of CAI combustion and achieve smooth transitions between SI and CAI combustion modes. However, SI–CAI hybrid combustion can produce significant cycle-to-cycle variations (CCV). In order to better understand the sources of CCV and minimize its occurrence, the large eddy simulation (LES) and Reynolds-averaged Navier–Stokes (RANS) approaches were employed in this study to model and understand the cyclic phenomenon of SI–CAI hybrid combustion. Both the multi-cycle LES and RANS simulations were analyzed against the experimental measurements in a single cylinder engine at 1500 rpm and a 5.43 bar average indicated the mean effective pressure (IMEP). The detailed analysis of the in-cylinder pressure traces, IMEP, in-cylinder peak pressure (PP), peak pressure rise rate (PPRR) and the crank angles with fuel mass burned fraction at 10%, 50%, 90% and mode transition was performed. The results indicate that overall, the adopted LES simulations could effectively predict the cyclic variations in the hybrid combustion observed in the experiments, while the RANS simulations failed to reproduce the cyclic characteristics at the chosen engine operating conditions. Based on the LES results, the correlation and visualization studies indicate that the cyclic variations in the local velocity around the spark plug lead to the variations in the early flame propagation, which in turn produce temperature fluctuations among the cycles and result in greater variations in the subsequent auto-ignition combustion events.

Analysis of Combustion and Cycle to Cycle Variations of an Ethanol (E100) Fueled Spark-Ignition Engine

Ethanol, produced from biomass, is a renewable fuel that can be an attractive alternative fuel to gasoline for the increasing concerns of energy security and the environment. In the present work, combustion characteristics of ethanol (E100) fueled spark-ignition engine are compared with base gasoline fuel, and consecutively evaluated the cycle-to-cycle variations (CCV) of the engine parameters using statistical analysis. These investigations were performed on a 250cc air-cooled four-stroke single-cylinder port-fuel injection spark-ignition fueled with ethanol and gasoline fuels. The engine was operated at 3500 rpm and 4000 rpm of engine speed while maintaining the same loads for ethanol and gasoline. It was found from the results that the peak in-cylinder pressure (Ppeak) is lower with ethanol than gasoline at a given load. The Ppeak decreased from 27.5 to 25.5 bar when ethanol was used in the engine at 9.8:1 compression ratio, 3500 rpm, and 5 Nm load. Moreover, the mass fraction burned of the air-fuel mixture was faster for ethanol as compared to gasoline. The CCV with Ppeak, location of Ppeak, engine speed, and indicated mean effective pressure (IMEP) were evaluated. The result showed that CCV of Ppeak and its location were lowered for ethanol related to gasoline. A decrement in the coefficient of variation (COV) of Ppeak and its location from 11.6% to 8.3% and 12.6% to 11.7%, respectively, were observed when ethanol was utilized in the engine instead of gasoline at 3500 rpm and 5 Nm load. This resulted in a decrease in the COV of IMEP and engine speed by 2.8% and 0.25%, respectively, for ethanol as compared to gasoline at 3500 rpm and 5 Nm. This study results indicate the combustion stability of spark ignition engine with ethanol (E100) is better than gasoline fuel.

Modeling Cycle-to-Cycle Variations of a Spark-Ignited Gas Engine Using Artificial Flow Fields Generated by a Variational Autoencoder

A deeper understanding of the physical nature of cycle-to-cycle variations (CCV) in internal combustion engines (ICE) as well as reliable simulation strategies to predict these CCV are indispensable for the development of modern highly efficient combustion engines. Since the combustion process in ICE strongly depends on the turbulent flow field in the cylinder and, for spark-ignited engines, especially around the spark plug, the prediction of CCV using computational fluid dynamics (CFD) is limited to the modeling of turbulent flows. One possible way to determine CCV is by applying large eddy simulation (LES), whose potential in this field has already been shown despite its drawback of requiring considerable computational time and resources. This paper presents a novel strategy based on unsteady Reynolds-averaged Navier–Stokes (uRANS) CFD in combination with variational autoencoders (VAEs). A VAE is trained with flow field data from presimulated cycles at a specific crank angle. Then, the VAE can be used to generate artificial flow fields that serve to initialize new CFD simulations of the combustion process. With this novel approach, a high number of individual cycles can be simulated in a fraction of the time that LES needs for the same amount of cycles. Since the VAE is trained on data from presimulated cycles, the physical information of the cycles is transferred to the generated artificial cycles.

READ-Optimized 28nm HKMG Multibit FeFET Synapses for Inference-Engine Applications

This paper reports 2bits/cell ferroelectric FET (FeFET) devices with 500 ns write pulse of maximum amplitude 4.5V for inference-engine applications. FeFET devices were fabricated using GlobalFoundries 28nm high-k-metal-gate (HKMG) process flow on a 300mm wafer. The devices were characterized, and statistical modeling of variations in the fabricated devices was carried out based on experimental data. Furthermore, the model was applied to multi-layer perceptron (MLP) neural network (NN) simulations using the CIMulator software platform. The neural network (NN) was trained offline, and the weights were transferred to the synaptic devices for an inference-only operation. Device-to-device (D2D) and cycle-tocycle (C2C) variations are limited by optimal process conditions and do not impact inference accuracy. However, due to short-term retention, read-to-read (R2R) variation significantly affects inference operation. This work proposes a synergistic READoptimization approach to mitigate the impact of short-term retention and device variation issues. The optimization technique fostered immunity in the MLP-NN towards R2R variations, and the MLP-NN maintains inference accuracy of 97.01%, while the software baseline is 98%.

The correlation between intake fluctuation and combustion CCV (cycle-to-cycle variations) on a high speed gasoline engine: A wide range operating condition study

In this work, the cycle-to-cycle variations (CCV) and their correlations to intake charge fluctuation of a motorcycle engine under wide operating conditions were investigated by experiment. Several key engine operating parameters (engine speed/load, air/fuel ratio, compression ratio and ignition mode) were considered and analyzed thoroughly. A new concept of intake charge time-pressure value (TPV) was proposed to estimate its charge capacity, and its fluctuation’s contribution to CCV was also assessed. Some important findings include: 1) the CCV goes up with engine speed. Engine speed is also the most important factor to affect intake charge fluctuation, coefficient of variation (COV) of TPV is only 3.38% at 3000 rpm under 60% load, but it rises up to 13.41% at 5000 rpm; 2) the contribution of charge cyclic fluctuation to CCV decreases with engine speed. At 3000 rpm, over 80% of CCV are caused by TPV’s fluctuation for all engine loads, but its contribution drops to about 30% at 5000 rpm under medium to full load; 3) engine load has moderate impact on intake charge fluctuation’s contribution to engine CCV within test range. This depends on the competition between intake pressure loss and trapped ability of residual gas; 4) the three factors of air/fuel ratio, compression ratio and ignition mode do not affect much the air exchange process, but largely influence the combustion process. The more they improve combustion process, the larger degree of participation in engine CCV by intake charge fluctuation.

A New Method to Determine the Impact of Individual Field Quantities on Cycle-to-Cycle Variations in a Spark-Ignited Gas Engine

Cycle-to-cycle variations (CCV) in spark-ignited (SI) engines impose performance limitations and in the extreme limit can lead to very strong, potentially damaging cycles. Thus, CCV force sub-optimal engine operating conditions. A deeper understanding of CCV is key to enabling control strategies, improving engine design and reducing the negative impact of CCV on engine operation. This paper presents a new simulation strategy which allows investigation of the impact of individual physical quantities (e.g., flow field or turbulence quantities) on CCV separately. As a first step, multi-cycle unsteady Reynolds-averaged Navier–Stokes (uRANS) computational fluid dynamics (CFD) simulations of a spark-ignited natural gas engine are performed. For each cycle, simulation results just prior to each spark timing are taken. Next, simulation results from different cycles are combined: one quantity, e.g., the flow field, is extracted from a snapshot of one given cycle, and all other quantities are taken from a snapshot from a different cycle. Such a combination yields a new snapshot. With the combined snapshot, the simulation is continued until the end of combustion. The results obtained with combined snapshots show that the velocity field seems to have the highest impact on CCV. Turbulence intensity, quantified by the turbulent kinetic energy and turbulent kinetic energy dissipation rate, has a similar value for all snapshots. Thus, their impact on CCV is small compared to the flow field. This novel methodology is very flexible and allows investigation of the sources of CCV which have been difficult to investigate in the past.

Manifold reduction techniques for the comparison of crank angle-resolved particle image velocimetry (PIV) data and Reynolds-averaged Navier-Stokes (RANS) simulations in a spark ignition direct injection (SIDI) engine

In this article, different manifold reduction techniques are implemented for the post-processing of Particle Image Velocimetry (PIV) images from a Spark Ignition Direct Injection (SIDI) engine. The methods are proposed to help make a more objective comparison between Reynolds-averaged Navier-Stokes (RANS) simulations and PIV experiments when Cycle-to-Cycle Variations (CCV) are present in the flow field. The two different methods used here are based on Singular Value Decomposition (SVD) principles where Proper Orthogonal Decomposition (POD) and Kernel Principal Component Analysis (KPCA) are used for representing linear and non-linear manifold reduction techniques. To the authors’ best knowledge, this is the first time a non-linear manifold reduction technique, such as KPCA, has ever been used in the study of in-cylinder flow fields. Both qualitative and quantitative studies are given to show the capability of each method in validating the simulation and incorporating CCV for each engine cycle. Traditional Relevance Index (RI) and two other previously developed novel indexes: the Weighted Relevance Index (WRI) and the Weighted Magnitude Index (WMI), are used for the quantitative study. The results indicate that both POD and KPCA show improvements in capturing the main flow field features compared to ensemble-averaged PIV experimental data and single cycle experimental flow fields while capturing CCV. Both methods present similar quantitative accuracy when using the three indexes. However, challenges were highlighted in the POD method for the selection of the number of POD modes needed for a representative reconstruction. When the flow field region presents a Gaussian distribution, the KPCA method is seen to provide a more objective numerical process as the reconstructed flow field will see convergence with an increasing number of modes due to its usage of Gaussian properties. No additional criterion is needed to determine how to reconstruct the main flow field feature. Using KPCA can, therefore, reduce the amount of analysis needed in the process of extracting the main flow field while incorporating CCV.

Nonlinear dynamics of cycle-to-cycle variations in a lean-burn natural gas engine with a non-uniform pre-mixture

The cycle-to-cycle variations (CCVs) in reciprocating internal combustion engines may cause negative influence on diving performance, fuel economy, and emissions. Especially, lean-burn technology or exhaust gas recirculation (EGR) was used to improve engine combustion efficiency and reduce NOx, and the combustion boundary was limited by increased CCVs. Therefore, it was important to identify the complex dynamics of CCVs and to take measures for inhibiting them. The CCVs based on indicated mean effective pressure (IMEP) time series were examined in a lean-burn natural gas engine with a non-uniform pre-mixture based on statistical and multifractal theories. Tests were conducted at an engine speed of 1000 rpm and low loads of 10% and 25%, and combustion data at an engine load of 10% were analysed in detail because the CCVs are less sensitive to changes of gas injection timing (GIT) at the higher engine load. The nonlinear dynamics of the CCVs was revealed at the different GIT from 1° to 120°CA ATDC. The statistical properties of IMEP time series were characterised by distributions of probability density functions (PDF), the multifractal complexities of the combustion fluctuations were quantitatively analysed by the singularity spectra in terms of the Holder exponent based on the theory of wavelet transform modulus maxima, and the primary source leading to the increased CCVs and complex dynamics in a natural gas engine with a non-uniform pre-mixture was identified using 3D-computational fluid dynamics simulations. Results show that as the GIT increased, the kurtosis of the IMEP time series systematically decreased from 592 to 1.8, the fast dynamics transitions from super-Gaussian to quasi-Gaussian distributions in combustion system were revealed, and the lower value of kurtosis implied the lower degree of intermittency. Except for the GIT of 60°CA ATDC, the value of the Holder exponent $$h_{0} > 0.5$$ , which implies that the CCVs for the other GITs behaved like a persistent walk or positive correlation. For GITs of 60° and 90°CA ATDC, the narrow broadness of the singularity spectrum implicated a monofractality, while the obvious multifractal properties could be identified for the other GITs. The transitions of the dynamic behaviours may be caused by the degree of mixture in-homogeneity combined with a new mechanism of “prior-cycle effects” proposed in this paper, and the effect mechanism was not only associated with the residual gas in the cylinder but also with the residual gas fuel in the intake port, rather than independent effect of residual gas in cylinder reported in previous researches. Our research results provided the deeper understanding on the dynamics of combustion system in multi-point injection natural gas engines and may be beneficial to achieve nonlinear prediction and to develop improved control strategies for inhibiting the CCVs.

Experimental study the impacts of the key operating and design parameters on the cycle-to-cycle variations of the natural gas SI engine

High cycle-to-cycle variations (CCV) in the spark ignition (SI) engine will not only lead to incomplete combustion and even misfire, but also affect the fuel economy and even influence the drivability and conformability of a vehicle, particularly at the lean-burn condition. However, the detail information about the root causes of CCV are still relatively limited, particularly the natural gas SI engine. In this paper, the impacts of the key operating and design parameters were investigated on the CCV of the high compression ratio SI natural gas engine under lean mixture condition. The results indicated that the in-cylinder pressure distribution tended to be concentrated, the number of partial burning or incomplete combustion cycles also reduced with increasing the engine speed. With increasing the load, the in-cylinder pressure distribution, peak combustion pressure and its location were tended to be concentrated. Furthermore, the number of the partial combustion or post combustion cycles declined from low-load to medium- and high-load zones. Moreover, the relationship between the peak combustion pressure and its location was almost linear with increasing the engine load. Unfortunately, with further increasing engine load, in-cylinder pressure traces started to disperse slightly. Under the low-load condition, in-cylinder pressure traces were tended to be concentrated with increasing compression ratio (CR). As CR continually increased, its influence on the in-cylinder pressure distribution decreased. In addition, the number of the partial combustion or post combustion cycles decreased with increasing CR. However, under the high-load condition, increasing the CR has slightly effects on the in-cylinder pressure distribution.

Analysis of cycle-to-cycle variations in a common-rail compression ignition engine fuelled with diesel and biodiesel fuels

Using wavelet and fractal theories, cycle-to-cycle variations (CCVs) in a common-rail compression ignition (CI) engine have been investigated at engine loads of 25% and 50%, biodiesel blend level was from 0% to 100%. Wavelet power spectrums and singularity spectra were calculated to identify the dominant oscillatory combustion modes and multifractal complexity. Reaction paths and component consumption sensitivity of n-heptane and methyl decanoate were studied to reveal the effect of biodiesel blend level on the combustion process of diesel fuel. Results reveal that the effect of biodiesel blend level on the CCVs is more significant at a low load, even when biodiesel blend level increases to 20%, the coefficients of variation decreases from 3.99% to 1.57%. The CCVs exhibit multiscale dynamics for all tested cases, and persistent high-frequency oscillations appear around a 16-cycle period persisting over the entire or several hundred of the engine cycles. As the biodiesel blend level increases, the periodic bands with the highest power were interrupted and combined with lower-frequency and high-frequency intermittent fluctuations. However, for the higher load, the dynamics of CCVs are mainly displayed in an intermittent fashion. The larger broadness of singularity spectra at higher engine loads suggests a higher degree of multifractality. For all of the tested cases, the dynamics of the CCVs behave like antipersistent walks. As a oxygenated fuel, biofuel substitution leads to increase of c7h15-1 concentration and radicals such as OH, O and H2O2, which are beneficial to decrease ignition delay and accelerate the chemical reaction rate of diesel fuel, and therefore inhibit the CCVs.

Laminar-to-turbulent flame transition and cycle-to-cycle variations in large eddy simulation of spark-ignition engines

This paper investigates the effect of laminar-to-turbulent flame transition modeling on the prediction of cycle-to-cycle variations (CCVs) in large eddy simulation (LES) of spark-ignition (SI) engines. A laminar-to-turbulent flame transition model that describes the non-equilibrium sub-filter flame speed evolution during an early stage of flame kernel growth is developed. In the present model, the flame transition is characterized by the flame kernel size at which the flame transition ends, defined here as the flame transition scale. The proposed model captures the effects that variations in a turbulent flow field have on the evolution of early-stage burning rates, through variations in the flame transition scale. The proposed flame transition model is combined with the front propagation formulation (FPF) method and a spark-ignition model to predict CCVs in a gasoline direct injection SI engine. It is found that multi-cycle LES with the proposed flame transition model reproduces experimentally-observed CCVs satisfactorily. When the transition model is not considered or when variations in the transition process are neglected, CCVs are significantly under-predicted for the case considered here. These results indicate the importance of modeling the laminar-to-turbulent flame transition and the effect of turbulence on the transition process, when predicting CCVs, under certain engine conditions. The LES results are also used to analyze sources for variations in the flame transition. It is found, for the present engine case, that the most important source is the cycle-to-cycle variation in the turbulence dissipation rate, which is used to measure the strength of turbulence in the proposed model, near a spark plug. The large-scale velocity field and the variations of the laminar flame speed due to the mixture composition and thermal stratification are also found to be important factors to contribute to the variations in the flame transition.

Analysis of Cyclic Variation Using Time-Resolved Tomographic Particle-Image Velocimetry

<div class="section abstract"><div class="htmlview paragraph">To achieve the strict legislative restrictions for emissions from combustion engines, vast improvements in engine emissions and efficiency are required. Two major impacting factors for emissions and efficiency are the reliable generation of an effective mixture before ignition and a fast, stable combustion process. While the mixture of air and injected fuel is generated by highly three-dimensional, time-dependent flow phenomena during the intake and compression stroke, the turbulent flame propagation is directly affected by the turbulence level in the flow close to the advancing flame front. However, the flow field in the combustion chamber is highly turbulent and subject to cycle-to-cycle variations (CCV). To understand the fundamental mechanisms and interactions, 3D flow measurements with combined high spatial and temporal resolution are required. Furthermore, most studies utilize velocity decompositions based on the standard Reynolds decomposition that yields a mean flow velocity and a fluctuation velocity, which comprises of fluctuations from CCV and turbulence. Therefore, time-resolved tomographic PIV measurements are coupled with a triple-velocity decomposition method based on a multiple-snapshot POD to investigate in-cylinder velocity fluctuations from cyclic variability. The method is extensively discussed and the most important steps are highlighted by examples. The results illustrate that the method yields valid findings, which allow a comparison of the impact of CCV between two parameter sets or between engine cycles.</div></div>

Investigation of cycle-to-cycle variations in a spark-ignition engine based on a machine learning analysis of the early flame kernel

High-speed Mie scatter imaging and in-cylinder pressure measurements are performed in an optically accessible four-stroke spark-ignition engine to investigate cycle-to-cycle variations (CCVs). Droplet Mie scattering is used to measure the cross-sectional flame contour. Machine learning (ML) methods are applied to predict combustion cycles of high maximum in-cylinder pressure based on flame cross-sections at -15 °CA (crank angle degrees before top dead center). All tested ML methods (decision tree-based, multi-layer perceptron, and logistic regression) are able to predict high energy engine cycles given only partial flame topology information at a temporal snapshot of the flame propagation. Feature importance analyses, employed using decision tree methods, reveal that a combination of flame position and shape features of the flames' cross-section is necessary in achieving a high prediction accuracy. These sensitivity analyses can be used to gain insight into the combustion processes and to strategically reduce the number of features. This enables the addition of new physical quantities in future investigations.

Development and Application of Bivariate 2D-EMD for the Analysis of Instantaneous Flow Structures and Cycle-to-Cycle Variations of In-cylinder Flow

The bivariate two dimensional empirical mode decomposition (Bivariate 2D-EMD) is extended to estimate the turbulent fluctuations and to identify cycle-to-cycle variations (CCV) of in-cylinder flow. The Bivariate 2D-EMD is an adaptive approach that is not restricted by statistical convergence criterion, hence it can be used for analyzing the nonlinear and non-stationary phenomena. The methodology is applied to a high-speed PIV dataset that measures the velocity field within the tumble symmetry plane of an optically accessible engine. The instantaneous velocity field is decomposed into a finite number of 2D spatial modes. Based on energy considerations, the in-cylinder flow large-scale organized motion is separated from turbulent fluctuations. This study is focused on the second half of the compression stroke. For most of the cycles, the maximum of turbulent fluctuations is located between 50 and 30 crank angle degrees before top dead center (TDC). In regards to the phase-averaged velocity field, the contribution of CCV to the fluctuating kinetic energy is approximately 55% near TDC.