Crank Angle Domain Research Articles

Evaluation of control-oriented flame propagation models for production control of a spark-assisted compression ignition engine

The drive to improve internal combustion engines has led to efficiency objectives that exceed the capability of conventional combustion strategies. As a result, advanced combustion modes are more attractive for production. These advanced combustion strategies typically add sensors, actuators, and degrees of freedom to the combustion process. Spark-assisted compression ignition (SACI) is an efficient production-viable advanced combustion strategy characterized by spark-ignited flame propagation that triggers autoignition in the remaining unburned gas. Modeling this complex combustion process for control demands a careful selection of model structure to maximize predictive accuracy within computational constraints. This work comprehensively evaluates a physics-based and a data-driven model. The physics-based model produces a burn duration by computing laminar flame speed as a function of test point conditions. The crank-angle domain is intentionally excluded to reduce computational expense. The data-driven model is an artificial neural network (ANN). The candidate models are compared to a one-dimensional engine model validated to experimental SACI engine data. Though both models capture the trends in burn rates, the ANN model has a root-mean square error (RMSE) of 1.4 CAD, significantly lower than the 10.4 CAD RMSE of the physics-based model. The exclusion of the crank-angle domain results in insufficient detail for the physics-based model, while the ANN can tolerate this exclusion.

In-cylinder pressure reconstruction by engine acoustic emission

The importance of the in-cylinder pressure transducer has been proven in revealing the information about combustion and exhaust pollution formation, as well as for its capability to classify knock. Due to their high price, they are not used commercially for engine health monitoring, which is of significant importance. Hence, this study will investigate the reconstruction of the in-cylinder pressure trace using a structure-borne acoustic emission (AE) sensor, which are relatively low cost sensors. As shown in the literature, AE indicators show a strong correlation with in-cylinder pressure parameters in both time and crank angle domain. In this study, to avoid the effect of engine speed fluctuations, the reconstruction is done in the crank angle domain by means of the Hilbert transform of AE. Complex cepstrum signal processing analysis with a feed-forward neural network is used to generate a reconstruction regime. Furthermore, the reconstructed signals are used to determine some of the important in-cylinder parameters such as peak pressure (PP), peak pressure timing (PPT), indicated mean effective pressure (IMEP) and pressure rise rate. Results showed that the combination of cepstrum analysis with neural network is capable of reconstructing pressure using AE, regardless of engine load, speed and fuel type. The reconstructed pressure can be used to reliably determine PP and PPT. IMEP can be estimated as well in a reasonable range.

A crank-kinematics-based engine cylinder pressure reconstruction model

A new inverse model is proposed for reconstructing steady-state and transient engine cylinder pressure using measured crank kinematics. An adaptive nonlinear time-dependent relationship is assumed between windowed-subsections of cylinder pressure and measured crank kinematics in a time-domain format (rather than in crank-angle domain). This relationship comprises a linear sum of four separate nonlinear functions of crank jerk, acceleration, velocity and crank angle. Each of these four nonlinear functions is obtained at each time instant by fitting separate m-term Chebyshev polynomial expansions, where the total 4 m instantaneous expansion coefficients are found using a standard (overdetermined) linear least-square solution method. A convergence check on the calibration accuracy shows that this initially improves as more Chebyshev polynomial terms are used, but with further increase, the overdetermined system becomes singular. Optimal accuracy Chebyshev expansions are found to be of degree m = 4, using 90 or more cycles of engine data to fit the model. To confirm the model accuracy in predictive mode, a defined measure is used, namely the ‘ calibration peak pressure error’. This measure allows effective a priori exclusion of occasionally unacceptable predictions. The method is tested using varying speed data taken from a three-cylinder direct-injection spark ignition engine fitted with cylinder pressure sensors and a high-resolution shaft encoder. Using appropriately filtered crank kinematics (plus the ‘calibration peak pressure error’), the model produces fast and accurate predictions for previously unseen data. Peak pressure predictions are consistently within 6.5% of target, whereas locations of peak pressure are consistently within ±2.7 °CA. The computational efficiency makes it very suitable for real-time implementation.

Acoustic emission characteristics of a single cylinder diesel generator at various loads and with a failing injector

Acoustic emission sensing techniques have been applied in recent years to dynamic machinery with varying degrees of success in diagnosing various component faults and distinguishing between operating conditions. This work explores basic properties of acoustic emission signals measured on a small single cylinder diesel engine in a laboratory setting. As reported in other works in the open literature, the measured acoustic emission on the engine is mostly continuous mode and individual burst events are generally not readily identifiable. Therefore, the AE are processed into the local (instantaneous) root mean square (rms) value of the signal which is averaged over many cycles to obtain a mean rms AE in the crank angle domain. Crank-resolved spectral representation of the AE is also given but rigorous investigation of the AE spectral qualities is left to future study. Cycle-to-cycle statistical dispersion of the AE signal is considered to highlight highly variable engine processes. Engine speed was held constant but load conditions are varied to investigate AE signal sensitivity to operating condition. Furthermore, during the course of testing the fuel injector developed a fault and acoustic emission signals were captured and several signal attributes were successful in distinguishing this altered condition. The sampling and use of instantaneous rms acoustic emission signal demonstrated promise for non-intrusive and economical change detection of engine injection, combustion and valve events.

Automotive Internal-Combustion-Engine Fault Detection and Classification Using Artificial Neural Network Techniques

In this paper, an engine fault detection and classification technique using vibration data in the crank angle domain is presented. These data are used in conjunction with artificial neural networks (ANNs), which are applied to detect faults in a four-stroke gasoline engine built for experimentation. A comparative study is provided between the popular backpropagation (BP) method, the Levenberg–Marquardt (LM) method, the quasi-Newton (QN) method, the extended Kalman filter (EKF), and the smooth variable structure filter (SVSF). The SVSF is a relatively new estimation strategy, based on the sliding mode concept. It has been formulated to efficiently train ANNs and is consequently referred to as the SVSF-ANN. The accuracy of the proposed method is compared with the standard accuracy of the Kalman-based filters and the popular BP algorithms in an effort to validate the SVSF-ANN performance and application to engine fault detection and classification. The customizable fault diagnostic system is able to detect known engine faults with various degrees of severity, such as defective lash adjuster, piston chirp (PC), and chain tensioner (CT) problems. The technique can be used at any dealership or assembly plant to considerably reduce warranty costs for the company and manufacturer.

Fault diagnosis of reciprocating compressor valve with the method integrating acoustic emission signal and simulated valve motion

This paper proposes a method of diagnosing faults in reciprocating compressor valves using the acoustic emission signal coupled with the simulated valve motion. The actual working condition of a valve can be obtained by analyzing the acoustic emission signal in the crank angle domain and the valve movement can be predicted by simulating the valve motion. The exact opening and closing locations of a normal valve, provided by the simulated valve motion, can be used as references for the valve fault diagnosis. The typical valve faults are diagnosed to validate the feasibility and accuracy of the proposed method. The experimental results indicate that this method can easily distinguish the normal valve, valve flutter and valve delayed closing conditions. The characteristic locations of the opening and closing of the suction and discharge valves can be clearly identified in the waveform of the acoustic emission signal and the simulated valve motion.

Burn angles and form factors for Wiebe function fits to mass fraction burned curves of a spark ignition engine with variable valve timing

The charge burn characteristics of a port-injected spark ignition engine with a pent-roof combustion chamber and variable valve timing have been investigated experimentally. The engine was run under stoichiometric mixture operating conditions over ranges of intake and exhaust valve timings, engine speed, and engine load. Empirical functions have been developed for the 0–90 per cent mass fraction burned angle and the form factor, which define Wiebe function fits to the mass fraction burned variation in the crank angle domain. The burn angle and form factor have been related to the level of charge dilution by burned gas, engine speed, ignition timing, and charge density at spark timing. The dilution level has the strongest influence on the burn rate and profile. The dilution level varied with intake and exhaust valve timings, external exhaust gas recirculation, and engine load. The results indicate that intake and exhaust valve timings influence combustion primarily by modifying the charge dilution with burned gas and charge density.

Torque balancing control in ignition event based scale for multi-cylinder SI engines

In this paper, the torque balancing problem for multi-cylinder SI engines is focused on. The deviation in average torque generation of sections with equal crank angle interval according to ignition event is introduced as index for balancing and an estimation method for this index from engine speed measurement is proposed based on crank shaft dynamics in crank angle domain. According to the coupling characteristic in torque generation between cylinders, a simple 2nd order ARMA model is established to describe the dynamics from spark ignition to the index. Based on this model, model predictive control technique is used in regulation. Finally, the validity of the proposed approach is demonstrated by simulation performed on a six-cylinder engine simulator with physical background.

Real-time brake torque estimation for internal combustion engines

Presented is a real-time engine brake torque estimation model where the instantaneous measured engine speed serves as the model input. The model is comprised of notch filters executed in the crank-angle domain to extract mean engine speed and the nth frequency component from the instantaneous engine speed signal in real-time. Here n denotes the number of engine cylinders. Moreover, the engine brake torque estimation is separated into two parts: steady-state and transient. It will be shown that the nth harmonic (in units of periods per engine cycle) of engine speed and mean engine speed are sufficient to estimate the engine brake torque. The steady-state portion of the model is developed using orthogonal least-squares estimation and results in a model with 15 regressors for our particular case. The transient portion is identified using a time domain identification method. Validation of the engine brake torque model is provided using a computational engine model for a 6-cylinder heavy duty diesel engine. Transient engine speed and torque conditions in the presence of sensor noise are evaluated as well as cylinder power imbalance scenarios.

Diesel Engine Indicated and Load Torque Estimation Using a Non-Linear Observer

This paper presents a non-linear observer to estimate the indicated torque and the load torque of a single-cylinder diesel engine from crankshaft and coupling angular velocity measurements. Since these variables can be measured using low-cost sensors, the observer may be useful in the implementation of the control or diagnostics strategies that require cylinder indicated torque and load torque, variables that are not easily measured and need expensive sensors. The observer operates in the crank angle domain and is based on a nonlinear dynamic engine model that includes an instantaneous friction model. The effects of inertia variations of the crankshaft assembly with piston pin offset are also included, which significantly increases the accuracy of the pressure estimation at high speeds. Strong chattering is avoided by modifying the sliding gain as a non-linear function of the crank angle. The stability of the observer is proved, and simulation results of the engine indicated torque and load torque are presented, which indicate good agreement with experimental data.

Reducing Engine Idle Speed Deviations Using the Internal Model Principle

Presented in this paper is a multivariable linear feedback controller design methodology for idle speed control of spark-ignition engines. The engine is modeled as a multi-input, single-output system. The proposed feedback control system employs both throttle and ignition timing to control engine speed and engine roughness. Throttle is used to attenuate low frequency components of the speed error and reject mean speed errors. Spark advance is used to reduce cylinder-to-cylinder differences in torque production by limiting high frequency speed deviations. The algorithm is executed in the crank-angle domain, and the internal model principle serves as the basis for cylinder torque balancing. The nonlinear relationship between ignition timing and torque production is explicitly incorporated into the design process using a sector bound. A loop shaping approach is proposed to design the feedback controller, and absolute stability of the nonlinear closed-loop system is guaranteed through the Tsypkin Criterion. Experimental results from implementation on a Ford 4.6L V-8 engine are provided.

Residual generation and statistical pattern recognition for engine misfire diagnostics

Methods for diagnosing misfire in internal combustion engines are presented in this paper. Crank-angle domain digital filters are used to extract features from the measured engine speed signal that are characteristic of misfire. Features for intermittent and continuous misfires are developed separately, since the engine speed responses for intermittent and continuous misfires are distinctly different. Also, the influence of crankshaft torsional vibration and repeatable measurement errors must be addressed differently in each case. The outputs from the digital filters serve as inputs to a pattern recognition network based on linear parametric classifiers. Experimental results from implementation on a Ford 4.6L V-8 engine are provided.

Indirect measurement of cylinder pressure from diesel engines using acoustic emission

Indirect measurement of the cylinder pressure from diesel engines is possible using acoustic emission (AE). A method is demonstrated for a large two-stroke marine diesel engine and a small four-stroke diesel engine, which involves reconstructing the cylinder crank angle domain diagram from the AE generated during the combustion phase. Raw AE was used for modelling and reconstructing the pressure waveform in the time domain but this could not be used to model the pressure rise (compression). To overcome this problem the signal was divided into two sections representing the compression part of the signal and the fuel injection/expansion stroke. The compression part of the pressure signal was reconstructed by using polynomial fitting. An auto-regressive technique was used during the injection/expansion stroke. The rms AE signal is well correlated with the pressure signal in the time and frequency domain and complex cepstrum analysis was used to model the pressure signal for the complete combustion phase (compression, injection and expansion). The main advantage of using cepstral analysis is that the model uses the frequency content of the rms AE signal rather than the energy content of the rms AE signal, which gives an advantage when the signal has lower energy content, during the compression process. By calculating the engine running speed from the rms AE signal and selecting the proper cepstrum model correlated to the combustion rms AE energy content, an analytical algorithm was developed to give a wide range of applicability over the different conditions of engine speed, engine type and load. The pressure reconstructed from both AE and acceleration data are compared. AE has the advantage of a much higher signal-to-noise ratio and improved time resolution and is shown to be better than the acceleration.

An investigation of crankshaft oscillations for cylinder health diagnostics

The vibrational characteristics of an internal combustion engine crankshaft are investigated from a cylinder health diagnostics point of view. Experimental results from a six-cylinder industrial diesel engine are presented to demonstrate the effects of cylinder imbalance on the individual harmonic components of the engine speed signal. A crank-angle domain numerical model of the crankshaft dynamics for a six-cylinder industrial diesel engine is also adopted to establish the effects of continuous low-power production in individual cylinders of a multi-cylinder engine. Outline of a diagnostics algorithm that makes use of the properties of crankshaft vibration behaviour is provided. In particular, crank-angle domain notch filters are employed to extact the harmonic components of engine speed. The outlined method can be implemented for individual cylinder health diagnostics across a family of multi-cylinder engines and can be formulated to handle changes in crankshaft characteristics due to replacement of mechanical components and/or wear.

Detailed analytical model of a single-cylinder diesel engine in the crank angle domain

A detailed analytical non-linear dynamic model for single-cylinder diesel engines is developed. The model describes the dynamic behaviour between fuelling and engine speed and includes models of the non-linear engine and dynamometer dynamics, the instantaneous friction terms and the engine thermodynamics. The model operates in the crank angle domain. The dynamometer model enables the study of the engine behaviour under loading. The instantaneous friction model takes into consideration the viscosity variations with temperature. Inertia variations with piston pin offset are presented. In-cycle calculations are performed at each crank angle, and the correct crank angles of ignition, speed variations, fuel supply and air as well as fuel burning rate are predicted. The model treats the cylinder strokes and the manifolds as thermodynamic control volumes by using the filling and emptying method. The model is validated using experimentally measured cylinder pressure and engine instantaneous speeds, under transient operating conditions, and gives good agreement. The model can be used as an engine simulator to aid diesel engines control system design and fault diagnostics.

An adaptive sliding mode controller for air-fuel ratio control of spark ignition engines

An adaptive dynamic sliding mode fuel injection control algorithm based on the measurement of a binary oxygen sensor to reduce the exhaust gas emissions is proposed. The controller suggested in this paper is designed on the basis of the two-state dynamic engine model developed in the crank angle domain, and it is composed of an adaptation law for fuel delivery model parameters and measurement bias in mass air flowrate. The control algorithm is mathematically compact enough to run in real time, and it is robust to modelling errors as well as to rapidly changing manoeuvres of the throttle. The simulation and experimental results show that this algorithm can substantially reduce the transient peaks in air-fuel ratio (AFR) while maintaining robustness to model errors and measurement delay.

Direct Fuzzy Control of Idle Speed in an Internal Combustion Engine

Idle speed control is currently an important problem in production vehicles. Increasing government regulations for vehicle efficiency motivates improvements to current systems. This paper explores the idle speed control problem with a goal of lowering idle speed, rejecting torque disturbances and reducing steady state variations in engine speed at idle. The distinguishing factor in this experimental study is that modeling and control design are carried out in the crank angle domain (as opposed to the time domain).Control schemes are designed and implemented on a Ford V8 engine. The approaches compared are РID, direct fuzzy control and fuzzy model reference learning control (FMRLC). The variable being controlled is the mass air flow through the idle bypass valve. The effect of reducing sampling rate is investigated for the PID and the direct fuzzy designs.

Real Time Estimation of Engine Torque for the Detection of Engine Misfires

The need for improvements in the on-line estimation of engine performance variables is greater nowadays as a result of more stringent emission control legislation. There is also a concurrent requirement for improved on-board diagnostics to detect different types of malfunctions. For example, recent California Air Resources Board (CARB) regulations mandate continuous monitoring of misfires, a problem which, short of an expensive measurement of combustion pressure in each cylinder, is most directly approached by estimating individual cylinder torque. This paper describes the theory and experimental results of a method for the estimation of individual cylinder torque in automative engines, with the intent of satisfying the CARB misfire detection requirements. Estimation, control, and diagnostic functions associated with automotive engines involve near periodic processes, due to the nature of multi-cylinder engines. The model of the engine dynamics used in this study fully exploits the inherent periodicity of the combustion process in the crank angle domain in order to obtain a simple deconvolution method for the estimation of the mean torque produced by each cylinder during each stroke from a measurement of crankshaft angular velocity. The deconvolution is actually performed in the spatial frequency domain, recognizing that the combustion energy is concentrated at discrete spatial frequencies, which are harmonics of the frequency of rotation of the crankshaft. Thus, the resulting deconvolution algorithm is independent of engine speed, and reduces to an algebraic operation in the frequency domain. It is necessary to perform a Discrete Fourier Transform (DFT) on the measured angular velocity signal, sampled at fixed uniform crank angle intervals. The paper discusses the model used in the study, and the experimental validation of the algorithm, which has been implemented in real time using a portable computer and has been tested extensively on different production vehicles on a chassis dynamometer and on the road.