Search Results

Some of the best teaching methods are laboratory courses in which students experience application of the principles being presented. Preparing young engineering students for a career in the automotive industry challenges us to provide comparable opportunities to explore the dynamic performance of motor vehicles in a controlled environment. Today we are fortunate to have accurate and easy-to-use software programs making it practical for students to simulate the performance of motor vehicles on “virtual” proving grounds. At the University of Michigan the CarSim® vehicle dynamics simulation program has been introduced as such a tool to augment the learning experience. The software is used in the Automotive Engineering course to supplement homework exercises analyzing acceleration, braking, aerodynamics, and cornering performance. This paper provides an overview of the use of simulation in this setting.

An accurate air flow rate model is critical for high-quality air-fuel ratio control in Spark-Ignition engines using a Three-Way-Catalyst. Emerging Variable Valve Timing technology complicates cylinder air charge estimation by increasing the number of independent variables. In our previous study (SAE 2004-01-3054), an Artificial Neural Network (ANN) has been used successfully to represent the air flow rate as a function of four independent variables: intake camshaft position, exhaust camshaft position, engine speed and intake manifold pressure. However, in more general terms the air flow rate also depends on ambient temperature and pressure, the latter being largely a function of altitude. With arbitrary cam phasing combinations, the ambient pressure effects in particular can be very complex. In this study, we propose using a separate neural network to compensate the effects of altitude on the air flow rate.

Sound attenuation and flow loss reduction are often two competing demands in vehicle breathing systems. The present study considers a full vehicle exhaust system and investigates both the sound attenuation and the flow performance of production configurations including the catalyst, the resonator, and the muffler. Dynamometer experiments have been conducted with a firing Ford 3.0L, V-6 engine at wide-open throttle with speeds ranging from 1000 to 5000 rpm. Measurements including the flow rates, the temperatures and the absolute dynamic pressures of the hot exhaust gases at key locations (upstream and downstream of every component) with fast-response, water-cooled piezo-resistive pressure transducers facilitate the calculation of acoustic performance of each component, as well as the determination of flow losses caused by these elements and their influence on the engine performance.

When structures may have dynamic instability complex eigenvalue analysis is a useful tool to predict it. Although the accurate prediction itself is significant, it is also crucial to obtain sensitivity of unstable eigensolutions in order to eliminate instability efficiently. Since the mathematical relationship between stiffness matrix and design variables may seldom be found in reality, finite difference method has been typically used to approximate the sensitivity. The novel way to accurately calculate the sensitivity is developed without implementing finite difference method. This paper shows the advantages of analytical sensitivity analysis compared to other methods for choosing the most important components' eigenvalues. It also provides necessary amount of frequency shift for each chosen components' eigenvalue to eliminate unstable eigenvalues.

This paper presents the real time implementation of detection filters for the diagnosis of incipient failures in electronically controlled internal combustion (IC) engines. The detection filters are implemented in a production vehicle. Recent results [1] have demonstrated the feasibility of a model-based failure detection and isolation (FDI) methodology for detecting partially failed components in electronically controlled vehicle subsystems. The present paper describes the real time application of the FDI concept to the detection of faults in sensors associated with the engine/controller In a detection filter, the performance of the engine/controller system is continuously compared to a prediction based on sensor measurements and an analytical model (typically a control model) of the system. Any discrepancy between actual and predicted performance is analyzed to identify the unique failure signatures related to specific system components.

A quasi-dimensional, multi-zone, direct injection (DI) diesel combustion model has been developed and implemented in a full cycle simulation of a turbocharged engine. The combustion model accounts for transient fuel spray evolution, fuel-air mixing, ignition, combustion and NO and soot pollutant formation. In the model, the fuel spray is divided into a number of zones, which are treated as open systems. While mass and energy equations are solved for each zone, a simplified momentum conservation equation is used to calculate the amount of air entrained into each zone. Details of the DI spray, combustion model and its implementation into the cycle simulation of Assanis and Heywood [1] are described in this paper. The model is validated with experimental data obtained in a constant volume chamber and engines. First, predictions of spray penetration and spray angle are validated against measurements in a pressurized constant volume chamber.

This paper reports the development of a model of diesel combustion and NO emissions, based on a modified eddy dissipation concept (EDC), and its implementation into the KIVA-3V multidimensional simulation. The EDC model allows for more realistic representation of the thin sub-grid scale reaction zone as well as the small-scale molecular mixing processes. Realistic chemical kinetic mechanisms for n-heptane combustion and NOx formation processes are fully incorporated. A model based on the normalized fuel mass fraction is implemented to transition between ignition and combustion. The modeling approach has been validated by comparison with experimental data for a range of operating conditions. Predicted cylinder pressure and heat release rates agree well with measurements. The predictions for NO concentration show a consistent trend with experiments. Overall, the results demonstrate the improved capability of the model for predictions of the combustion process.

Analytical and experimental investigations of a diesel engine cylinder block are performed. An attempt is made to reduce modeling and analysis costs in the design process of an engine. Traditionally, the engine has been modeled using either 8-node or 20-node solid elements for stress and thermal analyses and modeled using 4-node plate and shell elements for the dynamic analysis. In this paper, a simpler finite element modeling technique using only 8 node solid elements for both dynamic and static analyses is presented. Based on this integrated modeling technique of finite elements, eigenvalues are calculated and compared with the experimental data obtained from modal testing of an actual engine cylinder block.

A combined experimental and analytical approach was followed in this work to study stress distributions and causes of failure in diesel cylinder heads under steady-state and transient operation. Experimental studies were conducted first to measure temperatures, heat fluxes and stresses under a series of steady-state operating conditions. Furthermore, by placing high temperature strain gages within the thermal penetration depth of the cylinder head, the effect of thermal shock loading under rapid transients was studied. A comparison of our steady-state and transient measurements suggests that the steady-state temperature gradients and the level of temperatures are the primary causes of thermal fatigue in cast-iron cylinder heads. Subsequently, a finite element analysis was conducted to predict the detailed steady-state temperature and stress distributions within the cylinder head. A comparison of the predicted steady-state temperatures and stresses compared well with our measurements.

Quantitative LIF measurements of liquid fuel films on the piston of direct-injected gasoline engines are difficult to achieve because generally these films are thin and the signal strength is low. Additionally, interference from scattered laser light or background signal can be substantial. The selection of a suitable fluorescence tracer and excitation wavelength plays an important role in the success of such measurements. We have investigated the possibility of using toluene as a tracer for fuel film measurements and compare it to the use of 3-pentanone. The fuel film dynamics in a motored engine at different engine speeds, temperatures and in-cylinder swirl levels is characterized and discussed.

A four-valve-pentroof, direct-injection, optical engine fueled with n-heptane has been operated at four different steady-state HCCI operating conditions including 10% and 65% residuals, both at low and high swirl conditions. Both, planar toluene LIF and volume chemiluminescence show large scale inhomogeneity in the ensemble averaged images. The interpretation of the toluene-tracer LIF signals (when premixed with the fresh-air charge) as a marker for reaction homogeneity is discussed. A binarization scheme and a statistical analysis of the LIF images were applied to the per-cycle planar-LIF images revealing inhomogeneities both from cycle-to-cycle and within the regions of individual cycles that track with the average heat release rate. Comparison of these two homogeneity metrics between the four operating conditions reveals weak but discernable differences.

Coordinate measurement gages dominate in the area of dimensional control and variation reduction of automotive body assembly processes. However, coordinate measurement gages do not have the capability to track certain measured features. This incapability introduces inherent measurement error created by part (feature) mislocation in constrained non-measured directions. This inherent measurement error weakens the methods used for process control and variation reduction. In this paper, a principle of measurement uncertainty is developed in order to estimate the measurement error caused by this deficiency. The developed principle describes measurement error, which is independent of any other error related to the mechanical or optical coordinate measurement machines (CMMs, OCMMs). Additionally, an error map determined by the measurement uncertainty principle is created for error compensation.

The interaction of fuel sprays and in-cylinder flow in direct-injection engines is expected to alter kinetic energy and integral length scales at least during some portions of the engine cycle. High-speed particle image velocimetry was implemented in an optical four-valve, pent-roof spark-ignition direct-injection single-cylinder engine to quantify this effect. Non-firing motored engine tests were performed at 1300 RPM with and without fuel injection. Two fuel injection timings were investigated: injection in early intake stroke represents quasi-homogenous engine condition; and injection in mid compression stroke mimics the stratified combustion strategy. Two-dimensional crank angle resolved velocity fields were measured to examine the kinetic energy and integral length scale through critical portions of the engine cycle. Reynolds decomposition was applied on the obtained engine flow fields to extract the fluctuations as an indicator for the turbulent flow.

For applications of planar laser induced fluorescence (PLIF) to measure the fuel or equivalence ratio distributions in internal combustion (IC) engines it is typically assumed that the addition of a fluorescence tracer to a base fuel does not alter the combustion performance. We have investigated the impact on combustion performance through the addition of various amounts of 3-pentanone or toluene to iso-octane fuel. Correlations between equivalence ratio for a range of fuel/tracer mixtures and engine parameters, like peak pressure, location of peak pressure, indicated mean effective pressure (IMEP), and peak burn rate are discussed for data obtained in a spark-ignition direct-injection (SIDI) gasoline engine operated with near homogeneous charge. For typical tracer concentrations the impact on combustion performance is mostly negligible.

The distribution of fuel-air mixtures in many L-head engines is not homogeneous. If local mixture is too rich or lean, incomplete combustion occurs. This can play a major role in unburned hydrocarbon and carbon monoxide emissions. Fuel-air mixture distribution depends on in-cylinder swirl and turbulence and is directly related to intake manifold configuration, fuel delivery system design and combustion chamber shape. Understanding the spatial mixture distribution may help improve the design of these aforementioned components. Consequently, a more complete combustion process may result, and emissions reduced. A method that measures the emission of CH and C2 radicals via the use of an optical fiber bundle was used in this research to map the mixture uniformity in the combustion chamber. The intensity ratio (IC2/ICH) was correlated to the fuel-air equivalence ratio. The mixture distribution measured was then correlated with the hydrocarbon emission sequence.

The increasing availability of camera-based displays for indirect vision in vehicles is providing new opportunities to supplement drivers' direct views of the roadway and surrounding traffic, and is also raising new issues about how drivers perceive the positions and movements of surrounding vehicles. We recently reported evidence that drivers' perception of the distance to rearward vehicles seen in camera-based displays is affected not only by the visual angles subtended by the images of those vehicles, but also by the sizes of those images relative to the sizes of the displays within which they are seen (an influence that we have referred to as a framing effect). There was also evidence for a similar, but weaker, effect with rearview mirrors.

In this paper, the formability of AA5754 aluminum laser-welded blanks produced by Nd:YAG laser welding is investigated under biaxial straining conditions. The mechanical behavior of the laser-welded blanks is first examined by uniaxial tensile tests conducted with the weld line perpendicular to the tensile axis. Shear failure in the weld metal is observed in the experiments. Finite element simulations under generalized plane strain conditions are then conducted in order to further understand the effects of weld geometry and strength on the shear failure and formability of these welded blanks. The strain histories of the material elements in the weld metal obtained from finite element computations are finally used in a theoretical failure analysis based on the material imperfection approach to predict the failure strains for the laser-welded blanks under biaxial straining conditions.

The spark process has previously been shown to heavily influence ignition stability, particularly in direct-injected gasoline engines. Despite this influence, few studies have addressed spark behavior in direct-injected engines. This study examines the role of environmental factors on the behavior of the spark. Through measurement of the spark duration, by way of the ignition current trace, several observations are made on the influence of external factors on the behavior of the spark. Changing the level of nitrogen in the cylinder (to simulate EGR), the level of wetting and velocity imparted by the spray, the ignition dwell time and the orientation of the ground strap, observations are made as to which conditions are likely to produce unfavorable (shorter) spark durations. Through collection of a statistically significant number of sample spark lengths under each condition, histograms have been assembled and compared under each case.

The importance of eye-to-display distance for distance perception in rear vision may depend on the type of display. At least in terms of its influence on the effective magnification of images, eye-to-display distance is almost irrelevant for flat rearview mirrors, but it is important for convex rearview mirrors and for other displays, such as video displays, that create images closer to the driver than the actual objects of interest. In the experiment we report here, we investigate the influence of eye-to-display distance on distance perception with both flat rearview mirrors and camera-based video displays. The results indicate that a simple model of distance perception based on the visual angles of images is not very successful. Visual angles may be important, but it appears that relationships between images of distant objects and the frames of the displays are also important. Further work is needed to fully understand how drivers might judge distance in camera-based displays.

The effects of image size on perceived distance have been of concern for convex rearview mirrors as well as camera-based rear vision systems. We suggest that the importance of image size is limited to cases-such as current rearview mirrors-in which the field of view is small. With larger, richer fields of view it is likely that other distance cues will dominate image size, thereby substantially diminishing the concern that distortions of size will result in distortions of distance perception. We report results from an experiment performed in a driving simulator, with static simulated rearward images, in which subjects were asked to make judgments about the distance to a rearward vehicle. The images showed a field of view substantially wider than provided by any of the individual rearview mirrors in current systems. The field of view was 38 degrees wide and was presented on displays that were either 16.7 or 8.5 degrees wide, thus minifying images by factors of 0.44 or 0.22.