Search Results

This paper describes the transient spray characteristics of a high pressure, single fluid injector, intended for use in a direct-injection spark-ignited (DISI) engine. The injector was a single hole, pintle type injector and was electronically controlled. A variety of measurement diagnostics, including full-field imaging and line-of-sight diffraction based particle sizing were employed for spray characterization. Transient patternator measurements were also performed to obtain temporally resolved average mass flux distributions. Particle size and obscuration measurements were performed at three locations in the spray and at three injection pressures: 3.45 MPa (500 psi), 4.83 Mpa (700 psi), and 6.21 MPa (900 psi). Results of the spray imaging experiments indicated that the spray shapes varied with time after the start of injection and contained a leading mass, or slug along the center line of the spray.

Cost reduction in the automotive industry becomes a widely-adopted operational strategy not only for Original Equipment Manufacturers (OEMs) that take cost leader generic corporation strategy, but also for many OEMs that take differentiation generic corporation strategy. Since differentiation generic strategy requires an organization to provide a product or service above the industry average level, a premium is typically included in the tag price for those products or services. Cost reduction measures could increase risks for the organizations that pursue differentiation strategy. Although manufacturers in the automotive industry dramatically improved production efficiency in past ten years, they are still facing the pressure of cost control. The big challenge in cost control for automakers and suppliers is increasing prices of raw materials, energy and labor costs. These costs create constraints for the traditional economic expansion model.

The application of Artificial Intelligence (AI) in the automotive industry can dramatically reshape the industry. In past decades, many Original Equipment Manufacturers (OEMs) applied neural network and pattern recognition technologies to powertrain calibration, emission prediction and virtual sensor development. The AI application is mostly focused on reducing product development and validation cost. AI technologies in these applications demonstrate certain cost-saving benefits, but are far from disruptive. A disruptive impact can be realized when AI applications finally bring cost-saving benefits directly to end users (e.g., automation of a vehicle or machine operation could dramatically improve the efficiency). However, there is still a gap between current technologies and those that can fully give a vehicle or machine intelligence, including reasoning, knowledge, planning and self-learning.

Neural networks are useful tools for optimization studies since they are very fast, so that while capturing the accuracy of multi-dimensional CFD calculations or experimental data, they can be run numerous times as required by many optimization techniques. This paper describes how a set of neural networks trained on a multi-dimensional CFD code to predict pressure, temperature, heat flux, torque and emissions, have been used by a genetic algorithm in combination with a hill-climbing type algorithm to optimize operating parameters of a diesel engine over the entire speed-torque map of the engine. The optimized parameters are mass of fuel injected per cycle, shape of the injection profile for dual split injection, start of injection, EGR level and boost pressure. These have been optimized for minimum emissions. Another set of neural networks have been trained to predict the optimized parameters, based on the speed-torque point of the engine.

A non-intrusive sensing technique to determine start of combustion for mixing-controlled compression-ignition engines was developed based on an accelerometer mounted to the engine block of a 4-cylinder automotive turbo-diesel engine. The sensing approach is based on a physics-based conceptual model for the signal generation process that relates engine block acceleration to the time derivative of heat release rate. The frequency content of the acceleration and pressure signals was analyzed using the magnitude-squared coherence, and a suitable filtering technique for the acceleration signal was selected based on the result. A method to determine start of combustion (SOC) from the acceleration measurements is presented and validated.

The University of Wisconsin - Madison Clean Snowmobile team has designed, constructed and now refined an electric snowmobile with 40 km (24 mi) range and acceleration comparable to a 75 kW (100 hp) internal-combustion-powered snowmobile. Starting with a Polaris IQ Fusion chassis, a direct-drive chain-case was engineered to couple a General Motors EV1 copper-bar rotor AC induction electric motor to the track drive shaft. The battery pack uses 104 28 V, 2.8 A-hr Lithium-Ion battery modules supplied by Milwaukee Tool to store 8.2 kW-hr of energy at a nominal voltage of 364 V. Power is transmitted to the electric motor via an Azure Dynamics DMOC445LLC motor controller. All of the components fit within the original sled envelope, leading to a vehicle with conventional appearance and a total mass of 313 kg (690 lb). The vehicle, dubbed the BuckEV, accelerates to 150 m (500 ft) in 6.9 seconds and has a top speed of 122 km/hr (76 mph) with a pass-by sound level of 55 dB.

Various single and split injection schemes are studied to provide a better understanding of fuel distribution during cold starting in DI diesel engines. Improved spray-wall interaction, fuel film and multicomponent vaporization models are used to analyze the combustion processes. Better combustion characteristics are obtained for the split injection schemes than with a single injection. An analysis of the fuel impingement processes identifies the mechanisms involved in producing the differences in vaporization and combustion of the fuel. A greater amount of splashing occurred for the split injections compared to a single injection. This behavior is attributed to the decreased film thickness (less dissipation of impingement energy), the decreased impingement area (obtained by increasing the impingement Weber number), and most importantly, the reduced frequency of drop impingement.

Neural networks have been used for engine computations in the recent past. One reason for using neural networks is to capture the accuracy of multi-dimensional CFD calculations or experimental data while saving computational time, so that system simulations can be performed within a reasonable time frame. This paper describes three methods to improve upon neural network predictions. Improvement is demonstrated for in-cylinder pressure predictions in particular. The first method incorporates a physical combustion model within the transfer function of the neural network, so that the network predictions incorporate physical relationships as well as mathematical models to fit the data. The second method shows how partitioning the data into different regimes based on different physical processes, and training different networks for different regimes, improves the accuracy of predictions.

A global optimization method has been developed for an engine simulation code and utilized in the search of optimal fuel injection strategies. This method uses a Lagrange interpolation function which interpolates engine output data generated at the vertices and the intermediate points of the input parameters. This interpolation function is then used to find a global minimum over the entire parameter set, which in turn becomes the starting point of a CFD-based optimization. The CFD optimization is based on a steepest descent method with an adaptive cost function, where the line searches are performed with a fast-converging backtracking algorithm. The adaptive cost function is based on the penalty method, where the penalty coefficient is increased after every line search. The parameter space is normalized and, thus, the optimization occurs over the unit cube in higher-dimensional space.

Lithium-ion and Lithium polymer batteries are fast becoming ubiquitous in high-discharge rate applications for military and non-military systems. Applications such as small aerial vehicles and energy transfer systems can often function at C-rates greater than 1. To maximize system endurance and battery health, there is a need for models capable of precisely estimating the battery state-of-charge (SoC) under all temperature and loading conditions. However, the ability to perform state estimation consistently and accurately to within 1% error has remained unsolved. Doing so can offer enhanced endurance, safety, reliability, and planning, and additionally, simplify energy management. Therefore, the work presented in this paper aims to study and develop experimentally validated mathematical models capable of high-accuracy battery SoC estimation.

Increasingly, the exchanges of data in complex ECM (Electronic Control Module) systems rely on multiple communication networks across various physical and network layers. This has greatly increased system flexibility and provided an excellent medium to create well-defined exchangeable interfaces between components; however this added flexibility comes with increased network complexity. A system-level approach allows for the optimization of data exchange and network configuration as well as the development of a comprehensive network failure strategy. Many current ECM systems utilize complex multi-network communication strategies to exchange and control data to components. Recently, Caterpillar implemented an HIL (Hardware-In-the-Loop) test system that provides an approach for developing and testing a comprehensive ECM network strategy.

Airflow characteristics surrounding non-evaporating transient diesel sprays were investigated using a 6-hole injector. Particle Image Velocimetry (PIV) was used to measure the gas velocities surrounding a spray plume as a function of space and time. A hydraulically actuated, electronically controlled unit injector (HEUI) system was used to supply the fuel into a pressurized constant volume chamber at room temperature. The chamber gas densities in this study were 10 kg/m3, 20 kg/m3 and 30 kg/m3. The injection pressure was 96.5 MPa. Two frequency doubled (532 nm) Nd:YAG lasers were used to create coincident laser sheets to illuminate the test section at two instances after start of injection (ASI). The double exposed images of sprays and Al2O3 seed particles were developed and velocity vectors of the gas surrounding the transient diesel sprays were obtained using a numerical autocorrelation PIV method.

We present an approach in which an open-source software infrastructure is used for testing the behavior of autonomous vehicles through computer simulation. This software infrastructure is called CAVE, from Connected Autonomous Vehicle Emulator. As a software platform that allows rapid, low-cost and risk-free testing of novel designs, methods and software components, CAVE accelerates and democratizes research and development activities in the field of autonomous navigation.

Despite numerous research efforts, there is no reliable and widely accepted tool for the prediction of erosion prone material surfaces due to collapse of cavitation bubbles. In the present paper an Erosion Aggressiveness Index (EAI) is proposed, based on the pressure loads which develop on the material surface and the material yield stress. EAI depends on parameters of the liquid quality and includes the fourth power of the maximum bubble radius and the bubble size number density distribution. Both the newly proposed EAI and the Cavitation Aggressiveness Index (CAI), which has been previously proposed by the authors based on the total derivative of pressure at locations of bubble collapse (DP/Dt>0, Dα/Dt<0), are computed for a cavitating flow orifice, for which experimental and numerical results on material erosion have been published. The predicted surface area prone to cavitation damage, as shown by the CAI and EAI indexes, is correlated with the experiments.

A procedure has been developed to build system level predictive models that incorporate physical laws as well as information derived from experimental data. In particular a soot model was developed, trained and tested using experimental data. It was seen that the model could fit available experimental data given sufficient training time. Future accuracy on data points not encountered during training was estimated and seen to be good. The approach relies on the physical phenomena predicted by an existing system level phenomenological soot model coupled with ‘weights’ which use experimental data to adjust the predicted physical sub-model parameters to fit the data. This approach has developed from attempts at incorporating physical phenomena into neural networks for predicting emissions. Model training uses neural network training concepts.