Search Results

This paper describes an analysis of the Diesel Oil Surrogate (DOS) model used at Chalmers University (Sweden), including 70 species participating in 310 reactions, and subsequent improvements prompted by the model's systematic tendency to under-predict the combustion intensity in simulations of kinetically-driven combustion modes, e.g. Homogeneous Charged Compression Ignition (HCCI). Key bases of the model are the properties of a model Diesel fuel with the molecular formula C14H28. In the vapor phase, a global reaction decomposes the starting fuel, C14H28, into its constituent components; n-heptane (C7H16) and toluene (C7H8). This global reaction was modified to yield a higher n-heptane:toluene ratio, due to the importance of preserving an n-heptane-like cetane number.

The aim of this study was to examine the influence of computer manikin users' background and knowledge for the results of a computer manikin simulation. Subjects taking part in the study were either production engineers or ergonomists. A manual task that presented production and ergonomics problems was used. The task was simulated prior to the subjects' sessions, using the computer manikin software Jack. During the sessions, the animated simulation was shown to the test subject. Results show that there are differences in how production engineers and ergonomists interpret results from a manikin simulation. Depending on the user's background, certain aspects that are difficult to visualise with the computer manikin were interpreted differently, regarding e.g. detected problems and holistic perspectives.

The automotive industry makes extensive use of virtual models to increase efficiency during the development stage. The complexity of such virtual models increases with the complexity of the process that they describe, and thus new methods for their development are constantly evaluated. Among many others, data-driven techniques and machine learning offer promising results, creating deep neural networks that map complex input-output relations. This work aims at comparing the performance of two different neural network architectures for the estimation of the engine state and emissions (flow fuel, NOx and soot). More specifically, Convolutional Neural Network (CNN) and Long-Short Term Memory (LSTM) will be evaluated in terms of performance, using different techniques to increase the model generalization. During the learning stage data from different engine cycles are fed to the neural networks.

The current work investigates a method in 1D modeling of cooling systems including discretized cooling package with non-uniform boundary conditions. In a stacked cooling package the heat transfer through each heat exchanger depends on the mass flows and temperature fields. These are a result of complex three-dimensional phenomena, which take place in the under-hood and are highly non-uniform. A typical approach in 1D simulations is to assume these to be uniform, which reduces the authenticity of the simulation and calls for additional calibrations, normally done with input from test measurements. The presented work employs 3D CFD simulations of complete vehicle in STAR-CCM+ to perform a comprehensive study of mass-flow and thermal distribution over the inlet of the cooling package of a Volvo FM commercial vehicle in several steady-state operating points.

The automotive industry continues to increase the utilization of computer-aided engineering. This put demands on finding reliable objective measures that correlate to subjective driver assessments on driving stability performance. However, the drivers’ subjective perception of driving stability can be difficult to quantify objectively, especially on test tracks where the wind conditions cannot be controlled. The advancement in driving simulator technology may enable evaluation of driving stability with high repeatability. The purpose of this study is to correlate the subjective assessment of driving stability to reliable objective measures and to evaluate the usefulness of a driving simulator for the subjective assessment. Two different driver clinic studies were performed in a state-of-the-art driving simulator. The first study included 38 drivers (professional, experienced and common drivers) and focused on crosswind gust sensitivity.

In vehicle development, a subjective evaluation of the vehicle’s behavior at high speeds is usually conducted by experienced drivers with the objective of assessing driving stability. To avoid late design changes, it is desirable to predict and resolve perceived instabilities early in the development phase. In this study, a mathematical model is developed from measurements during on-road tests to predict the driver’s ability to identify vehicle instabilities under excitations such as aerodynamic excitations. A vehicle is fitted with add-ons to create aerodynamic excitations and is driven by multiple drivers on a high-speed track. Drivers’ evaluation, responses, cabin motion, and crosswind conditions are recorded. The influence of yaw and roll rates, lateral acceleration, and steering angle at various frequency ranges when predicting the drivers’ evaluation of induced excitation is demonstrated. The drivers’ evaluation of vehicle behavior is influenced by driver-vehicle interactions.