Search Results

LiDAR sensors have played more and more important role on Intelligent and Connected Vehicles (ICV) and Advanced Driver Assistance Systems (ADAS) .However, the development and testing of LiDAR sensors under real driving environment for ADAS applications are greatly limited by various factors, and often are impossible due to safety concerns. This paper proposed a novel functional LiDAR model under virtual driving environment to support development of LiDAR-based ADAS applications under early stage. Unlike traditional approaches on LiDAR sensor modeling, the proposed method includes both geometrical modeling approach and physical modeling approach. While geometric model mainly produces ideal scanning results based on computer graphics, the physical model further brings physical influences on top of the geometric model. The range detection is derived and optimized based on its physical detection and measurement mechanism.

This paper proposes a novel allocation-based control method for four-wheel independently actuated electric vehicles. In the proposed method, both actuator dynamics and input/output constraints are fully taken into consideration in the control design. First, the actuators are modeled as first-order dynamic systems with delay. Then, the control allocation is formulated as an optimization problem, with the primary objective of minimizing errors between the actual and desired control outputs. Other objectives include minimizing the power consumption and the slew rate of the actuator outputs. As a result, this leads to frequency-dependent allocation that reflects the bandwidth of each actuator. To solve the optimization problem, an efficient numerical algorithm is employed. Finally the proposed control allocation method is implemented to control a four-wheel independently actuated electric vehicle.

In order to understand the flow and wind noise characteristics generated by the outside rearview (OSRV) mirror, a series of wind noise measurements for two production mirrors was conducted at the GM Aerodynamics Lab (GMAL) wind tunnel. These measurements included the time-averaged static pressures, surface noise sources, and far field propagation noise. The data obtained in this investigation will be used for future CFD numerical validations. The two mirrors chosen for the test are the GMT360 (a truck mirror) and the GMX320 (a sedan mirror). The test mirror was mounted on an elevated table which was specially designed for the current project to avoid any significant flow boundary layer buildup on the wind tunnel floor. The test conditions reported in this paper include four inlet speeds of 30, 50, 70 and 90 mph at 0 yaw angle. To record the wind noise sources, nine surface flush-mount microphones were used.

This paper describes a computational and experimental effort to document the detailed flow field around a pickup truck. The major objective was to benchmark several different computational approaches through a series of validation simulations performed at Clemson University (CU) and overseen by those performing the experiments at the GM R&D Center. Consequently, no experimental results were shared until after the simulations were completed. This flow represented an excellent test case for turbulence modeling capabilities developed at CU. Computationally, three different turbulence models were employed. One steady simulation used the realizable k-ε model. The second approach was an unsteady RANS simulation, which included a turbulence closure model developed in-house. This simulation captured the unsteady shear layer rollup and breakdown over the front of the hood that was expected and seen in the experiments but unattainable with other off-the-shelf turbulence models.

Computational Fluid Dynamics (CFD) has gained popularity as a tool for many airflow situations including road vehicle aerodynamics. This trend, to bring CFD to bear on vehicle aerodynamic design issues, is appropriate and timely in view of the increasing competitive and regulative pressures being faced by the automotive industry. For a large portion of the engineering community, the primary source of CFD capabilities is through the purchase of commercial CFD codes. This paper summarizes the results of a series of benchmark external aerodynamic simulations that were carried out for a generic pickup truck model using two commercial CFD codes, namely Fluent and the PowerFLOW. For direct comparisons the computations and the experiments were performed for the same model (vehicle) geometry and under similar flow conditions.

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.

This paper describes the computational results of the flow field around two vehicle geometries using the Immersed Boundary (IB) technique in conjunction with a steady RANS CFD solver. The IB approach allows the computation of the flow around objects without requiring the grid lines to be aligned with the body surfaces. In the IB approach instead of specifying body boundary conditions, a body force is introduced in the governing equations to model the effect of the presence of an object on the flow. This approach reduces the time necessary for meshing and allows utilization of more efficient and fast CFD solvers. The simulations are carried out for an SUV and a pickup truck models at a Reynolds number of 8×105. Cartesian meshes (non-uniform) with local grid refinement are used to increase the resolution close to the boundaries. The simulation results are compared with the existing measurements in terms of surface pressures, velocity profiles, and drag coefficients.

Joint failure, e.g. resistance spot weld (RSW) joint failure, has been identified as one of the key failure modes when vehicle crash occurs. Currently, a beam element or a rigid link is commonly used to represent a RSW in a full vehicle analysis. Several failure criteria have been implemented in the math-based tools to model the spot weld failure, namely the failure strain and the resultant force based failure criterion. The former is not robust and the latter requires a lot of tests to be done for different designs of the resistance spot welded joints. In the case when the RSW joint has pullout failure, i.e. the failure occurs in the base metal, a material model is used in the present work to model the RSW joint pullout failure by modeling the failure of the base material. Only simple tensile tests are needed in order to use this method while the geometry effects of the joint design and the strain rate dependence of the base material are handled by the material model itself.

One way to achieve low temperature regeneration on diesel particulate traps is to employ NO2 as the oxidant. However, the engine may not produce sufficient NOx to achieve the required particulate regeneration. An RRT (rolling regeneration trap) was proposed as a way to enhance the effective concentration in diesel engine exhaust. The RRT uses catalyzed ceramic foam, which allows repeated use of existing NOx in the exhaust stream. The ceramic foam is a filter itself: thus, it can reduce the particulate loading on the wall-flow filter, which is more prone to plugging by the particulate deposits. Furthermore, the presence of particulate matter in the catalyzed section circumvents the high temperature limit imposed by thermodynamic equilibrium of a simple NO+NO2 +O2 system. However, experimental results revealed that the regeneration efficiency on the wall-flow section was very low and NO2 slip is an issue.

Simple devices have been shown to be capable of tailoring the flow field around a vehicle and reducing aerodynamic drag. An experimental and computational investigation of a drag reduction device for bluff bodies in ground proximity has been conducted. The main goal of the research is to gain a better understanding of the drag reduction mechanisms in bluff-body square-back geometries. In principle, the device modifies the flow field behind the test model by disturbing the shear layer. As a consequence, the closure of the wake is altered and reductions in aerodynamic drag of more than 20 percent are observed. We report unsteady base pressure, hot-wire velocity fluctuations and Particle Image Velocimetry (PIV) measurements of the near wake of the two models (baseline and the modified models). In addition, the flows around the two configurations are simulated using the Reynolds Averaged Navier-Stokes (RANS) equations in conjunction with the V2F turbulence model.

This paper describes the initial phases of an on-going project in the GM Motorsports Safety Technology Research Program to investigate Indy car crashes using an on-board impact recorder as the primary data collection tool. The development of a database consisting of crash investigation data patterned after national highway crash databases is discussed. The data gathered and coded includes track and incident scene information, vehicle damage, and driver injuries, as well as the vehicle decelerations measured by the impact recorder. The paper discusses the development of specifications for the impact device, the selection of the specific recorder and its implementation on a routine basis in Indy car racing. The results from incidents that produced significant data during the 1993, 1994 and 1995 racing seasons are summarized.