Search Results

A finite element model utilizing the smoothed particle hydrodynamics (SPH) algorithm in LS-DYNA is developed to simulate the behavior of automotive fuel tanks in drop tests. Limited physical tests described in the open literature have demonstrated failures in certain reconditioned fuel tanks, but the tests were inconclusive. The modeling effort described in this paper is aimed at providing insight into the complex loading on the tanks resulting from the fluid-structure interaction. Application of the SPH algorithm to model a variety of fuel tanks as well as laboratory test specimens shows distinct correlations between tank geometry design and its propensity for failure, as well as differences in loading resulting from substitution of water for gasoline.

According to National Highway Traffic Safety Administration (NHTSA) speed-related traffic fatalities accounted for 31% of total fatalities on U.S. roadways in 2003. Traditional speed control methods suffer from significant shortcomings. Adaptation (ISA) systems hold the promise of safer roadways through improving driver compliance with speed limits. This paper describes the development of a new multi-modal speed adaptation system to be tested in the CISR car-driving simulator. The system is capable of adapting to the driver's driving style and provides appropriate warning for over speeding based on the vehicle speed, speed limit, driver individual preferences, and risk factor. A hierarchical manager module determines the warning strategy. The adequate warning strategy is specific to driving situations and individual characteristics. Modes of warnings being considered include VISUAL, and HAPTIC.

An active torque control steering system is developed and implemented in a car simulator. The simulator has a comprehensive and accurate full vehicle dynamics and road/environment models. A simple model of the driving simulator’s vehicle was developed and a PID controller, which uses the vehicle’s yaw angle, and position, was designed to control vehicle steering torque. The controller is then integrated with the driving simulator program, emulating the real world conditions. The developed system was tested in various obstacle avoidance and lane change scenarios in the car simulator, and the vehicle was able to avoid the stationary obstacles autonomously.

Vehicle steering control can provide assistance to drivers for lane keeping, automated trajectory following, or more extreme evasive maneuvers. An active torque control steering system is designed using Linear Quadratic Regulator (LQR), and its performance was evaluated using the commercial software CARSIM. The system is developed to maintain a desired trajectory for the vehicle in performing evasive maneuvers to avoid imminent crash scenarios. In order to better understand the behavior of the system with different controllers, a simple bicycle model of the vehicle was developed, and an LQR controller was developed to control vehicle steering torque. The controller uses yaw angle, yaw rate, velocity, and position of the vehicle to generate the required steering torque to follow the desired trajectory. An observer was developed to estimate non-measured parameters. Trajectories are generated to follow a lane change before reaching the obstacle.

A previously defined methodology for the development of Finite Element (FE) detailed models of road vehicles has been utilized to create four different FE models of Child Restraint Systems (CRS). The resulting CRS “fleet” includes two convertible toddler seats, one infant rear-facing seat and a booster seat. Model dimensions range from 5,865 nodes, 4,168 elements and 2 parts for the reduced booster seat to 18,204 nodes, 21,345 elements and more than 20 parts for the most complex convertible one. All adjustable reclining/rocking positions have been considered during the model definition. The main characteristics of the models, as well as the reverse engineering process, are discussed in this paper. Particular material properties have not been studied in depth, but some insight is also offered on this subject to the prospective analyst or modeler. Versatility of the models and future research work required to fully validate the models are briefly commented.

Current link layer protocols for safety-related inter-vehicle communication networks suffer from significant scalability and security challenges. Carrier Sense Multiple Access (CSMA) approaches may produce excessive transmission collisions at high vehicle densities and are vulnerable to a variety of Denial of Service (DoS) attacks. Explicit time slot allocation approaches tend to be limited by either the need for a fixed infrastructure, a high number of control messages, or poor bandwidth utilization, particularly in low-density traffic. This paper will present a Communications Architecture for Adaptive Reliable Adhoc Networks (CARAVAN). CARAVAN includes novel adaptations of explicit timeslot allocation protocols for IVC networks.

This paper investigates the effects of driver airbag inflation characteristics, airbag relative position, airbag to dummy relative velocity, and steering column characteristics using a finite element model of a vehicle, air bag, and Hybrid III 50% male dummy. Simulation is conducted in a static test environment using a validated finite element model. Several static simulation tests are performed where the air bag module's position is mounted in a rigid steering wheel and the vertical and horizontal distances are varied relative to the dummy. Three vertical alignments are used: one position corresponds to the head centered on module, another position corresponds to the neck centered on module, and the third position centers the chest on the module. Horizontal alignments vary from 0 mm to 50 mm to 100 mm. All of these tests are simulated using a typical pre-1998 type inflation curve (mass flow rate of gas entering the bag).

This paper explains the characteristics of a highly detailed finite element model of a 1997 Dodge Grand Caravan, intended primarily for crashworthiness applications. The interest of the model, which was funded by a federal grant, stems in part from the its public domain availability. It consists of approximately 540 parts, 337,000 elements, 349,000 nodes and 6,500 connections. Since it will be mainly used to represent its class of vehicle, in this case the minivan type, an alternative concept of validation for a category of vehicles is discussed at the end of the article, together with the versatility and limitations of the model.