The computer program LS-DYNA3D was used to simulate the behavior of a specific, though representative, heavy truck cab-over tractor-trailer vehicle during a full 180° rollover event. These simulations provide a key component in the development of a physical testing procedure for evaluating structural integrity and occupant crash protection system designs in heavy trucks.
Vehicle crashworthiness for passenger automobiles has been the focus of numerous research studies in recent years ((Hallquist, 1989), (Schwer, 1996)), but comparatively little attention has been paid to the crashworthiness of heavy trucks. The United States government, vehicle manufacturers, safety researchers, and private groups have all agreed on the need to better understand and improve heavy truck crashworthiness. Based on these concerns, the Society of Automotive Engineers (SAE) initiated a three-phase research program to evaluate various crashworthiness issues for heavy trucks. Phase I of the SAE Heavy Truck Crashworthiness program entails development of characteristic crash pulses and analysis of truck occupant dynamics (Cheng, et al., 1996a, 1996b, 1996c). Phase II of the program entails both computational and experimental tasks for the investigation of 180° rollovers, and is the subject of this paper. The Phase III work establishes recommended practices for evaluating truck crashworthiness designs.
The 180° dynamic rollover analyses that form Phase II of the SAE study provide critical information on the magnitude, direction and spatial distribution of ground contact forces during a rollover. The contact force vectors can be used to plan efficient full-scale static tests that more faithfully reproduce actual rollover forces on the cab structure. The analysis also provides a means for comparing different cab structural parameters, although this is not addressed in this paper.
Whereas typical barrier or vehicle-to-vehicle impacts have a duration of approximately 0.1 seconds, rollover events are typically one to three seconds, or more, in length. Such long duration events, which necessarily involve large deformation structural response, inelastic material behavior and element failure, present a significant computational challenge. Since the initial phase of the rollover event is essentially rigid, a kinematic (rigid-body) vehicle analysis can be used up to the instant the cab contacts with the ground. The deformable phase begins at the instant of cab-to-ground contact, and in this study, is carried forward for 1.25 seconds, using a deformable (finite element) model.
The remainder of this paper explains how the rigid body and the finite element analyses were developed then coupled together to form the complete rollover analysis. Key results from the rollover simulation are also presented.
Finite Element Model
The computer program ANSYS was used to create the finite element model of the cab. The commercial computer program LS-DYNA3D / LS-TAURUS (Hallquist, 1994) was selected for the FE analysis because of its capabilities in handling contact, material degradation and tearing, weld failure, rigid body/flexible body coupling, and because the user has good control of time step size through mass scaling. Model size and mesh density were carefully optimized in order to avoid computation times of several days to more than a week on workstation-class computers. Because LS-DYNA3D uses an explicit solution method, mesh density was particularly critical for the few small or stiff elements that determine critical time step size.
An actual cab-over heavy truck cab design was used to construct the finite element model. In addition to detailed engineering drawings, three exemplar specimens of the cab structure were obtained; two were used for full-scale testing (described later in the paper), the third served as a modeling reference. Altogether, more than 60 component parts were included in the cab model alone. Close attention was paid to the front panel and door areas, and also to the connection details, as they have a profound influence on the load path, especially for nonlinear analyses. The final cab model, shown in Fig. 1, consists of 45 combinations of material types and shell thicknesses, 15 different beam sections, 8000 elements, and 9800 nodes.
The complete finite element model for the rollover simulation includes the flexible cab, a rigid chassis, and a rigid trailer. The geometry of the rigid bodies was defined with a finite element mesh, then assigned rigid material properties, then explicitly assigned inertial properties (CG location, mass and rotational inertias). Nodal constraints were used to join the cab to the chassis at four support points, and a ball-and-socket joint connects the chassis and trailer at the fifthwheel. A nonlinear rotational spring was lumped at the fifthwheel to provide resistance against relative torsion (about the longitudinal axis), and to approximate the windup energy stored in the torsionally flexible trailer.
T. Kim Parnell, PhD, PE is a Mechanical Engineering consultant with strong experience in a number of technology areas. He holds PhD and MSME degrees from Stanford University in Mechanical Engineering. Dr. Parnell specializes in the mechanical engineering design and behavior of Biomedical Devices, Shape Memory Metals, Bioabsorbable Polymers, MEMs, Electronic, and Miniature Components. He consults actively in these areas, as well as in failure analysis and reliability.
©Copyright - All Rights Reserved
DO NOT REPRODUCE WITHOUT WRITTEN PERMISSION BY AUTHOR.