Date of Award
12-2008
Document Type
Thesis
Degree Name
Master of Engineering (ME)
Legacy Department
Mechanical Engineering
Committee Chair/Advisor
Thompson, Lonny L
Committee Member
Biggers , Sherrill B
Committee Member
Joseph , Paul F
Abstract
In order to design car body structures which are safe during crash, modern automotive manufacturers perform both full-scale experimental crash tests and computer simulation of vehicle crash events using commercially available Finite Element Analysis (FEA) packages such as ABAQUS or LS-DYNA. Use of crash simulations significantly reduces the number of real time crash experiments needed and reduces the time required for design changes. However, in order to capture accurately crash behavior during high-speed impact, a large amount of detailed FEA modeling features such as number and types of elements, mesh element size, number of components, different types of connectors, material properties, and other detailed features are needed. Crash simulation requires explicit time-stepping procedures, which can be computationally expensive for complicated full vehicle models with many components.
An important feature in crash simulation is the amount of detail included in modeling spot weld connections. Traditionally for efficiency, simple node-to-node rigid connections for modeling spot weld connections between different components are used, especially when many components are connected in a full vehicle crash model. Recent studies have shown the importance of accurate modeling including elastic stiffness and failure modes for spot welds due to high impact loads in automotive crash analysis. For efficiency and convenience, most commercially available FEA packages now include the option of creating mesh independent spot welds, which allow the user to define the location of the center point of the spot weld and define the spot weld radius on adjacent surfaces of connected components. A distributed coupling to nodes within the radius specified is automatically created which approximates the behavior of a spot weld of finite size. In addition, the size of the rigid spot weld model provides greater accuracy compared to the simple node-to-node connection. However, it has not been until very recent that some researchers and commercially available FEA software have the ability to include important spot weld elastic properties and failure modes combining pull, peal, shear, and torsion.
In this work, different levels of complexity in spot weld modeling are examined in terms of sufficient accuracy which can be used efficiently for impact analysis of large connected components and full vehicle crash models. In order of increasing complexity, the following spot weld models are considered and results compared: (a) simple node-to-node rigid connection, (b) rigid mesh independent spot welds, (c) elastic mesh independent spot welds, and (d) elastic with failure mesh independent spot welds. In order to study the fundamental behavior of the different mesh-independent spot weld models, pullout and peal tests between two thin ductile steel plates are performed which isolate different failure modes. Comparisons of reaction force versus displacement curves and internal energy versus displacement for all the different spot weld models are given. Results indicate that the rigid connected results in peak reaction forces which are much larger than elastic spot welds. The spot weld model, which includes failure, follows the same path as the elastic weld but when reaching the particular failure force the reaction remains constant with additional applied displacement.
To better understand the behavior of the spot-weld models for crash analysis on a realistic and important automotive component which exhibits complex crushing modes with combined axial and bending a frontal longitudinal rail designed for strength and energy absorption was studied with a node-to-node rigid spot weld compared with mesh independent rigid and elastic spot weld connections. The frontal longitudinal rail is a thin walled closed section located in between the front bumper and the firewall manufactured from two stamped sheets with spot welds on both sides of flanges at discrete intervals along the length. In addition to spot welds, the effect of various shape and size parameter changes including waves, beads, and a small rib for crush initiation that significantly increase energy absorption and crush force efficiency for the rail component are proposed.
Recommended Citation
Vimalanathan, Srinivasan, "An Investigation on Spot-Weld Modeling Complexity for Crash Simulation" (2008). All Theses. 526.
https://open.clemson.edu/all_theses/526