Options
2015
Conference Paper
Title
Crash simulation of adhesively bonded structures
Abstract
In recent years structural adhesives have become more and more popular in the automotive industry. The main reason for this is the potential to reduce vehicle weight and cost by reducing the amount of rivets and bolts used in the structure. At the same time, the standards for passenger safety are high and crashworthiness has to be considered during the early design phase. Consequently, finite element (FE) crash simulation plays an important role in the automotive design process. Within this framework there is a need for numerically efficient, predictive models describing the rate-dependent response of adhesively bonded joints as structural adhesives are increasingly used in crash safety relevant components. Additionally, industry requires a straight forward model calibration procedure based on limited amount of simple tests. Cohesive zone models have in the past been successfully used to describing crack initiation and subsequent propagation in composite materials or adhesively bonded joints subjected to quasi-static loading conditions. This type of model can nowadays be found in all commercial codes. However, cohesive zone models dedicated to modeling crash by considering rate-dependent material parameters such as strength and fracture toughness are not yet available. In this paper, a physically-based rate-dependent cohesive zone model is presented. The model combines a bi-linear traction-separation law describing the peel response with a trapezoidal traction-separation law describing the shear response of the adhesive. The model takes into account four rate-dependent quantities: mode I (peel) strength, mode II (shear) strength, mode I fracture toughness, and mode II fracture toughness of the adhesive. The model was implemented into the commercial finite element code Abaqus/Explicit by means of a user-defined material. The model was calibrated for the crash-optimized adhesive DOW BETAMATE 1496V using four different types of test at different rates of loading. The rate-dependent mode I strength was determined using butt-joint tests. The rate-dependent mode II strength was determined using compressive double lap shear tests. The rate-dependent mode I and mode II fracture toughness were measured using the tapered double cantilever beam test and the tapered end notched flexure test, respectively. The model was validated against experiments performed on adhesively bonded metallic T-joints subjected to different loading directions and loading velocities (up to 5 m/s). The numerical predictions showed an excellent correlation to the experimental evidence, both in terms of the qualitative damage and failure sequence as well as the force-displacement response.