Robust numerical modeling of adhesively-bonded joints for safety assessment of vehicle body structures subjected to extreme loading

includes bibliographical references and index

PhD; 2023; Centre for Product Design and Manufacturing

Continuous adhesively-bonded joints can lead to stiffer vehicle body structures in
contrast to conventional spot-welded steel sheet metal-based unitized vehicle body
design. For a given target stiffness, adhesive bonding, in lieu of spot welding, can
thus provide scope for lightweighting which in turn can make vehicles more fuel
efficient. The performance of adhesively bonded joints, however, has not been
adequately proven under severe dynamic loading conditions such as arising from
impact and blast-induced shock wave propagation. There is also a deficiency in
published literature on robust numerical modelling of adhesively-bonded joints
exhibiting predominantly shear and peel stresses, and applying the same towards
prediction of the behavior of structural components such as steel hat sections
subjected to axial and transverse impact. Additionally, the system-level performance
of adhesively-bonded joints in full vehicle body design under extreme loading
conditions, which probably goes beyond the general purview of academic research,
does not appear to have been reported in open literature.
Keeping the above points in mind, the current research is aimed at systematically
studying the mechanical behaviors of adhesively-bonded joints with steel substrates
at coupon, component and full vehicle levels with a judicious combination of physical
testing and nonlinear explicit finite element analysis (FEA). With the stated objective,
coupon specimens of various joint configurations have been tested in a UTM to begin
with till failure, and their behaviors reproduced using cohesive zone modeling in LSDYNA, an explicit contact-impact FEA solver, with uncoupled mode I and mode II
fracture mechanics properties. The right constitutive model had to be arrived at
through extensive comparison of relevant material models including those based on
von Mises and Drucker-Prager yield criteria for classical elasto-plastic stress-strain
behaviors of materials. As part of the systematic study envisioned, thin-walled hollow
members with double-hat sectional profiles of a given length were fabricated with spot welding, only adhesive bonding, and with hybrid joining i.e. predominantly
adhesive bonding with sparse spot welding. These specimens were then subjected to
axial and transverse impact loading in a drop-weight test setup. Not only valuable
insights into the relative behaviors of the components of diverse joining techniques
were obtained, but also the nonlinear dynamic responses of the components became
valuable data for further validating the finite element (FE) modelling procedure
developed previously at the level of coupon specimens. With an emphasis on realworld applications, conventional spot welded front rails were replaced in turn with
only adhesively-bonded and hybrid welded-bonded rails, in a detailed FE model of a
compact passenger car and subjected to full frontal impact against a rigid barrier as
in a US-NCAP test with a closing speed of 56 kmph. This study provided a framework
for assessing the performance of a purely adhesively-bonded rail in a full vehicle
crash test. Finally, using an ALE (Arbitrary Lagrangian-Eulerian) modelling
procedure, the effect of an underbody blast under a vehicle resulting in a shock wave
propagating through the ambient air and striking underneath the floor of a vehicle was
captured and the potentially severe injury caused to the lower limb of an occupant
predicted. In this connection, the FE model of a military lower leg extremity (MILLX) anthropomorphic test device was formulated and various novel floor-based
countermeasures with spot welding and adhesive bonding compared for mitigation of
lower limb injuries to vehicle occupants.