Failure Of Composite Beams Research Articles

Fire behavior of composite beams with simple connections: Benchmarking of numerical models

This paper focuses on the development and benchmarking of 3D finite element models for predicting the behavior and failure of composite beams with simple (shear) connections subjected to gravity loads and fire conditions, which include both heating and cooling phases. The simple connections included are the shear-tab, single-angle, and double-angle connections commonly used in the U.S. engineering practice. All the structural components including the steel beam, concrete slab, shear connectors, and connection elements (bolts, plates, angles etc.) were modeled in detail using appropriate finite elements, contact interaction models, and material and component behavior at elevated temperature. The detailed finite element models were benchmarked using results from large-scale tests of composite beams and connections subjected to fire loading including heating and cooling phases. The numerical results including the midspan displacement histories and the connection rotation histories were compared with those measured experimentally by the authors. These comparisons indicate that the detailed models reasonably predict the experimental deformation histories and observed failure modes. The models were used to provide additional insight into the axial forces developed in the connections, particularly during the cooling phase, thus enhancing experimental findings. The benchmarked model is recommended for conducting analytical parametric studies to supplement experimental investigations of composite beams and simple connections under fire.

Dynamic loading of layered composite beams with damage of delamination type

Failure of composite beams containing initial delamination is investigated under low-speed impact by a dropping weight. Bolotin's theory of dynamic failure (1992) is used. The start and stop time of growth of a nonsymmetric delamination crack is studied. It is found that the initial extension starts and subsequently continues at the crack tip, which is located closer to the impact point. Then extension is observed on the second crack tip, and finally a synchronous growth of delamination occurs at both crack tips. With constant impact energy, the final size of delamination does not depend on its initial size. The start and stop time of delamination growth increases as the initial defect approaches the beam surface.

Scaling effects in the flexural response and failure of composite beams

An experimental program was conducted in which scale model graphite-epoxy composite beams were loaded in bending until failure to investigate the effects of specimen size on flexural response and strength. Tests were performed on unidirectiona l, angle-ply, cross-ply, and quasi-isotropic beams of eight different scaled sizes ranging from 1/6 scale to full scale. The beams were subjected to an eccentric axial compressive load designed to promote large bending deformations and rotations. Results indicated that, although normalized flexural response was independent of specimen size, a significant scale effect in strength was observed for all laminate types. Typically, failure stresses and strains decreased as the size of the beam specimen increased from subscale to the full-scale prototype. Standard failure criteria for composite materials, such as the maximum stress, maximum strain, and tensor polynomial criteria, cannot predict the strength degradation with increasing specimen size. Consequently, a Weibull statistical model and a fracture mechanics model were used to analyze the strength scale effect.