Date of Award

12-2001

Level of Access Assigned by Author

Campus-Only Thesis

Degree Name

Master of Science (MS)

Department

Civil Engineering

Advisor

William Davids

Second Committee Member

Eric N. Landis

Third Committee Member

Roberto Lopez-Anido

Abstract

The use of reinforced glulam beams as bridge girders raises several concerns. Bridges see many repetitions of high loads throughout their life span, which may generate fatigue related issues in either the wood or the FRP. Furthermore, bridge girders are subjected to the atmospheric variations, under which the wood will exhibit significant dimensional changes whereas the FRP will undergo only small dimensional changes. Since these materials are bonded together, significant stresses may develop near the material interface.

In this study, four 164 beams reinforced with glass FRP were fatigued at design loads. Two of the beams had full reinforcement and two of the beams had partial length reinforcement. Two reinforcement types were used: wet-preg glass with a MOE in tension of 5750 ksi and pultruded glass with a MOE in tension of 63 1 1 ksi. Of the beams tested one was loaded to its design moment and 72% of its design shear and the other three were loaded to both their design moment and design shear simultaneously. Three of the beams exhibited shear failures and the beam with partial length pultruded reinforcement had an FRP delamination failure.

A moisture transport program was developed to predict moisture gradients through a wood beam's cross section due to seasonal moisture variations. The program uses Fickian diffusion and has the ability to model both linear and nonlinear diffusion coefficients. It has the ability to use a variety of boundary conditions including impervious surfaces. It can account for various grain orientations and conveys its output in both graphical and tabular forms. A finite-element model was also incorporated with the moisture transport model. The finite-element model used predictions from the moisture model to estimate dimensional changes and the resulting stresses and strains caused at the wood-FRP interface in a reinforced glulam beam.

A clip gage with a long gage length was developed to reliably measure strains in a heterogeneous material. The gage consists of a thin metal strip of spring steel that is arched and clipped between two tabs that are adhered to a specimen. Strain is measured by relating the displacement of the two tabs to the change in arch of the spring steel. The degree to which the spring steel is arched is measured by bonding foil gages at the steels midspan in a full or half Wheatstone bridge configuration. The end result is an accurate gage that is constructed in-house, made out of standard materials, reusable, inexpensive, and easy to attach.

A finite-element analysis was done on FRP-glulam beams with partial-length tension reinforcement to investigate stress concentrations at the ends of the reinforcement. The procedure included the use of the FEA program ANSYS to analyze girders in four-point bending. A modeling scheme, which incorporated the use of submodels, was used to obtain the desired mesh refinement at the ends of the reinforcement. A fully fixed, dimensionless bond line was assumed with no adhesive soaking into either material. The resulting stresses were averaged over short distance from the end of the reinforcement to counteract the affects of a stress singularity at that point. The results show that stresses at the ends of the FRP reinforcement are significantly higher than beam theory predicts. These stresses can contribute toward peeling the FRP reinforcement away from the wood in the glulam.

Download

Share