Research On Long Term Performance Of GFRP Bars Embedded In Concrete

In the rapid development of construction industry in the past100years, the reinforced concrete structure has been widely used and most of industrial, civil and other requirements have been meted. However the problem of insufficient tension member’s durability of steel-reinforced is exposed as well. Therefore, to meet the corrosion resistance of atmospheric attack, GFRP bars received extensive attention. Long term performance of new materials is the focus of research as security is particularly important for the industry, civil structures. For the status of little practical application and research data of long term performance of GFRP bars, mechanism of aging of GFRP and the impact of environmental factors on mechanical properties were analyzed. Besides, based on the theory of Fick’s Law a new Long-term durability prediction model with random error is created.Accelerated aging test is used in most long term durability studies of GFRP bars. Specimens are exposed under the condition which is hasher than the real environment. To amend the experimental results, the study on the mechanism of aging is necessary. Physical corrosion, atmospheric corrosion, damp corrosion is analyzed to determine the reasons for the decrease of durability of the material.GFRP bars will be subject to a variety of environmental and human impact in service which will produce continuing Negative effects. In this paper, the effects are analyzed on the basis of experimental data. Important performance indicators such as glass transition temperature, tensile strength, bond strength under the action of moisture, temperature, alkaline ion and saline water are discussed. The effects of the protective layer of concrete are also analyzed. On the basis of the above study and the comparison of the two models, Arrhenius Equation and Fick’s Law, a new Long-term durability prediction model with random error is created to further improve the accuracy of predictions. In the model building process, Bayesian updating theory and WinBUGS are used to determine the unknown parameters. Compared with the long term performance experimental data, the model achieved good results.