CONSTRUCTION STAGE BEHAVIOUR AND LONG TERM MONITORING OF THE PRESTRESSED CONCRETE BRIDGE

Abstract

Prestressed concrete girder bridges form a significant component of the bridge stock of Indian railways. The long term performance of these bridges is a cause of concern to the bridge authorities, and several of these bridges have shown varying degrees of deterioration over the time of their use. The problem is further exacerbated by the immense problem of dismantling and disposal of these girders if they are found to be inadequate. While several tests have been conducted in laboratory conditions, a full-scale real life testing on a bridge intended for mixed traffic of freight and metropolitan transport is relatively rare. With a view to a better understanding of the behaviour of the bridge deck, two prestressed girders have been comprehensively instrumented with fibre Bragg grating (FBG) sensors that were embedded in the bridge during construction. The main aim of the installation scheme is to access the vulnerability level of the PSC girders at any stage during the design life of the sensors. This system is also beneficial to find out the thermal effects during the initial ages of hydration, as sensors were installed before casting. In addition to routine inspections for maintenance purposes, the monitoring system can be used to validate the design assumptions and find out the stress loss during different stages of constructions and service stages. Data has been acquired over the different stages of its construction including curing period, prestressing over different stages, erection stages, casting of the diaphragm, casting of the slab, load test, and service stages. The initial hydration period of concrete is always associated with heat development and related temperature variation in the structure due to the heat of hydration. If the concrete cannot move freely, thermal stresses will generate, and early age cracks may occur when thermal stresses are tensile in nature and exceed the tensile strength of green concrete. Several thermal models exist in the literature; these models often give different results based on the assumptions and changing boundary conditions. In this thesis, various such models are evaluated and calibrated with the existing data acquired during the setting process of two girders, and a model that accurately resembles the site behaviour is selected. The numerical method presented characterise the hydration temperature and stress/strain evolution in the concrete during the first days of post concreting. The formulated thermo-mechanical model is validated with the experimental results. Numerically predicted results from the FE analysis were in good agreement with the monitored data. A maximum temperature difference of 36°C was observed between the core of the end diaphragm with the surface of the girder. The simulation results indicated that tensile stress calculated were mainly located in the haunch portion between the end diaphragm and intermediate diaphragm. While calibrations of heat models have been done extensively in the laboratory, a full-scale test of this nature serves to give realistic estimates of the behaviour of PSC girders in actual site conditions. Erection stresses are rarely considered in the design; in the data analysis, it was found out that not only are they reasonably high but also form a critical load case. In addition, torsional moments developed during erection stages, tensile stresses exceeded the allowable limit recommended by the IRC 18-2000. The results show that erection procedures followed in the field need to be changed. The presented information and recommendation given to avoid the current erection stage issues will help in developing future policies of the practice of bridge construction. The behaviour during the prestressing of one girder over a companion girder in the composite section is studied. Theoretical design values are compared with the monitored values, good agreement with the design values was obtained. A model is proposed to find out the elastic curve of the PSC girder during the service stages; the proposed model is validated from the comprehensive field load test conducted on the bridge. The effect of erection stress and companion girder prestressing on the final service stage of the bridge is studied. Finally, The monitored values of creep and shrinkage strain are compared with the values calculated using EC2, ACI 209R-92, Bazant-Baweja B3, and GL2000 models for a duration of about 2.5 years by fitting experimentally measured and theoretically fitted elastic strain inputs. The ACI 209R-92 model was found to be giving less error during different construction stages as well as service stages. PSC girder is designed as a Type 1 member; however, from the analysis, it was found that tensile stress was developed during service stages, factors accountable for these effects were finally discussed. The periodical monitoring of the bridge throughout construction stages and service stages proves to be an excellent tool for verifying the design models and quality assurance in the construction control.