Research On Key Technologies For Evaluation Of Long-span PC Cable-stayed Bridge

At present, many problems exist on the evaluation of long-span PC(prestressed concrete) cable-stayed bridges. In order to solve these problems, key technologies for evaluation of long-span PC cable-stayed bridges are studied in depth in this thesis. The main content of the research includes five aspects as follows:(1) Load efficiency coefficient decreasing optimization is discussed at the first time. For reducing load efficiency coefficient, the load test conditions of a practical example are designed. The linear regression equations of structural response (stress, strain and cable forces, etc.) and load efficiency coefficient are obtained by regression analysis, and test of significance of the equations is finished. Using the linear regression equations, the structural response can be predicted. Finally, the decreasing optimization of load efficiency coefficient of a long-span PC bridge is presented, and the effectiveness and accuracy of this method are verified. The research results show that the linear regression equations are highly significant and can predict structural response accurately. Generally, the prediction error is decreasing with the increasing of load efficiency. In addition, the load efficiency of long-span PC bridge should be decreased. Furthermore, the load efficiency of long-span PC cable-stayed bridge can be reduced to 0.6.(2) According to the existing problems of load test of cable-stayed bridge, the loading characteristics of different components of cable-stayed bridge are studied. Then, the basic test conditions, test items and secondary test conditions of long-span cable-stayed bridge are determined. Based on the load efficiency coefficient decreasing optimization, the problem of mergence of load test conditions is proposed. Finally, the mergence of load conditions of a practical PC cable-stayed bridge is presented, and the effectiveness and practicability of load test condition mergence are verified. Generally, the condition of maximum deflection of main beam can merge with maximum positive moment condition of main beam; and maximum horizontal displacement condition of tower can merge with maximum positive moment condition of tower.(3) The ranges of verification coefficient aiming at different components are proposed in the first time. Using statistics method of engineering mathematics, the samples of different components'verification coefficients are analyzed with statistics software named Minitab. The range of verification coefficient of main beam, cable and tower are determined. The research results show that the bearing capacity of cable-stayed bridge can be evaluated based on the verification coefficients of main beam, cable and tower; the reasonable range of verification coefficient of main beam strain is 0.60~1.00, the range of main beam deflection is 0.72~1.00; the range of cables is 0.73~1.00; the range of tower stain is 0.18~1.00, the range of tower horizontal displacement is 0.47~1.00. (4) Based on fuzzy theory, a fuzzy safety evaluation model of PC cable-stayed bridge is established, and the safety evaluation of a practical bridge is completed using fuzzy evaluation theory. The results show that the whole safety level of Dezhou bridge is in the first class; cable safety level is in the first class; tower safety level is in the first class; main beam safety level is in the second class; foundation safety level is in the first class. The obtained classes of the structural components are verified with the classes ranged by verification coefficient, and the whole safety evaluation class of the bridge is consistent with the class calculated by JTG H11-2004《Code for Maintenance of Highway Bridges and Culverts》.(5) According to the theory of analytic hierarchy process, a complete durability evaluation model of PC cable-stayed bridge is established. Then, combined with fuzzy theory and neural network theory, a durability evaluation model based on fuzzy neural network is proposed. Finally, taking the cable module as an example, the validity of the model based on fuzzy neural network is verified. The research results show that: the durability evaluation model of long-span PC cable-stayed can be established using analytic hierarchy process method, and the first level evaluating indexes include the durability of main beams, cable, tower, foundation and accessory facilities. Dynamic BP algorithm is prior to common BP algorithm in network's training speed. The trained fuzzy-neural network can store the experts'knowledge and experiences, and can be used in durability evaluation of cable module.