Research On Fire Temperature Distribution And Intelligent Prediction Of Dense Cross Beam Of Double Deck Suspension Bridge

As the number of bridges increases,the safety risks of bridges also increase.Fire is one of the important risks faced by bridges.As bridges are mainly composed of steel,the high temperature of fire can have a significant impact on the mechanical properties of bridges,posing a threat to the safety of bridge structures.For the confined space of the lower lane of the double-layer suspension bridge,after the fire,the heat dissipation is limited,and the ambient temperature will rise rapidly,which will directly affect the safety of the dense beam structure.In serious cases,the bridge structure may be partially damaged or collapse as a whole,leading to huge public property losses and heavy casualties.Therefore,it is of great significance to study the fire temperature distribution of dense beams in some confined spaces of double-layer suspension bridge under the influence of fire.In this thesis,the temperature field of dense beam of double deck suspension bridge is obtained by carrying out fire scale experiment and numerical simulation research of double deck suspension bridge;The temperature distribution law of the dense beam under the coupling effect of different ambient wind speeds and the spacing between the dense beam and the diaphragm in the tank car fire is revealed;A prediction model for the maximum temperature rise distribution of a dense crossbeam fire was established,taking into account the coupling of environmental wind speed and the spacing between diaphragms;Furthermore,the reliability of the numerical simulation results was verified by comparing and analyzing them with the scaled experimental results;On this basis,the method of combining numerical simulation with neural network algorithm is used to intelligently predict the temperature distribution of the dense beam.By comparing and analyzing the prediction performance of different prediction models,the optimal prediction model of the temperature distribution of the dense beam of the bridge is selected.The main achievements are as follows:1.When the spacing between the diaphragm plates of the dense beam of the double-layer suspension bridge is constant,the maximum temperature rise under the ceiling gradually decreases with the gradual increase of the ambient wind speed.When the characteristic wind speed is less than 0.19,the maximum temperature rise of the ceiling decreases with the decrease of the spacing between the transverse partitions,and the spacing between the transverse partitions of the dense crossbeam has a significant impact on the maximum temperature rise;When the characteristic wind speed is greater than 0.19,the influence of the spacing between the dense crossbeam and the transverse partition on the maximum temperature rise gradually decreases,and the environmental wind plays a dominant role.Further revealed the mechanism of the influence of environmental wind speed and spacing between diaphragms on the maximum temperature rise of the dense beam.2.Based on dimensional analysis,a theoretical prediction model for the dimensionless maximum temperature rise of the dense cross beam coupled with the dimensionless fire source power,external wind speed and diaphragm spacing was established.3.Compare and analyze the numerical simulation results with the scaled experimental results,and the error of the numerical simulation results is within 20%;Based on the theoretical model derived earlier,the dimensionless maximum temperature rise of the dense beam was predicted,and the predicted results were in good agreement with the numerical simulation results,verifying the reliability of the numerical simulation and laying the foundation for neural network data training.4.GRU(Gated Cyclic Unit)has good applicability and stability in the prediction of the temperature time series data of the dense beam fire,followed by LSTM(Long Short Term Memory Network)and RNN(Cyclic Neural Network).Therefore,the GRU model can be preferred for the prediction of the temperature time series data of the dense beam in some confined spaces of the double-layer suspension bridge.The network was used to predict the multi-point temperature of a dense crossbeam fire under different working conditions,and the predicted values were in good agreement with the actual values.The research in this thesis is of great significance to the formulation and improvement of fire codes for double-layer suspension bridge,fire risk assessment,fire prevention and control,and post disaster fire accident investigation.This thesis consists of 58 figures,13 tables,and 94 references.