Research On Bidirectional Air Inter-Exchage Longitudinal Ventilation In Road Tunnels

With rapid development of construction of road tunnels, there has being a large number of long and complicated road tunnels nowadays, and the design of tunnel ventilation and control system is increasingly difficult. Based on the Jinping construct-assistance tunnel, the paper detailedly researches on tunnel ventilation for super long road tunnels with twin tubes, programs the road tunnel ventilation system, and carried on experiments in a model tunnel.For road tunnels within small vehicle flux and large contamination per vehicle, "line-source vehicle model" cannot truly and accurately simulate the distribution of contamination in tunnels, it proposes the "point-source vehicle model", and founds the numerical simulation model for the ventilation of long road tunnels with twin tubes. Base on the theory of network ventilation and emergency ventilation in tunnel fires, it sets up the numerical simulation model for network ventilation and emergency ventilation, educes their specific solutions, and designs a program for normal and emergency ventilation in road tunnels.Due to terrain and vehicle density, the ventilation rates of twin tubes for fulfilling the requirement for temperature and pollution concentration is asymmetric. For saving the cost of ventilation system, it does detailed research on the application of air inter-exchange in load-asymmetric road tunnels, proposes the specific formulae and approaches, and uses the technique to simulate the ventilation system of Jinping construct-assistance tunnel.As the vehicle flux varies, the ventilation control system can be controlled efficiently by turning on or off a part of fans, respectively corresponding to increasing or decreasing vehicle flux. It establishes the neural network on-line control model for single-line road tunnel using bidirectional air inter-exchange ventilation system, including specific procedures and programs, based on the theory of tunnel ventilation network and of the neural network on-line control.The resistance characteristics of Jinping construct-assistance tunnel is large-coarseness and nonuniform-roughness. And air flow in emergency stop location is very complicated. The flow resistance is mostly made up of friction resistance and local resistance. According to similitude theory, it sets up the scale model test-bed for testing the friction resistance and the local resistance, and carried out experiments under different Reynold Number conditions. The friction resistance coefficient ,λ, both Tunnel A and Tunnel B, is 0.04. The positive and passive local resistance coefficients for tunnel B are respectively 0.349 and 0.336. Then set up another test-bed for testing the pressure rise of impulse fans, and confirm the correctness of CFD modeling and numerical method.It analyzes the distribution of pollution concentration in tunnels under normal traffic, blocked traffic and jam conditions, and simulates the time beyond the limited pollution concentration for both blocked traffic and jam. After simulating the distribution of pressure in both tunnel A and tunnel B with different collocation of fans, it proposes the proper collocation of fans for these tunnels. Then uses the fire module to simulate the fire scenario of heavy good vehicle (HGV) in three different places, presents the distribution of air temperature, wall temperature, and smoke concentration in both tunnel A and tunnel B. In addition, it simulates the impacts of high smoke temperature on the impulse fans, and proposes some constructive suggestion, based on the numerical results.It simulates the ventilation system of Jinping construct-assistance tunnel under normal traffic, blocked traffic and jam conditions, using the tunnel ventilation control program, to investigate the effect of the neural network on-line control system on ventilation in road tunnels. The results shows that the control system can modify the control strategy and turn on fans appropriately to prevent excess of the pollution concentration. And confirm the robustness of the designed neural network on-line control system applied to single-line road tunnel using bidirectional air exchange ventilation.