Study On Nuclear Accident Aerosol Dispersion Model On Sea Surface And Dry Deposition Experiment

In the early stage after a nuclear accident,the source of the accident must be accurately obtained to provide scientific support for the evaluation of the consequences of the accident,thereby ensuring the health of the public and conducting reasonable emergency actions.The accuracy of the radionuclide atmospheric dispersion model is one of the most important factors affecting the source term estimation.The atmospheric dispersion models currently proposed are mostly based on the diffusion of nuclide on the underlying land surface.There is still a large difference between the its physical assumptions and the actual situation,because of the influence of various factors such as diffusing materials,deposition types,underlying surface types,etc.This may lead to errors of at least one order of magnitude in the results of source term inversion in other cases.In particular,island or offshore platform nuclear power plants generally face the ocean underlying surface,and the nuclide aerosol leaking from the containment shell or the chimney after an accident will spread to the land through the ocean.The ocean underlying surface is constantly moving,and the waves actively capture aerosols.As a result,its deposition rate is quite different from that on land,which increases the error of the diffusion model.Therefore,the dispersion model needs to be modified according to the aerosol deposition rate on the sea surface to improve the accuracy of the dispersion model and inversion model.For the problems above,this paper conducts theoretical and experimental research on the dispersion and the deposition of aerosol on the sea surface.Considering the capture of aerosols on the sea surface,this paper modified a Gaussian plume model based on the aerosol deposition model on the sea surface,and proposed a "Gauss-sink" coupling model.Studies show that the change trend of this model with increasing distance in the downwind direction is similar to the original model,and both follow the basic form of the Gaussian plume model.The peak value of the concentration is about1/3 of the original model,and the rate of increase and decrease of the concentration is also slow.The aerosol concentration of the modified model is much smaller than that of the original model near the sea surface.In order to verify the "Gauss-sink" coupling model proposed in this paper,an environmental wind tunnel experiment is needed.Based on the fan wind tunnels in Shanghai Jiaotong University,this paper establishes a multi-fan controlled wind tunnel control algorithm to improve the accuracy of the wind tunnel simulation flow field.Considering the influence of the wind tunnel boundary on the flow field and the influence between adjacent fan generated flow fields,a fast and accurate method for fitting and predicting wind speed profiles based on convolutional neural networks is proposed.The experiments find that the method increases the wind speed fitting speed to the minute level,and reduces the relative mean absolute error between the predicted wind speed profile and the actual measured wind speed profile in the wind tunnel to 7%,which can greatly improve the accuracy and reliability of the experiment.In order to measure the concentration of aerosol puff moving in the wind tunnel,this paper proposes a high-speed camera tracking device to improve the reliability of the measurement.The high-speed camera has a small frame,and the experimental aerosol plume moves quickly and easily moves outside the frame,which makes it difficult to capture its dispersion process in experiment.Therefore,in this paper,the correlation filter tracking and target detection algorithms are used to track the puff in real time with a mobile high-speed camera.Computer simulation and physical experiments were used to verify its feasibility.The results show that when the threshold of overlap score is 0.6,the tracking success rate of this method reaches0.7,which can keep the tracked puff within the camera frame and meet the requirements of high-speed image acquisition.Finally,based on the improved wind tunnel control method and measurement method,an atmospheric dispersion experiment was performed to verify the feasibility of the "Gauss-sink" model proposed in this paper.A comparative analysis of the dispersion on the solid surface,the dispersion on the calm water surface,and the dispersion on the wavy water surface has been performed.Experiments show that the rate of aerosol deposition on the simulated sea surface is higher than that on the solid underlying surface,and its dispersion model is consistent with the modified model proposed in this paper,which proves the effectiveness of the model.