Research On Air-flow Organization In Thermal Channel Of Glass Curtain Wall

Energy-saving is a very serious common issue during the course of human development, building energy conservation is an important component of it. The building energy conservation is not only includes the construction energy-saving, but also including the energy-saving during the course of use, a reasonable glass curtain wall contribute to the building energy-saving. Therefore, the study of the glass curtain wall is great significant for energy-saving.This paper consists of two part, ANSYS is employed to simulate and analyze the air-flow in the thermal channel of glass curtain wall to ensure whether the glass curtain wall can achieve"thermal chimney effect"in summer in the first part. Firstly, the mathematical model of the glass curtain wall is established, the common summer weather conditions as the boundary conditions loaded on the model. Through simulation and analysis we found that the"thermal chimney effect"can not be achieved. This shows the energy-saving can not be achieved. The various speed (2.0m/s,2.5m/s,3.0m/s,3.5m/s,4.0m/s,4.5m/s,5.0m/s)is loaded on the nether vent , through comprehensive analysis, the proper mechanical ventilation speed is 4.5m/s. By calculation, the fan motor power is 861W and energy-saving is 84%. In this paper, the different-sized glass curtain wall are simulated and analyzed such as wall spacing are 0.2m,0.4m,0.6,0.8m,1.0m,1.2m,1.4m,1.6m,2.0m,2.2m,2.5m respectively and vent size are 0.1m,0.15m,0.2m,0.25m,0.3m,0.35m,0.4m,0.45m,0.5m respectively. Finally, the economy of representative glass curtain wall is calculated and the energy-saving is 30%. In order to examine the model established in this paper correct or not ,the paper cited experimental date published in the literature. Comparison of simulation and experimental data, the greatest mean absolute error is 0.0681 which proves the model and research methods are accurate.In this paper, the fluid simulation model is standards k ?εturbulent model, fluid is incompressible fluid.In the second part, the author use neural network to fit the relation between the glass curtain wall size and the parameters on the nether vent. In these ANN models, three-layer Perceptron (BP network) is always employed as the universal approximator to build the relationship between the input and output parameters, log-sig and Pure linear are employed as a hidden layer and output layer transfer function. Compared with the simulated average speed data, the average, mean and standard deviations of the trained neural network are 0.3312%, 4.526% and 7.4654% respectively, and 100% of the deviations are within±15%, and approximately 82.5% of the deviations are within±10%. Compared with the simulated average temperature data, the average, mean and standard deviations of the trained neural network are 0.00018%, 0.1023% and 0.13987% respectively, and 100% of the deviations are within±5%. Finally, the glass curtain wall neural network correlation is achieved.If we input parameters in the correlation, the average temperature and the average velocity on the nether vent can be obtained directly.Several results are achieved in this paper, such as simulating the thermal channel glass curtain wall and identifying the defects, simulating the various size glass curtain wall, energy-saving calculation, experimental verification, and the neural network correlation.In this paper, the model of the glass curtain wall has higher precision and good versatility.The research result can guide the construction of the thermal channel glass curtain wall. In addition, scientific research workers can refer to the research methods in this paper.