Research On Aerosol Particle Deposition Characteristics In Ventilation Duct Systems

Owing to the worsening of outdoor and indoor air pollution nowadays, indoor air quality are attracting more and more attentions from government and individual persons. The distribution of aerosol particulate, one of the most important pollutants deteriorating building environment and affecting faculties'health, is crucial to the evaluation of exposure concentrations and quantitative effects on the people. Past researches on the movement and distribution of particles indoors generally neglect particle deposition in ventilation duct systems, which is an important sink of aerosol particulates. Basing on investigation of a typical building's air-conditoning system, this dissertation focusing on theoretical analysis, experimental testing, and numerical simulation of particle depostion rules in ventilation ducts. The main works and relevant results in the paper include:A. Afer summary of the contents of large amounts of references, topics like aerosol particulate pollution and its health hazards, the influence of HVAC systems on indoor air quality (IAQ), the distribution of particulates in HVAC systems, controlling or decreasing the effects of particulate matters'pollution are outlined. Having defined some basic parameters of aerosol particulate, the movement of particles in the vicinity of wall boundary and interaction of particles and the wall are elaborated.B. An analytical prediction program, particle deposition analytical prediction model (PDAPM), is written by simplifying and modifying former researchers'models. After comparing the outputs of the PDAPM code and recent experimental data, this program is able to predict the particle deposition velocities in fully developed turbulent flows in ventilation ducts perfectly. Conclusions from the analytical prediction model include: (1) the most important region in the vicinity of the wall deciding particle depostion velocities is turbulent parameters in the y+≤10 zone, namely in the turbulent sublayer and buffer layer; (2) wall roughness may elevate particle deposition velocity dramatically, most evident to medium size particles; (3) in the same predefined conditions, dimensionless deposition velocity of a definite diameter particle increases by the elevation of air velocity. Evidently, these results improve our understanding on particle deposition rules in fully developed turbulent flows. C. A public building in Changsha was investigated about the particle concentration distribution and particle deposition on ventilation duct floor in its central air-conditioning duct systems. Due to serious indoor pollution sources like smoking and the short distance from return air inlets to pollution sources, particle concentrations and deposition masses in return air ducts are bigger than those in supply air ducts, and particle concentrations decay along flow path for the reason of particle depostition. Particle concentrations and particle deposition masses are largger in winter than those in summer owing to the more serious pollution in Changsha's winter. Generally speaking, main factors influencing particle deposition masses on ventilation duct floors include air velocity, pollution source intensity, distances from the pollution sources, and so on.D. In order to improve understanding on particle deposition on different surfaces of galvanized steel ventilation ducts having different width height ratio, we did relatively systematic experiments utilizing an aerosol generator designed by ourself and two kinds of nearly monodisperse talcum powders with different average diameters. From the experiment results in the nearly developed turbulent flow region of experimental duct systems and their comparison with the prediction results of PDAPM code, calculated experimental deposition velocities fitts well with those prediction results, and the experimental data collected by weaveless cloths and aluminum foils directly are reliable and repeatable. The analysis of particle depositon on walls of a square duct after a 90 degree bend finds that secondary flow built up in the bend have significant effect on particle deposition velocity on the wall, especially near the bend exit region. Empirical equations which is convenient to apply in engineering are fitted to experimental data achieved by weaveless cloths and aluminum foils referring to the principles of particle deposition.E. After introducing recent advancement on nonlinear anisotropic two-equation turbulent models contrary to traditional linear isotropic eddy viscosity models, a simplified nonlinear k-ωtwo-equation turbulent model is deduced. Then the simulation of fully-developed turbulent flow in a rough square duct testifies the existence of secondary flow and rough wall law. Further simulations on developing turbulent flow in a long square duct using linear renormalization group (RNG) k-εand nonlinear k-ωtwo-equation turbulent models reveal that both models can catch the development of centerline mean velocity along the flow path. However, the nonlinear turbulent model's simulation results coincide better with direct numerical simulation results than the linear turbulent model's regarding to turbulent velocity perpendicular to the wall. According to mean and turbulent velocities in and after the 90°square bend, simulation results of nonlinear k-ωtwo-equation turbulent model are also closer to experimental data than the linear RNG k-εmodel.F. Based on the above flow fields of long square ducts and 90°square bend, particle distribution and movement ruels are simulated by a Lagrangian stochastic trajectory model. As regard to the simulated particle deposition velocities in the nearly developed region of the long duct and its comparison to typical experimental results, particle deposition velocities in nonlinear k-ωtwo-equation turbulent model's flow are obviously closer to the experimental data than in linear RNG k-εmodel's, especially to smaller diameter particles. Particle deposition velocities decay along flow pass from inlet to outlet, which is the most prominent to medium diameter particles. To the simulation results in and after the 90°square bend using nonlinear k-ωtwo-equation turbulent model, (1) particle penetrations in the bend are in accord with results from a typical empirical equation; (2) simulated deposition velocities on the walls after the bend catch the trend of our experimental data, and illustrate the important effect of secondary flow near the bend outlet.