ATMOSPHERIC DISPERSION MODELLING USING SMOOTHED PARTICLE HYDRODYNAMICS METHOD
In the military field, the pollutants dispersion phenomenon in the atmosphere is used as the basis for the development of NBC (Nuclear, Biology, Chemical) weapons detection sensor. Atmospheric dispersion is a combination of diffusion due to turbulent eddies and advection due to wind, which is influe...
Saved in:
Main Author: | |
---|---|
Format: | Theses |
Language: | Indonesia |
Subjects: | |
Online Access: | https://digilib.itb.ac.id/gdl/view/52038 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Institut Teknologi Bandung |
Language: | Indonesia |
Summary: | In the military field, the pollutants dispersion phenomenon in the atmosphere is used as the basis for the development of NBC (Nuclear, Biology, Chemical) weapons detection sensor. Atmospheric dispersion is a combination of diffusion due to turbulent eddies and advection due to wind, which is influenced by many meteorological factors. This phenomenon usually occurs in large domains consisting of flow in the presence of temperature differences and obstacles. The difference in temperature causes the buoyancy force to appear which can affect the flow motion. The time average concentration profile in atmospheric dispersion forms a Gaussian profile. With the presence of atmospheric dispersion, measurement of air quality at any given time and place is important but difficult to be done. As a solution, various types of dispersion modeling have been developed. In this thesis, the dispersion is modeled using Computational Fluid Dynamics as a single-phase fluid using the Smoothed Particle Hydrodynamics method. SPH is a lagrangian mesh-free method that replaces the use of a grid with a bunch of particles. For each particle, the property approximation at each time step is influenced by the cluster of particles around it. Each particle evolves with time based on a governing equation. In this thesis, the linked-list algorithm (neighboring particle searching method) is applied to reduce the long computational time due to large domains. The buoyancy force is modeled using the SPH approximation of the energy equation in terms of temperature and the Boussinesq approximation. The obstacles in the flow is modeled by solid wall boundary condition, while the wind is modeled by the inflow/outflow boundary condition. To overcome numerical problems in the SPH method, namely tensile instabilities, particle shifting technique is used. Parameters that influence modeling accuracy are particle resolution, domain size, and smoothing length. The SPH algorithm is then applied to model elevated single and multiple line source cases in a uniform flow and to model cases with atmospheric boundary layer and buoyancy. The concentration properties obtained were then compared with the exact and FVM solution. The SPH modeling carried out resulted in sufficiently accurate concentrations for all cases, except for cases where buoyancy was present. |
---|