Calculation of Particle Resuspension Using Static and Dynamic Particle Models
MetadataVis full innførsel
The aim of this master work was to investigate particle resuspension from fouled heatexchanger surfaces. In metallurgical industries, the waste gas may contain dust and particles which can deposit on the heat exchanger surface. Particulate fouling leads to a reduction in the heat transfer, which may be compensated by external heating or cooling or an increase in the heat exchanger surface. Both compensations lead to increased costs.A literature study on particle resuspension was carried out. The simplest resuspensionmodels are based on static force and moment balances. The relevant forces acting on aparticle are the aerodynamic forces and the adhesion forces. Rotation is the dominantmotion leading to particles detaching from a surface and it was assumed that a detachedparticle would resuspend. Kinetic resuspension models were investigated. The kineticmodels are classified as quasi‐static ‐ or force and moment balance models and dynamic or energy accumulation models. The quasi‐static models are based on similar principles as the static models, but include the effects of turbulence and surface roughness, through statistical distributions. The energy accumulation models are derived for a turbulent energy transfer to the particle, keeping it in constant vibration on the wall. These models allow for resuspension by aerodynamic forces which are smaller in magnitude than the adhesion forces. This is a result of resonant energy transfer, but calculations and experiments have shown that the resonance contribution can be neglected. For no resonance, models based on a quasi‐static balance of aerodynamic‐ and adhesive moments, describe resuspension accurately.Three models were implemented in the computer code MATLAB, a static force and momentbalance model and two kinetic models, the quasi‐static Rock n Roll model and the VZFG,energy accumulation model. The models have been used to calculate particle resuspensionfrom smooth and rough surfaces. The static model confirmed that rotation is the dominantmotion. The calculations with the kinetic models were used to discuss the influence ofsurface roughness, turbulence, particle size and flow exposure time.The two most uncertain parameters, turbulence and surface roughness were investigatedseparately. Correlations for the adhesion force for rough surface contact were implemented, and it was suggested to implement results from numerical studies of turbulent flow in the models.A brief section of experimental work is presented. The flow through a test section for fouling studies was investigated. The experimental studies are currently being initiated and are suggested as a work of further studies.