Klimatsimulering av ett kontorsrum
Hur de termiska stigkrafterna påverkar ventilationseffektiviteten
AbstractIn urban societies people spend more and more time indoors, which put great demands on the indoor environment. If the ventilation is to manage the removal of unwanted substances in the indoor air, a good mixing of the air is required. The question arises that, whether the buoyancy forces will counteract a good mixing of the indoor air or not?This report aims to clear up the difficulties whether you need to consider the buoyancy forces or not, when adjusting a ventilation unit. For studying the indoor mixing of the air a two dimensional model is constructed in COMSOL Multiphysics. In this model several supply air temperatures is simulated to see how much the temperature affects the mixing of the air.The physical relations that are being used in this model is the Navier-Stoke?s equation for Non-Isothermal flow, the relations for General Heat transfer and the relation for Convection and Diffusion. These relations exist as tools in COMSOL Multiphysics.It wasn?t possible to get the model to converge at low air velocities, because the model is dependent of accuracy in the field of cm or mm.The lowest velocity, in the zone of occupancy, which the model was able to simulate was 0.5 m/s but it takes velocities below 0.2 m/s to fulfill the indoor climate requirements due to the problems concerning draft.It was possible to see that the buoyancy forces affect the mixing of the air to a certain degree, despite high air velocities.Unfortunately it wasn?t possible to see to what degree you have to consider the buoyancy forces when adjusting a ventilation unit. To see the affects of the buoyancy forces or not, is highly dependent on the velocity of the air. Because the force generated by the supply air increases by the square of the velocity change.When constructing a more detailed model, which is a requirement for convergence when having low velocities, the memory of the computer isn?t enough.The results of the simulations show that a computer with more calculation capacity, than the Karlstad University can offer, is needed to make a successful simulation.