Master Thesis - Towards a Virtual Climate Chamber : A numerical study using CFD software

Abstract: For each generation of electronic equipment there is a trend towards higher power den-sities. Increased heat generation is an undesired consequence that the thermal design unit in a company must handle. The goal of thermal design engineer/unit is to utilizethe same volume to more efficiently transfer more heat from the equipment. This can bedone by exploring more complex and advanced heat sink geometries, optimizing the finshapes and so on. The new prototypes developed will be tested for their reliability and endurance in special chambers called climate chambers, that simulate desired environ-ments. The measurements by thermal design teams in these kind of climate chambers are mainly of outdoor products, whose cooling is based on natural convection. Forcedcooling using fans is optional for these outdoor products. The climate chambers in general provides temperature measurement as the outputto the analysis, though there are other important parameters that define the operationalfunctionality of an equipment. The ability to visualize the flow characteristics duringthe process of testing is a valuable aid in the design process. A virtual/CFD form of thephysical climate chamber (CC) would empower the design process, while alleviating theusage of the climate chambers for such analyses. CFD offers a wide range of capabilitiesthat lets the user change the boundary conditions with great ease compared to that ofthe experimental setup. The numerical model developed in this thesis project provides results, that help inunderstanding the physics involved in fluid flow inside the physical climate chamber.Turbulence quantification of the flow is the main aim of this thesis project, which wouldbe resourceful in future works. Experiments are conducted inside the climate chamber, in order to aid the construction of numerical model as well as serve as source of vali-dation for the numerical results. Laminar transient case simulations are preferred over use of any turbulence models, to limit any kind of predictions made by these turbulencemodels. Integral length scales and turbulence intensities are compared and reason fordiscrepancies are addressed. The results from the comparisons show that, the numerical model emulates physicsof actual flow inside the climate chamber. However, there are many factors that directlyaffect the results, making it difficult to precisely quantify the error, within the time periodof this thesis project.