Abstract
The objective of this thesis was to determine which RAN’s turbulent model and mesh inflation settings produce results within a determined 5% experimental accuracy while allowing for at least 30 design iterations per week, with respect to a conjugate heat transfer CFD analysis on a Solid State RF amplifying circuit using ANSYS CFX. The RAN’s k-ε, k-ω, and SST turbulent models were selected to use during the simulations. Specific target Y-Plus values established the mesh inflation settings. Simulation industrial time analysis variables of solver process time and Computer RAM requirements determined industrial efficiency of the simulations. A test Solid State RF amplifying circuit was built, mounted, and cooled on an aluminum finned heat sink. Experimental pressure drop, air volumetric flow rate, and thermocouple temperature values in specific zones were measured and recorded to determine the accuracy of each simulation. When using a RAN’s turbulent model and mesh inflation settings, while performing a conjugate heat transfer CFD analysis on a Solid State RF amplifying circuit, it was determined that the most efficient configuration producing accurate results was the simulation using the SST model with a 0.6 Y-Plus value. This configuration calculated a 3.40% pressure drop error, a 3.18% individual thermocouple location temperature error, an average 2.58% couple zone temperature error, solved in 40.83 minutes, and demanded 11.3GB of RAM.