- B. Carroll
- University of Texas at Austin
Heat transfer by convection is a complicated process involving transfer of thermal energy through momentum and thermal gradients. The convection model developed for the VTB makes simplifying assumptions to minimize the solver computation time and to reduce number of independent equations that must be solved at each time step. These assumptions are:
- Control volume with only conservation of energy employed across the boundaries
- Explicit and transient in time and with a lumped element spatial representation; spatial gradients cannot be resolved
- Thermal physical properties are temperature dependent with an option to remove this dependency
- Heat transfer by conduction and radiation are neglected
The model uses two thermal nature ports and is represented schematically by a cooling fin. The port located at the icon’s center is the input node (port 0) and the output node (port 1) located near the bottom is connected to either a temperature source/sink or it can be linked to another thermal port on a separate model. It is not necessary to adhere to this input/output guideline as long as the user understands that all state data follows this convention. For instance, Temperature0 is the temperature of port 0.
The driving force behind developing the convection model was to remove the user burden of specifying the convective coefficient, historically a difficult and computationally intensive task. This procedure is aided through the use of empirical and experimental correlations that have been documented extensively in the literature. Although such correlations have, at best, 10-20% accuracy, they provide basic understanding of how a system may respond given user identifiable parameters such as geometric attributes, free stream velocities, and surface orientation relative to the flow field. Such attributes are conveyed to the user through a custom dialog window. Convection parameters can be accessed and changed before and during run time, providing a parametric approach to system behavior. The complete list of correlations and the steps required to determine the correct correlations are presented in the Appendix.
Use of convective correlations requires intrinsic thermo-physical property data, such as density, specific heat, kinematic viscosity, thermal conductivity, coefficient of thermal expansion, etc. Property values are computed at each time step using curve-fitted equations derived from published tables. If the user chooses to use a specified heat transfer coefficient, thermo-physical properties are no longer required and this solve block is not called by the solver.