Modelisation of Convective Heat Transfer Using Novel Adaptive Parametric Models and Novel Empirical Models Obtained From Measured Surface Temperatures

In most real-life applications, thermal modeling must be done in outdoor, extremely varying, weather conditions, complexifying this thermal modeling. One of the phenomena significantly influencing the predicted temperatures is heat convection. This paper discusses heat convection modeling. Two classes of models exist. The first numerically solves the fluid motion equations and the second designs empirical models from measurements. However, numerical models have prohibitive computation time. Empirical models are faster and simpler. However, they are often designed for dedicated object dimensions in well-controlled conditions (wind tunnels). To our knowledge, simple, fast models that can be adapted to any object geometry in outdoor, extremely varying environments do not exist. We then define a new class of models: adaptive convection models. These models include parameters that can be tuned to any geometry, environment, and fluid properties. This paper proposes two adaptive models. We also propose a new empirical model deduced from measurements obtained from a plane plate. We finally compare all these models with existing convection models using measured temperatures at different geographical locations and for different objects.