Bubble Growth From First Principals

We consider the simplest relevant problem in the foaming of molten plastics, the growth of a single bubble in a sea of highly viscous Newtonian fluid, and without interference from other bubbles. This simplest problem has defied accurate solution from first principles. Despite plenty of research on foaming, classical approaches from first principles have neglected the temperature rise in the surrounding fluid, and we find that this oversimplification greatly accelerates bubble growth prediction. We use a transport phenomena approach to analyze the growth of a solitary bubble, expanding under its own pressure. We consider a bubble of ideal gas growing without the accelerating contribution from mass transfer into the bubble. We explore the roles of viscous forces, fluid inertia, and viscous dissipation. We find that bubble growth depends upon the nucleus radius and nucleus pressure. We begin with a detailed examination of the classical approaches (thermodynamics without viscous heating). Our failure to fit experimental data with these classical approaches, sets up the second part of our paper, a novel exploration of the essential decelerating role of viscous heating. We explore both isothermal and adiabatic bubble expansion, and also the decelerating role of surface tension. The adiabatic analysis accounts for the slight deceleration due to the cooling of the expanding gas, which depends on gas polyatomicity. We also explore the pressure profile, and the components of the extra stress tensor, in the fluid surrounding the growing bubble. These stresses can eventually be frozen into foamed plastics. We find that our new theory compares well with measured bubble behavior.

Heat Transfer in an Airfoil Shaped Strut

The heat transfer from a hot primary flow stream passing over the outside of an airfoil shaped strut to a cool secondary flow stream passing through the inside of that strut was studied experimentally and numerically. The results showed that the heat transfer on the inside of the strut could be reliably modeled as a developing flow and described using a power law model. The heat transfer on the outside of the strut was complicated by flow separation and stall on the suction side of the strut at high angles of attack. This separation was quite sensitive to the condition of the turbulence in the flow passing over the strut, with the size of the separated wake changing significantly as the mean magnitude and levels of anisotropy were varied. The point of first stall moved by as much as 15% of the chord, while average heat transfer levels changed by 2-5% as the inlet condition was varied. This dependence on inlet conditions meant that comparisons between experiment and steady RANS based CFD were quite poor. Differences between the CFD and experiment were attributed to anisotropic and unsteady effects. The coupling between the two flows was shown to be quite low - that is to say, heat transfer coefficients on both the inner and outer surfaces of the strut were relatively unaffected by the temperature of the strut, and it was possible to predict the temperature on the strut surface quite reliably using heat transfer data from decoupled tests, especially for CFD simulations.