The Local Natural Convection Heat Transfer Coefficient On A Heated Horizontal Cylinder Oscillating In Water PDF Download

An experimental study has been made of the local natural convective heat transfer coefficient around the circumference of a heated horizontal cylinder oscillating vertically in water. The heat transfer surface consisted of a 1 3/8-inch diameter cylinder with a small test section imbedded in its surface. This enabled data to be taken so that the local and overall values of the heat; transfer coefficient could be determined. The cylinder was oscillated sinusoidally in a tank of distilled water at a frequency of 0 to 25-cps with an amplitude of 0 to 0.100-inch. The temperature difference between the water bath and the test cylinder was held at approximately twenty degrees. Observations of the flow patterns around the cylinder were made using a shadowgraph technique and a dye stream visualization, The local heat transfer coefficient versus position data were taken at six different conditions of frequency and amplitude. These conditions were: (1) stationary, (2) n = 500 rpm, a = 0.100-inch, (3) n = 750 rpm, a = 0.0667-inch, (4) n = 1000 rpm, a = 0.100-inch, (5) n = 1500 rpm, a = 0.0667-inch, and (6) n = 1500 rpm, a = 0.100-inch. The overall cylinder results were similar to the results found by V.H. Swanson and by Martinelli and Boelter in similar work. The maximum increase in the overall cylinder heat transfer rate was of the order of 200 percent. The data for the local heat transfer coefficient showed that the maximum increase in the heat transfer coefficient occurred at the top of the cylinder and was on the order of 290 percent. At the same condition of oscillation the coefficient at the side increased 230 percent while the coefficient at the bottom increased 72 percent. In comparing the shapes of the distributions of local Nusselts number with the shapes Fand, Roos, Cheng, and Kaye found by imposing a sound field on a air-cylinder system, a difference was noted which can be attributed to the difference in the direction of oscillation between the two investigations. In the present investigation the cylinder was oscillated vertically while Fand, Roos, Cheng, and Kaye used a horizontal oscillation of the fluid particles. The resulting differences in the acoustic streaming pattern account for the differences noted in the shapes of the local heat transfer coefficient versus position curves. The shapes did show that the effect of mechanical oscillation and the effect of a sound field on the convective heat transfer rate were similar. A dye stream visualization of the flow pattern indicated Fand, Roos, Cheng, and Kaye were correct when they concluded that the shape of the distribution of Nusselt number was caused by the interaction of a natural convection flow pattern and acoustic streaming. This study sheds some light on the mechanism causing the increase in the natural convection heat transfer coefficient when oscillation is introduced, and it shows the need for more experimental investigation into the distribution of the local heat transfer coefficient around cylinders.