Experimental and numerical study of free convective heat transfer in air performed simultaneously from the upper and lower surfaces of a horizontal plate heated on both sides

Abstract

This paper presents an experimental study, supported by numerical analysis, on convective heat transfer in air from a newly designed, horizontal, double-sided heated Type II plate. Unlike the single-sided heated Type I plate, where the heating surface faces either upwards or downwards, the Type II plate heats both surfaces simultaneously.

The tested plate was constructed from three glued laminates, each 0.6 mm thick, resulting in a total thickness of 2.1 mm. Minimal lateral-surface heat loss (<4.2%) ensured high measurement accuracy. Surface resistance thermometers (T_low/up) were etched into the copper layer on one side, while two resistance heaters (N_low/up) were embedded in the copper-coated middle laminate.

Numerical calculations (Num.II and Num.I) for Type II and I plates revealed discrepancies under the UWT (Uniform Wall Temperature) condition: C_Exp._II/C_Num.II = 33.8% (upper) and -8.4% (lower), and for UHF (Uniform Heat Flux): 17.8% (upper) and 6.1% (lower). Similar discrepancies in C_Num.II/ C_Num.I (-36.1% UWT_up, 0.5% UWT_low, -26.9% UHF_up, 0.7% UHF_low) confirmed different heat transfer mechanisms on the upper heating surfaces of Type II and I plates.

In the absence of the mean literature correlation (MLC.II) for Type II plates, numerical verification (C_Num.II) was conducted, yielding C_Exp./C_Num.II = 1.338 (upper) and 0.916 (lower) for UWT, and 1.178 (upper) and 1.061 (lower) for UHF. Indirect verification using Type I plate data (MLC.I) required numerical transformation, yielding C_Exp.II/C_MLC.I = 1.426 (upper) and 0.748 (lower) for UWT, and 1.017 (upper) and 0.896 (lower) for UHF. These results confirm the reliability of the Type II plate tests.