Journal of Heat Transfer · 1999 · 124 citations · 16 references
EngineeringChemical EngineeringMechanical EngineeringHeat TransferThermodynamicsTransport PhenomenaHeat ExchangeHeat ExchangersHeat FlowHeat Transfer AnalysisHeat Transfer In FluidsHeat Transfer ModelingFin Spacing EffectsFinned-tube Heat ExchangersFinsFins Heat TransferFriction Factor CorrelationsHeat Exchanger PerformanceHeat Exchanger Performance AnalysisHeat Transfer Coefficient Correlations
Plain fin‑and‑tube heat exchangers with staggered tube arrays are the focus of this study. The authors aim to develop air‑side heat‑transfer coefficient and friction‑factor correlations that depend on Reynolds number and key geometric parameters such as tube diameter, pitch, and fin spacing. A multiple‑regression analysis of 47 experimental data sets was employed to formulate the correlations. The correlations perform well for small‑diameter or high pitch‑to‑diameter ratio exchangers, with the heat‑transfer model for ≥3 rows and for 1–2 rows predicting 94 % of the data within ±20 % and the friction model predicting 90 % within ±20 %.
This paper deals with heat exchangers having plain fins on a staggered array of circular tubes. Correlations are developed to predict the air-side heat transfer coefficient and friction factor as a function of the Reynolds number and geometric variables of the heat exchanger such as tube diameter, tube pitch, fin spacing, etc. A multiple regression technique was used to correlate 47 sets of heat exchanger data to develop the heat transfer and friction correlation. The correlations are applicable to heat exchangers having small diameter tubes (or large tube pitch to tube diameter ratio), whose performance previous correlations failed to predict adequately. The heat transfer correlation applicable to three or more row configuration predicts 94 percent of the data within ±20 percent, and the heat transfer correlation applicable to one- or two-row configuration predicts 94 percent of the data within ±20 percent. The friction correlation predicts 90 percent of the data within ±20 percent.
16