Mean temperature difference in cross-flow heat exchange applied to multipass air-cooled fin-tube units with a finite number of rows

Abstract

The effective mean temperature difference (M.T.D.) in a heat exchanger depends on the terminal temperatures of the two streams, the distributions of flows over the transfer area with the associated local mixing effects and, most important, the relative directions of flow of the two streams. Air-cooled fin-tube exchangers, of the type considered here, are arranged for cross-flow. Measurements for several counter-cross-flow row-and-pass arrangements, on a test rig using commercially available fintubes, have shown that models (presented in the literature or derived here) which assume that the air stream contacts the tube rows sequentially, adequately predicted the overall performance (effective M.T.D.) of such exchangers. For predicting ZocaZ temperature changes from row to row, however, a more sophisticated model allowing for partial air bypassing of alternate rows, as well as local partial transverse mixing of air at different temperatures, appears necessary, particuarly for wide fin-tip clearances. The fin-tube geometry effectively prevents significant longitudinal (tube length direction) mixing, and models assuming complete longitudinal mixing of the air at any cross-section predict over-conservative (low) M.T.D.'s.

Die effektiewe gemiddelde temperatuurverskil (G.T.V.) in 'n hitteruiler is afhanklik van die eindpunttemperature van die twee strorne, die verspreiding van die vloei oor die oordragoppervlak inagnemend die mate van plaaslike menging en, die belangrikste, die relatiewe vloeirigtings van die twee strome. Dwarsvloei is van toepassing by die lugverkoelde vinbuisruilers, hier onder bespreking. Verskeie dwarsvloeirangskikkings, waar die effektiewe vloei van die twee strome in teenoorgestelde rigtings is, is eksperimenteel ondersoek. Dit is gevind dat die algehele warmteoordragvermoe (effektiewe G.T.V.) van sodanige ruilers effektief voorspel kan word deur modelle, of afgelei of uit die literatuur verkry, wat gebaseer is op die aannarne <lat die lugstroom opeenvolgend in kontak kom met die rye buise. 'n Meer komplekse model, waarin toegelaat word vir 'n gedeeltelike omloop van lug oor alternatiewe rye en vir gedeeltelike dwarsmenging van lug by verskillende temperature, is blykbaar nodig, veral by hitteruilers waar die vinrandspasiering groot is, om plaaslike temperatuurveranderings van ry tot ry te voorspel. Weens die geometrie van die vinbuisopstelling vind daar bykans geen menging plaas in 'n longitudinale rigting (dws rigting van buislengte) nie. Modelle gebaseer op die aann~me van algehele menging in hierdie rigting sal te klein waardes vir die G.T.V. voorspel.