Book Description
This thesis is concerned with the determination of thelocal and average mass/heat transfer--from swirling jetsimpinging orthogonally onto flat surfaces. ApplicatioCn>of swirl alters the flow field of the jet considerablyand eventually the maximum velocity in the jet is displacedfrom the axis resulting in a typical double-peak profile. Further increase of swirl, can even result in arecirculation in the inner core of the flow. The turbulencecharacteristics of the jet are also affected. itis expected that these changes will, in addition, modifythe heat transfer behaviour. Consequently, a study of both single free jets and asquare array (of 3x 3) of jets was undertaken. Therange of swirls examined in this study was from zerothrough to weak and then medium swirl, i. e. the swirlnumber S was varied from 0 to 0.48. The other parametersvaried in the experiments were the nozzle-to-targetspacing z/D from 2 to 12, the nozzle pitches x/D (forarrays of jets) which were 3.2,4.8 and 6.4, and the jGtReynolds number. In the single frce jet tests, two flowrates corresponding to ReD-":2 32pOOO and 60,000 were studiedwhilst, for the multiple jets, the measurementswere confined. to the lower of these Reynolds numbers.(ReD is based on the mean exit velocity in the non-swirlingcase and the diameter of the nozzle.)Limited velocity and turbulence measurements were alsoundertaken on the single jet to characterise the flowand also to compare the behaviour of the present jetswith those in previous studies. The flows were similarto those observed previously for swirling jets so thatthe heat transfer results should be generally applicable. A thin-film naphthalene sublimation technique was usedto measure mass transfers over the target surface. Theheat transfer coefficients were then derived using theChilton-Colburn analogy. A rig was developed to spraya uniform coating of naphthalene on the target surface. This thin-film technique was found to provide repeatableresults and the validity of the experiments was further assessed by comparing the results with previouslypublished data for the no swirl case. These were inreasonable agreement. For the single free jets, the application of swirl wasfound to continuously reduce the heat transfers. Theheat transfers, however, became more uniform. An empiricalcorrelation has also been suggested for the averageNusselt number associated with these single swirling jets and is valid for S= 0.12-to 0.48. In some circumstancesin the multiple jet tests (e. g. at close nozzle-to-target spacings) the average heat transfers increasedto a maximum of IS =0.24 (approximately). Further increasesin the degree of swirl brought about a subsequentreduction in average heat transfer coefficients untileventually the performance of the swirling jets was poorerthan that of the non-swirling flows. This discrepancyin behaviour can be explained in terms of changes in thelocal heat transfer distributions.