Discussion: “Heat Transfer and Fluid Flow Analysis of Interrupted-Wall Channels, with Application to Heat Exchangers” (Sparrow, E. M., Baliga, B. R., and Patankar, S. V., 1977, ASME J. Heat Transfer, 99, pp. 4–11)

Abstract

R. K. Shah. Heat transfer and flow friction characteristics of compact heat exchanger surfaces are mainly determined experimentally. Theoretical/analytical solutions that are available are primarily useful only for continuous cylindrical passage type compact surfaces. The authors of this paper are the first investigators to analyze the complex flows and heat transfer in one type of interrupted wall compact surface which is referred to as strip fin, serrated fin or offset fin. Analyses of this type are long needed and are most welcome. My congratulations to the authors for their very fine and practical paper. As the authors have clearly mentioned in their Introduction, the heat transfer coefficients in the entrance region of a duct are substantially higher than those at locations further downstream. This fact has motivated having heat exchanger surfaces interrupted in the flow direction. It is a common presumption that the shorter the interrupted length, the higher the performance of the surface and subsequently of the exchanger. The reduction in the strip length is presently limited by the manufacturing technology, although significant advances have been made by the industry in the past decade. One simple way to predict the heat transfer performance of the strip fin may be the use of the conventional plain duct thermal entry length solutions. The idealizations made in such solutions are: uniform temperature profile and uniform (or developed) velocity profile at the start of each strip fin and the conventional boundary layer growth over it. The applicable theoretical mean Nusseit numbers correspond to those for the entrance region of a plain duct of length L. In Fig. 10, such Num_r for the constant wall temperature boundary condition are plotted as a function of L* = L/(iH Re Pr) for parallel plates for two flow conditions: developing velocity profiles, and fully developed velocity profiles [10]. Superimposed are Nup from Table 1 of the paper for the periodic fully developed flow, and N u m j (=7.541) for the fully developed flow through parallel plates. The following observations may be made from this figure: 1 For L* > 0.017, the strip length mean Nusseit numbers associated with a periodic fully developed flow are greater than those for a simultaneously developing flow for parallel plates. The augmentation characteristics of the interrupted wall is clearly seen in this region. 2 As expected, periodic fully developed Nup are always greater than 7.541 for the stable fully developed flow through parallel plates.