| Abstract
| - Solid−fluid heat-transfer coefficients have an important role in the design of chemical processing equipment.The major resistance to heat transfer lies in a region very close to the wall, where experimental measurementsare very difficult. The validity and accuracy of the models developed for the estimation of the heat- andmass-transfer coefficient still do not have general applicability for the entire range of Reynolds and Prandtlnumbers, because of the limited knowledge of near-wall turbulence. There have been two approaches forsuch model development: one is an analytical approach, which considers the momentum, mass, and heattransfer to be analogous in nature and the understanding of one of these processes can be used to predict theother two; the other approach is heuristic, based on the visualization of the behavior of the coherent structuresin the near-wall region. The continuous movement of fluid elements to and away from the wall (coherentstructures) affects the transport phenomena. The models for the quantification of this behavior have beendeveloped for the estimation of heat- and mass-transfer rates in the literature. However, both these approachescontain parameters fitted empirically to obtain good agreement with the experimental heat- and mass-transferdata. These models must be tested for their formulation and empirical constants on the basis of accuratesolutions of governing equations of heat, mass, and momentum transfer. This is possible using direct numericalsimulation (DNS) and large eddy simulation (LES), which can accurately predict the near-wall flow pattern.An attempt has been made to exploit the ability of DNS and LES to develop insight into hitherto used models,based on analogies and/or heuristic arguments.
|
