Investigation of local heat transfer coefficients in plate heat exchangers with temperature oscillation IR thermography and CFD

Abstract

A method for the measurement of local convective heat transfer coefficients from the outside of a heat-transferring wall has been developed. This method is contact-free and fluid independent, employing radiant heating by laser or halogen spotlights and an IR camera for surface temperature measurements; it allows for the rapid evaluation of the heat transfer coefficient distribution of sizable heat exchanger areas. The technique relies first on experimental data of the phase lag of the outer surface temperature response to periodic heating, and second on a simplified numerical model of the heat exchanger wall to compute the local heat transfer coefficients from the processed data. The IR temperature data processing includes an algorithm for temperature drift compensation, phase synchronization between the periodic heat flux and the measured temperatures, and Single Frequency Discrete Fourier Transformations. The ill-posed inverse heat conduction problem of deriving a surface map of heat transfer coefficients from the phase-lag data is solved with a complex number finite-difference method applied to the heat exchanger wall. The relation between the local and the mean heat transfer coefficients is illuminated, calculation procedures based on the thermal boundary conditions are given. The results from measurements on a plate heat exchanger are presented, along with measurements conducted on pipe flow for validation. The results show high-resolution surface maps of the heat transfer coefficients for a chevron-type plate for three turbulent Reynolds numbers, including a promising approach of visualizing the flow field of the entire plate. The area-integrated values agree well with literature data. CFD calculations with an SST and an EASM–RSM were carried out on a section of a PHE channel. A comparison with the measured data indicates the shortcomings of even advanced turbulence models for the prediction of heat transfer coefficients but confirms the advantages of EASM–RSM in complex flows.