A look-up table for trans-critical heat transfer in water-cooled tubes

This article describes the development and validation of a trans-critical heat transfer look-up table for water at high subcritical and supercritical pressures. As a basis for constructing the table, an extensive database of near-critical and supercritical heat transfer measurements was compiled and upgraded by the rejection of unreliable or inappropriate data, the removal of duplicates and outliers and the reduction of data scatter. A large number of available single-phase and supercritical heat transfer correlations were assessed against the database and the most accurate correlations for each heat transfer regime were identified. These correlations were then used to construct a skeleton table, which provides values of the heat transfer coefficient for a matrix of combinations of 11 values of pressure in the range from 19 to 30 MPa, 9 values of mass flux in the range from 100 to 5000 kg/m2 s, 17 values of bulk enthalpy in the range from 1000 to 3000 kJ/kg, and 8 values of wall superheat in the range from 10 to 500 K. For the construction of the final table, the predictions of correlations were replaced by experimental values, adjusted following established trends to conform to the skeleton table value matrix. Unlike all previous prediction methods, the table applies not only to normal heat transfer conditions but also to conditions with heat transfer deterioration and enhancement, as it includes data obtained under such conditions. The table values were further adjusted so that apparent discontinuities that were not related to physically plausible changes in heat transfer were smoothened out. The predictions of the table were assessed statistically against the experimental database. When compared to predictions of other available methods, the predictions of the present table were found to be generally closer to the experimental values for wide ranges of conditions. Finally, the trends followed by the smoothened table predictions for different heat transfer modes were found to represent more closely the trends of experimental data than trends followed by other prediction methods.