R134a has the advantage of conducting visible critical flow experiments because of the lower temperature and pressure at the critical point than water. Whether the water critical flow model is suitable for predicting the flow rate and the pressure profile of R134a critical flow is unknown. An experiment on the critical flow rate and the pressure profile of R134a critical flow in a slit, mimicking the shape of a crack in the pipe, has been conducted. The dimension of the slit and the subcooling degree at the entrance are changed in the experiment. The Delayed Equilibrium Model (DEM), which is accurate for the water critical flow mass flux in a slit, is verified based on the experiment results. The empirical coefficients in the original DEM are adjusted according to the approximate saturated curve of R134a. The conservative and aggressive approaches to changing empirical coefficients are verified, respectively. For R134a, the ratio of boiling pressure to saturated inlet pressure is determined to minimize the relative error of volume flow rate and pressure profile in the slit with 0.3 mm width. It proves that the aggressive approach is lower in relative errors and differences in relative errors among different slit widths and different subcooling degrees. For the aggressive approach, the abstract value of the relative error of volume flow rate is lower than 20% in most cases. The abstract value of the mean relative error of pressure is lower than 15% in most cases. Besides, the differences in relative errors among different slit widths are not significant, and the differences in relative errors tend to be zero as the subcooling degree increases. DEM is adequate for predicting the volume flow rate and pressure profile for the R134a critical flow in a slit. Nevertheless, the relative error of the volume flow rate should disperse around zero at low subcooling degree, which is not achieved in this research. Further optimization requires more rigorous analytical methods and experiment data to find the best set of empirical coefficients, and the effect of velocity difference may be introduced.
Read full abstract