This paper presents the assessment of TRACE (version v4.160) against the Critical Heat Flux (CHF) experiments in annular tubes performed at the Royal Institute of Technology (KTH) in Stockholm, Sweden. The experimental database includes data for coolant mass fluxes between 250 and 2500 kg/m²s and inlet subcoolings of 10 and 40 K at a pressure of 70 bar.
The work presented in this paper supplements the calculations of single round tube experiments carried out earlier and provides a broader scope of validated geometries. In addition to the Biasi and CISE-GE CHF correlations available in the code, a number of experimental points at low flow conditions are available for the annular geometry experiments, which also permitted the assessment of the Biasi/Zuber CHF correlation used in TRACE v4.160 for low flow conditions. Experiments with different axial power distribution were simulated and the effects of the axial power profile and the coolant inlet subcooling on the TRACE predictions were investigated.
The results of this work show that the Biasi/Zuber correlation provides good estimation of the CHF at 70 bar, and, for the same conditions, the simulation of the annular experiments resulted in the calculation of lower CHF values compared to single-tube experiments. The analysis of the performance of the standard TRACE CHF correlations shows that the CISE-GE correlation yields critical qualities (quality at CHF) closer to the experimental values at 70 bar than the Biasi correlation for annular flow conditions. Regarding the power profile, the results of the TRACE calculations seem to be very sensitive to its shape, since, depending on the profile, different accuracies in the predictions were noted while other system conditions remained constant. The inlet coolant subcooling was also an important factor in the accuracy of TRACE CHF predictions. Thus, an increase in the inlet subcooling led to a clear improvement in the estimation of the critical quality with both Biasi and CISE-GE correlations.
To complement the work, three additional CHF correlations were implemented in TRACE v4.160, namely the Bowring, Tong W-3 and Levitan-Lantsman CHF models, in order to assess the applicability of these correlations to simulate the CHF in annular tubes. The improvement of CHF predictions for low coolant mass flows (up to 1500kg/m²s) is noted when applying Bowring CHF correlation. However, the increase in the inlet subcooling increases the error in predicted critical quality with the Bowring correlation. The Levitan-Lantsman and Tong-W-3 correlations provide results similar to the Biasi model. Therefore, the most correct CHF predictions among the investigated correlations were obtained using CISE-GE model in the standard TRAC v4.160 code.