This research investigated the condensation heat transfer and the flashing-induced instability, which were the two primary heat transfer mechanisms of the passive containment cooling system (PCCS) in iPOWER.
In the first part, the present research focused on the experimental study of the condensation phenomena with the presence of non-condensable gas on a slender vertical cylinder to apply the PCCS in iPOWER. The condensation experiments were performed using 1m long, 31.75mm diameter tube in 2~5 bar, 0.1~0.8 of air mass fraction, and 10~60$^\circ K$ of wall subcooling degrees. The experimental heat transfer coefficients were increased at the higher pressure conditions. Also, the heat transfer coefficients were increased in the decreasing of the air mass fraction and the wall subcooling degrees. During the experiments, the condensate flow characteristics were observed. In the high condensation rate condition, the high growth rate of drop-wise condensation was taken place in the top surface of the tube. In the middle and bottom of the tube surface, the active rivulet flow was observed. The rivulet flow could increase the condensation driving force. In the low condensation rate condition, the stagnant droplets covered the whole surface of the tube.
The rivulet flow effect was indirectly tested by CFD analysis considering the liquid film flow in a CFD model. The CFD analysis showed that the liquid film flow was not significant effects on the heat transfer mechanism, only limited in 20% of enhancement effect. The diffusion layer model with the heat and mass transfer analogy has been widely adopted in the condensation modeling with the presence of non-condensable gas. The film-wise condensation-based diffusion layer model produced about 10~40% of under-prediction in the low wall subcooling degree at the low pressure condition while it predicted the lower heat transfer coefficients by about 50~100% than the experimental data in the low wall subcooling degree at the high pressure condition. As a result, we postulated that the drop-wise condensation was significant effects on the condensation heat transfer. Based on the relevant analysis and the observed phenomena, a new correction factor was empirically introduced into the diffusion layer model to account for rivulet flow-induced condensation enhancement and the drop-wise condensation. It was confirmed that the new diffusion model produced accurate condensation heat transfer coefficients with 4.2% of the mean absolute deviation in comparison with the present experimental data. Also, predictions by the new model were in good agreement with the prototype experimental results with 6.5% of the mean absolute deviation. We can say that the present model has a good prediction capability for the longer tube and the expandability to the low wall subcooling region.
In the second part, the present research focused on the development of flashing-induced instability (FII) maps for PCCS in iPOWER. The nodalization of PCCS was modeled using the original plant design data. A series of sensitivity analysis were performed to identify the sensitivities of the nodalization scheme, selection of time step size, and alternative models and correlations in the MARS code. The coarse nodalization of the riser section induced artificial flashing oscillations, and fine nodes were preferred to estimate accurate vapor generation in the riser section. The oscillation patterns were not much sensitive to the selection of the time step size, but marginally sensitive to alternative user options in MARS code.
Based on the standard input and the analysis methodology, the stability maps for the PCCS were developed. Typically, the instability was initiated as a sinusoidal instability near the stability boundary, and the sinusoidal instability changed to the intermittent instability at the high phase change number. In respect to the stability boundary criterion, we found out that the calculated equilibrium quality at the exit of the riser as a stability boundary was approximately 1.2% above 5 kW/$m^2$, which was in good agreement with Furuya’s results. However, in a very low heat flux condition, the onset of the FII occurred at the lower equilibrium quality than the experimental data. Also, it was confirmed that inlet throttling reduced the unstable region. However, the design of the inlet restrictor for minimizing FII should be carefully determined by considering its negative effect on the heat removal performance of the natural circulation system.
본 학위논문은 피동 원자로건물냉각계통의 주요 열전달 메커니즘인 응축과 증발현상에 대한 연구를 수행하였다. 먼저, 피동냉각계통 열교환기의 열제거 성능을 검증하기 위해 원형과 동일 직경의 1m 길이 단일관에 대하여 비응축성 가스가 존재하는 조건에서 응축열전달계수를 측정하였다. 실험결과, 압력이 높을수록, 공기분율과 벽면과냉각도가 낮을수록, 응축열전달계수가 증가되는 것을 확인하였다. 응축벽면에서 발생하는 응축수의 유동양식을 관찰한 결과, 응축이 활발한 조건에서 관 상부표면에는 생성과 소멸이 활발한 적응축을 관찰하였고, 중심 및 하부표면에는 표면액적의 결합으로 생성된 하향 리뷸릿유동을 관찰하였다. 증축이 저하된 조건에서는 응축표면 대부분이 적응축이 지배하는 것을 확인하였다.
응축열전달 메커니즘이 적응축 및 하향리뷸릿 유동의 영향으로 증진되는 가설을 세우고, 이를 검증하였다. 먼저, 하향리뷸릿 유동에 대해 전산유체역학 분석결과, 응축열전달 증진효과가 20% 정도로 평가되었다. 열물질전달 상사성 기반 확산경계층모델의 분석결과, 확산경계측모델 대비 하향 리뷸릿유동의 증진효과는 최대 30%, 적응축의 경우, 최대 100% 증진효과가 있는 것으로 평가되었다. 적응축의 열전달 메커니즘에 연관성이 높은 압력과 벽면과냉각도의 함수로 표현된 새로운 수정상관식을 제시하여, 확산경계층모델을 개선하였다. 개선된 모델은 본 실험과 실규모 실험결과를 모두 6.5%의 평균절대오차, 8%의 평균제곱근오차 이내로 예측함을 확인하였다.
다음으로 피동냉각계통의 자연순환유로내에서 발생하는 증발기인 유동불안정성에 관한 해석적 연구를 수행하였다. 사고 후 장시간이 지나면, 계통내 유체의 온도가 포화상태에 이르게 되는데, 수직 16m 높이를 가지는 출구배관의 수두압으로 인해, 열교환기 관내에는 증발이 발생하지 않고, 수두손실로 인해 출구배관 최상단부에서 증발이 일어나게 된다. 이때 계통의 유동불안정성이 발생하는데 이를 증발기인 유동불안정성이라 한다. PCCS 계통내에 증발기인 유동불안정성 현상을 분석하기 위해 설계자료를 기반으로 피동냉각계통의 자연순환유로를 MARS 코드로 모델링하였다.
격자, 시구간, 사용자 옵션에 대한 코드 민감도 평가결과, 시구간, 사용자 옵션은 큰 영향이 없었으나, 증발이 발생하는 출구배관에서의 성긴격자가 가상의 유동불안정성을 예측할 수 있음을 확인하였고, 조밀격자로 최종 해석모델을 수립하였다. 검증된 해석모델을 기반으로 다양한 열속조건에 대하여 증발기인-유동불안정성 지도를 개발하였다. 분석 결과, 5kW/$m^2$ 이상의 열속조건에서는 출구건도가 1.2%에 근접할 때, 안정상태에서 불안정상태로 천이하는 것을 확인하였으며, 이는 기존 실험연구에서 제시된 결과와 동일하였다. 그러나 3kW/$m^2$ 이하의 열속조건에서는 출구건도가 1.2%보다 낮은 조건에서 유동불안정성이 개시되는 것을 확인하였다.
