(A) study on the onset of flow instability and critical heat flux in a narrow rectangular channel

Abstract

The onset of the flow instability (OFI) and critical heat flux (CHF) in narrow rectangular channels are very crucial subjects for design and safety analysis concerning the systems such as high power research reactors (RRs) using plate-type fuels, since the phenomena could lead damage to fuels due to the initiation of local degradation of heat transfer. Therefore, experimental and analytical studies listed below have been performed to improve the understanding and predicting performance of both phenomena in a narrow rectangular channel heated on both sides, focusing on low pressure and low mass flux (LPLF) conditions, in consideration of transient or abnormal conditions of RRs. First, two-phase flow excursion instability (FI) especially for a downward flow was experimentally investigated. Three test channels with different gap sizes (2.5, 3.3 and 4.1 mm) for the same channel width (40 mm), the heated width (30 mm) and the length (350 mm) were adopted for the experiment. In addition, top plenum was opened to simulate a single channel of multi-channels in open-pool-type RRs. For several im-posed heat flux values ($ranging 0-1000 kW/m^2$) and high inlet subcooling (43-75 K), subcooled flow boiling was investigated while reducing the fluid mass flux starting from a sufficient high-mass-flux condition (i.e. mass-flux-controlled system) up to the occurrence of the flow excursion. The minimum mass flux conditions in which a stable flow is sustainable were identified via 47 FI data points and were compared with relevant correlations. The results showed that FI was highly affected by gap size of channel, and therefore new empirical correlation reflecting the gap effect were suggested. The visualization of the flow boiling using a high speed camera was clearly demonstrating that the FI is triggered by the abrupt increase of pressure drop due to the coalescence of facing bubbles (for Pe < 14000) or wavy vapors (for Pe > 14000) on opposing heated surfaces, rather than onset of significant void (OSV), which has been generally conceptualized. Secondly, the characteristics of premature and stable CHF for downward flow were experimentally investigated using a narrow rectangular channel with gap of 2.35 mm. The flow boiling was developed as the wall heat flux was increased for the imposed mass flux of 500 and $1000 kg/m^2s$, and inlet subcooling of 52 -74 K (i.e. heat-flux-controlled system). In addition, top plenum was closed to adopt inlet throttling, while lower plenum was connected to the open pool so as to maintain atmospheric pressure at the exit. The results showed that pressure drop fluctuation was inherently involved in the flow boiling, regardless of the amount of throttling, after a distinct initiating point called as the onset of pressure drop fluctuation (OPDF). It was also verified that coalescence of bubble (or vapor) layers on opposing heated surfaces triggered the OPDF as the expansion of the liquid-vapor interface became bounded by the other layer on the opposite side. The fluctuation was amplified as the exit quality was increased and the flow regime passed through the unstable slug or churn flow. For low inlet throttling conditions, premature CHF was induced at the maximum fluctuation during those flow regimes. On the other hand, stable CHF was obtained in the re-stabilized flow regime as the annular flow was developed only for large inlet throttling conditions with an additional pressure drop across the valve of more than 0.4 bar for the experimental configuration. Thirdly, a database including 62 OPDF data and 54 CHF data was constructed with additional experimental data for various system settings. 24 CHF data were stable CHFs, which were identified based on the characteristic of pressure drop fluctuations. Parametric trends analysis, flow regime analysis and correlation assessment were performed with the database. Based on the analysis, the upper and lower limits of CHF were newly proposed. Especially for stable CHF, it was verified that the onset of liquid entrainment in the annular flow triggered stable CHFs for downward flow in a narrow rectangular channel, instead of onset of slug or churn-annular flow transition, which was proposed in relevant literatures. In addition, it was certified that the limited systematic effects such as plenum or heater material effect was involved in the database. Fourthly, we conducted more detailed investigations focusing on the growth of bubble layers (BLs) not only for downward flow but also for upward flow. Furthermore, image processing algorithm was newly developed to automatically extract quantified bubble layer thickness (BLT) from vast number of images. It was verified that triggering mechanism is identical for both upward and downward flow from visualization, that is bubble blockage due to mergence of facing BLs on opposite side, but OFI for upward flow occurred at 25 % higher heat flux under same fluid conditions. Total 194 BLT data were obtained for various fluid conditions by using image processing algorithm. Parametric trends analysis showed that the maximum BLT increases with increasing heat flux, and decreasing mass flux and inlet subcooling. From the results, R factor which has been utilized as a key parameter in existing OFI and OSV correlations was adopted for development of new correlation. It was well represented that OFIs were enhanced for upward flow against downward flow due to delay of BL growth and the merging phenomena. Finally, new mechanistic model based on force balance approach was proposed for the prediction of the merging point and therefore OFI, since maximum BLTs were comparable to bubble departure diameter.