M. Ishii’s research while affiliated with Purdue University West Lafayette and other places

Purdue Novel Modular Reactor (NMR) is a BWR-type small modular reactor design with an electric output of 50 MWe that relies on natural circulation driven flows for both normal operation and accident management. NMR and other similar natural circulation driven systems are often subject to flow instabilities, which could affect system control and safety under low-power and low-pressure conditions due to small driving force compared to forced circulation. Therefore, flow instabilities during the initial startup transients of NMR were experimentally investigated in a natural circulation test facility with an operating pressure limit of 1.0 MPa. The overall height of this test facility is seven meter, close to that of the prototype design. This makes it appropriate to study flow instability phenomena (mainly flashing instability occurred in the chimney due to reduced hydrostatic head) during reactor startup transients. In our study, four initial startup procedures with different power ramp rates were experimentally investigated with the aim of eliminating the flow instabilities observed from tests using a normal startup procedure scaled from that for SBWR-600. Specifically, a very slow startup transient test and two pressurized startup transient tests were performed. The experimental results indicated that both startup procedures were applicable to the initial startup of NMR. However, the pressurized startup procedure might be preferred due to its shorter operating hours required.

Two-phase flows in large diameter channels have a great deal of importance in a wide variety of industrial applications. Nuclear systems, petroleum refineries, and chemical processes make extensive use of larger systems. Flows in such channels have very different properties from flows in smaller channels which are typically used in experimental research. In this paper, the various differences between flows in large and small channels are highlighted using the results of previous experimental and analytical research. This review is followed by a review of recent experiments in and model development for flows in large diameter channels performed by the authors. The topics of these research efforts range from void fraction and interfacial area concentration measurement to flow regime identification and modeling, drift-flux modeling for high void fraction conditions, and evaluation of interfacial area transport models for large diameter channels.

In the two-fluid model, it is important to give an accurate prediction for the interfacial area concentration. In order to achieve this goal, the interfacial area transport equation has been developed. This study focuses on the benchmark of IATE performance in a rod bundle geometry. A set of interfacial area concentration source and sink term models are proposed for a rod bundle geometry based on the confined channel IATE model. This model was selected as a basis because of the relative similarity of the two geometries. Benchmarking of the new model with interfacial area concentration data in an 8×8 rod bundle test section which has been scaled from an actual BWR fuel bundle is performed. The model shows good agreement in bubbly and cap-bubbly flows, which are similar in many types of geometries, while it shows some discrepancy in churn-turbulent flow regime. This discrepancy may be due to the geometrical differences between the actual rod bundle test facility and the facility used to collect the data which benchmarked the original source and sink models.

In commercially available Computational Fluid Dynamics (CFD) codes such as ANSYS CFX and Fluent, the interfacial shear term is missing in the field momentum equations. The derivation of the two-fluid model (Ishii and Hibiki, 2011) indicates the presence of this term as a momentum source in the right hand side of the field momentum equation. The inclusion of this term is considered important for proper modeling of the interfacial momentum coupling between phases. For separated flows, such as annular flow, the importance of the shear term is understood in the one-dimensional (1-D) form as the major mechanism by which the wall shear is transferred to the gas phase (Ishii and Mishima, 1984). For gas dispersed two-phase flow CFD simulations, it is important to assess the significance of this term in the prediction of phase distributions. In the first part of this work, the closure of this term in three-dimensional (3-D) form in a CFD code is investigated. For dispersed gas–liquid flow, such as bubbly or churn-turbulent flow, bubbles are dispersed in the shear layer of the continuous phase. The continuous phase shear stress is mainly due to the presence of the wall and the modeling of turbulence through the Boussinesq hypothesis. In a 3-D simulation, the continuous phase shear stress can be calculated from the continuous fluid velocity gradient, so that the interfacial shear term can be closed using the local values of the volume fraction and the total stress of liquid phase. This form also assures that the term acts as an action-reaction force for multiple phases. In the second part of this work, the effect of this term on the volume fraction distribution is investigated. For testing the model two-phase flow data measured at Purdue University is assessed. The interfacial shear term is assembled in ANSYS CFX. Simulation results are presented to assess the effect of the interfacial shear term on the phase distribution.