Simulation of the SPEL Small Scale Moderator Experiments Using the General Purpose Fluid-Flow Heat Transfer Code Phoenics

Abstract

A small scale experimental tank was previously constructed at SPEL (Sheridan Park Engineering Laboratory) to improve our understanding of moderator circulation inside the calandria of a typical CANDU reactor. The apparatus was not a scale model of any existing calandria but embodied the salient features of a typical calandria cross-section. Seven ports on the top portion of the cylindrical shell were provided for temperature measurement by probe insertion. Temperatures were measured at various locations along a vertical line at each port. The prime reason for construction the test facility was to provide experimental verification of the AECL in-house code PDMOTH which is used to simulate thermalhydrualic behaviour in an actual CANDU moderator. The code was developed by Carlucci et al., at CRNL (Chalk River Nuclear Laboratories). PDMOTH temperature predictions were in reasonable agreement with SPEL test results.


In applying the 20 code PDMOTH to actual moderator simulations, various approximations/assumptions have to be made, eg: treatment of axial heatlead variation, neglect of axial flow, etc. In view of these assumptions/approximations, it was decided to perform some 3D simulations using the general purpose fluid-flow, heat transfer code PHOENICS. Prior to undertaking these 3D simulations, it was natural to simulate some of the SPEL tests using the PHOENICS code. Both 2D and 3D simulations were performed.


In conclusion the agreement between PHOENICS predictions and the experimentally observed temperatures was good. There was little difference between the 2D and 3D predictions, indicating that the flow in the experimental apparatus was essentially two-dimensional.