Search

At present, there are two possible approaches, which practically are used for the thermal-hydraulic analysis of the reactor core in simulations, where crossflow effects are relevant: the representation of the core by a system of separated parallel flow subchannels with provision of a forced crossflow mixing by implementation of additional formulations and the porous body approach, in which the core geometry is replaced by a structure of homogenized zones of porous media. When crossflow is comparable with flow along the bundles, the porous body approach is the only one suitable method to perform steady-state and transient flow calculations for safety analyses. The CFD – code CFX-4 offers a porous region model for the modelling of the core geometry as a homogenized medium. This model is characterised by a set of properties (volume porosity, resistance to flow and so on). The Core Crossflow Experimental Facility (CCEF) was built with the objective to obtain experimental data for the flow in a rod bundle under the conditions of forced crossflow with relatively low Reynolds numbers and variable flow angle. The test section of CCEF (a plexi-glass model) contains the test rod bundle. The test rod bundle is based on the geometry of a typical PWR with a pitch/diameter ratio of 1.33 and was scaled with a factor 1.5. The crossflow was induced by an asymmetrical outlet condition for the bundle. The test bundle consists of the 100 plexiglass rods, arranged in 4 rows. Additional internals, which are typical for a reactor core geometry, were not installed in the test section. The working fluid in the experiments was water. For the velocity measurements the one-component laser-Doppler anemometer (LDA) was used. Calculations for the comparison with experimental data were performed in two different approaches. The detailed geometry was created for the calculation of flow fields with k-e turbulence model and a relative simple geometry of the test section for the flow simulations in the porous region approach. The results of the experimental investigations on the CCEF and the comparison with the calculations, performed with CFX-4 Code in the geometry of the test section are presented in this paper.

An experimental program was carried out, where heat transfer and hydraulic resistance were measured in a facility that simulated a fuel assembly of a Soviet VVER-440 nuclear power reactor. The experiments were performed in a test loop with four different rod bundles that consisted of 19 full height, directly electrically heated rods. Most of the tests were conducted close to nominal conditions of the simulated reactor. Heat transfer crisis of dryout type was measured at mass fluxes from 15 to 40 % of the nominal flow of the reactor, when inlet temperature of flow and heat flux were varied in the neighbourhood of the nominal reactor values. Occurrence of heat transfer crisis was independent of heat flux in the covered large steam qualities and a small mass fluxes. No crisis correlation predicted this phenomenon, but the GE and the Hsu-Beckner formulas obtained on an average similar values to those of the experiments. In slow flow decay transients the crisis appeared later than in the stationary experiments. Single phase heat transfer coefficient in the rod bundles was found to be close to the Dittus-Boelter correlation. Two-phase heat transfer coefficient was smaller than expected, but the measurements were disturbed by surface fouling. Single phase friction of turbulent flow was the same as had earlier been obtained in similar conditions. Two phase friction was measured up to large steam qualities beyond dryout in small mass fluxes. Local minimum of friction was observed around the crisis point. The Baroczy curves predict two-phase friction best of the correlations that were tested against the experimental data.

An experimental program was carried out, where heat transfer and hydraulic resistance were measured in a facility that simulated a fuel assembly of a Soviet VVER-440 nuclear power reactor. The experiments were performed in a test loop with four different rod bundles that consisted of 19 full height, directly electrically heated rods. Most of the tests were conducted close to nominal conditions of the simulated reactor. Heat transfer crisis of dryout type was measured at mass fluxes from 15 to 40 % of the nominal flow of the reactor, when inlet temperature of flow and heat flux were varied in the neighbourhood of the nominal reactor values. Occurrence of heat transfer crisis was independent of heat flux in the covered large steam qualities and a small mass fluxes. No crisis correlation predicted this phenomenon, but the GE and the Hsu-Beckner formulas obtained on an average similar values to those of the experiments. In slow flow decay transients the crisis appeared later than in the stationary experiments. Single phase heat transfer coefficient in the rod bundles was found to be close to the Dittus-Boelter correlation. Two-phase heat transfer coefficient was smaller than expected, but the measurements were disturbed by surface fouling. Single phase friction of turbulent flow was the same as had earlier been obtained in similar conditions. Two phase friction was measured up to large steam qualities beyond dryout in small mass fluxes. Local minimum of friction was observed around the crisis point. The Baroczy curves predict two-phase friction best of the correlations that were tested against the experimental data.

Experiments on flow-induced vibrations using a close-packed triangular rod array with a pitch-to-diameter ratio of 1.1 in water cross-flow was carried out and data analysis performed. The bundle consists of 21 rows with five rods in each. One flexibly mounted test rod (TR) is located in the fourth row in an otherwise fixed array. The rod can freely move in the transverse and in-line direction. Two accelerometer sensors were attached at the both ends of the TR to measure the rod response on the fluid flow. Kalman type filters was used in order to restore TR kinematics from simultaneous measurements of accelerometers and less sensitive laser extensometers. The effect of Reynolds number on the stability of the flexibly mounted TR has been analysed. Flow history (different flow increasing and decreasing scenarios) impact on the stability of TR oscillating motion was studied as well. Obtained experimental results was used for validation of 2-dimensional Computational Fluid Dynamics (CFD) model. In the model rods are represented with the cylindrical surfaces. Spring-mounted TR was modelled in the fourth row, as in experimental installation. Due to developed turbulence system of Unsteady Reynolds-Averaged Navier-Stokes (URANS) equations was used for fluid flow description. Unsteady pressure and viscous forces acting on the TR then are given by URANS solution. Open sourced CFD toolkit OpenFOAM was used for coupled solution of both - systems of RANS PDE’s and ODE’s of solid body dynamics with the built-in six degree-of-freedom solver. In order to reduce computational time several array size reduction techniques was applied.