Rod Bundle Array Research Articles

Improvement of macroscopic turbulence model for subchannel analysis in the rod bundle array

The PRIUS program was established to generate an experimental database for the 6 × 4 and 12 × 6 rod bundle geometry. The database will be used to address the subchannel and CFD code analysis required for modeling and validation. This is necessary because Small Break Loss of Coolant Accident (SBLOCA) and Intermediate Break Loss of Coolant Accident (IBLOCA) present three-dimensional phenomena in the core due to the radial power profile, crossflow, and diffusion-dispersion. Therefore, specific experimental programs are required, especially during core reflooding, to investigate the large-scale three-dimensional effects. However, validating each sensitive model of the code separately in the presence of 3D effects is not possible due to the inability to implement instrumentation at high pressure and temperature steam-water flow conditions. The PRIUS test program uses a single-phase flow test to simulate a non-homogeneous velocity distribution and provide information on crossflow with radial mixing effects between subchannels. The CUPID code, which uses a macroscopic turbulence model, has been validated using the PRIUS-II experimental database. Existing macroscopic turbulence models were also validated for their prediction capabilities with different inlet flow conditions. However, the validation revealed significant errors in the shear region between subchannels. An improved macroscopic turbulence model showed promising results in predicting turbulence kinetic energy in porous media analysis.

Experimental investigation and correlation development for liquid carryover during reflood

A correlation is developed for the liquid carryover fraction during both constant and oscillatory reflood with application to nuclear power accidents. The correlation, semi-empirical in nature, is based on extensive experiments in a 7 × 7 rod bundle array at the Nuclear Regulatory Commission/Pennsylvania State University Rod Bundle Heat Transfer (NRC/PSU RBHT) facility. Liquid carryover and entrainment can significantly impact quench behavior and maximum cladding temperatures during a postulated accident; therefore, accurate predictions of these phenomena are critical. The parameters associated with the stability at the liquid–vapor interface, and hence the liquid entrainment and subsequent carryover are established. These parameters provide the basic form of the correlation with experimental data from the NRC/PSU RBHT facility to determine the model coefficients. Using a weighted least-squares method that minimizes the error between modeled and experimental carryover fraction, we implemented a genetic algorithm to identify the parameters of this correlation. The proposed model is compared with experimental data and predictions from the NRC’s TRACE code and lies within 20% error margin.

Adventures in study of the pressure losses in wire-wrapped rod bundle arrays

For the occasion of celebrating the 40th anniversary of the NURETH conference series, we present a brief overview of our work on the thermal-hydraulics of wire-wrapped rod bundles over these past four decades. In this context, this short technical note presents a personal perspective about this topic in a colloquial narrative. Particular focus is placed on a model for calculating hydraulic performance of rod bundles with wire spacers as used in sodium- and lead-cooled fast reactors. Since its first publication in 1986, this model has been often discussed by the international community at events like NURETH, sparking some suggestions for improvement. These issues have been addressed in a recent series of articles leading to an upgraded version published in 2018. This upgraded model yields a better overall prediction of all available experimental data and accurately represents some relevant geometrical aspects, such as the friction factor for two otherwise identical bundles but with different number of pins (bundle size effect). An outlook on the next steps for this effort is given, particularly from the point of view of using this model for licensing purposes.

Experimental visualization of flow structure inside subchannels of a 4 × 6 rod-bundle

A fuel element geometry frequently used in nuclear reactors is the rod-bundle. In it, coolant flows axially through subchannels formed between the rods. The mixing of cooling fluid in a rod-bundle reduces the temperature differences in the coolant and along the perimeter of the rods. To ensure thermal performance of a nuclear reactor, detailed information about heat transfer and turbulent mixing taking place within subchannels is required. To improve the prediction ability of the subchannel analysis using computer code, a considerable amount of experimental data should be utilized for developing a new model for subchannel analysis. Experimental work has been conducted in the PRIUS-I (in-PWR Rod-bundle Investigation of Undeveloped mixing flow across Subchannel) test facility, which has adopted the matched index-of-refraction (MIR) technique to visualize the multi-dimensional velocity and turbulence fields in an unheated 4 × 6 rod-bundle geometry as well as to provide benchmark data for computer code validation. The 4 × 6 rod-bundle array has the same pitch-to-diameter (P/D) as the prototype fuel assembly. PRIUS-I test facility can simulate the uniform and non-uniform inlet flow conditions in order to validate the effect in crossflow between the fuel assemblies of PWR. In this study, the crossflow and mixing phenomena that exist between two adjacent subchannels through the gaps are visualized in any cross section of the rod-bundle geometry using a MIR technique. For non-uniform inlet velocity condition, the flow characteristics near the inlet where the flow mixing occurs actively is compared with computational fluid dynamics (CFD) calculation results. The experimental data can be used as benchmark data for the system safety analysis code or CFD code validation.

The Experimental Study on Mixing Characteristics in a Square Subchannel Geometry with Typical Flow Deflectors

The lateral flow structures in a subchannel geometry were carefully examined in an enlarged 5 × 5 rod bundle array by using a two-color LDA system. The detailed measurement data of lateral velocities in subchannels have been obtained as fabricating the rod bundle array 2.6 times larger than the real bundle size. Two types of flow deflectors were selected for the test: a conventional split-type mixing vane and a swirl-type vane that had been designed to enhance the swirling flow in a subchannel. Experiments were performed in a cold test loop facility at KAERI. The experimental conditions were normal pressure (1.5 bar) and temperature (35°C). The average axial velocity was maintained at 1.5 m/s and was equivalent to Re = 50,000. From the experimental results, specific flow characteristics for each vane types were observed.

Enhancement of heat transfer in a typical pressurized water reactor by different mixing vanes on spacer grids

The flow mixing devices on a grid spacer are designed to enhance the turbulence and heat transfer in sub-channels. The present study evaluates the effects of different mixing vane configurations on flow pattern and heat transfer in the downstream of the mixing vanes in the sub-channels of fuel assembly. This is done by obtaining velocity and pressure fields, turbulent intensity and the heat transfer coefficient using a three-dimensional CFD analysis. Two blade groups, two-dimensional and three-dimensional, were modeled. The two-dimensional blades are classified in ring type, diamond type, square type, homographic type and multiply type. Three-dimensional blades containing split vane, ripped open, trumpet shaped and split trumpet shaped are applied in a 2 × 2 rod bundle array. Also, a 3 × 3 rod bundle array is modeled to evaluate the neighborhood effects on the thermo-hydraulic parameters. The latter model shows a good comparison with the available experimental data by 5–12% difference. A standard K-epsilon model is used as a turbulence model and the symmetry condition is set as boundary conditions. It was confirmed that the turbulence in the sub-channel was significantly promoted by spacer and mixing devices. However, their effects rapidly decreased to a fully developed level after passing approximately 10 times the hydraulic diameter downstream of the spacer. The CFD results showed good agreement with the measurements. The static pressure of the fluid in the flow direction drops rapidly, then in a very short distance rise up, followed by a decrease near to the linear slope down stream. The split trumpet vane was the best with a 14% increase in the heat transfer coefficient, however regarding the manufacturing possibilities, the split vane spacers are expected to significantly enhance the overall heat transfer of a nuclear fuel assembly about 9.82% with a reasonable increase in the pumping cost.

Phenomenological investigations on the turbulent flow structures in a rod bundle array with mixing devices

Detailed turbulent flow profiles have been measured on a square sub-channel geometry with typical mixing devices. For a fine examination of the lateral flow structure on a sub-channel geometry with 2D LDA, a 5 × 5 rod bundle array was fabricated as 2.6 times larger than the real bundle size. The mixing devices used were a typical split type and a swirl type. The experiments were performed at the condition of Re = 48,000 (axial bulk velocity 1.48 m/s) and the water loop was maintained at the conditions of 35 °C and 1.5 bar during an operation. As for the results, distinct intrinsic flow features were observed according to the type of mixing devices. In a typical split type, there was no remarkable swirling flow within a sub-channel and the lateral flow was vigorous in the gaps. In the swirl type, a single swirling flow was dominant within a sub-channel and there were relatively small lateral flows in the gaps.

A Numerical Investigation of Turbulent Interchange Mixing of Axial Coolant Flow in Rod Bundle Geometries

This article is concerned with modeling turbulent interchange mixing within rod bundle arrays. Using a three-dimensional numerical solution, the characteristics and appropriateness of an isotropic κ–∊ turbulence model were investigated by comparing code predictions with experimental results. A novel methodology for solving for the fluid temperature field was also developed. Application of this methodology resulted in significant savings in computational effort for the problem of interest. Analysis of the results indicated that the isotropic model accurately predicted the radial component of turbulent eddy viscosity, but underpredicted the azimuthal component. As a result, radial diffusion processes (such as pressure drop) were predicted accurately, but azimuthal diffusion processes (such as turbulent interchange mixing) were underpredicted. Higher-order turbulence models such as algebraic stress models are therefore required if the details of the turbulent mixing are to be resolved.

The French Thermal-Hydraulic Program Addressing the Requirements of Future Pressurized Water Reactors

In the framework of feasibility studies of future pressurized water reactors, a thermal-hydraulic program has been established in order to qualify computer codes in three fields: critical heat flux and reflooding heat transfer in triangular rod bundle arrays with tight lattices and the coolability of movable fertile rods in their guide tubes. Tests carried out at the Commissariat a l'Energie Atomique in cooperation with Electricite de France and Framatome are now in progress, and preliminary results are shown for these three studies.

Fully developed rod bundle flow over a large range of Reynolds number

In an investigation of the fluid mechanics of single phase reactor cores, extensive measurements of mean axial velocity, wall shear stress and all six Reynolds stresses have been made in fully developed flow through a square pitched rod bundle array with pitch to diameter ratio of 1.107. The range of Reynolds numbers, based on bulk velocity and hydraulic diameter was 22600 to 207600. The mean secondary flow velocities could not be measured at any Reynolds number, implying that they were always less than about 1% of the bulk velocity. The axial momentum integral equation is used to show that the wall shear stress distribution is determined primarily by the pressure gradient and the transverse shear stress |ovbar|uw, a result that confirms the negligible size of the mean secondary flow. The implications of the results for current engineering calculation methods are discussed.

Flow pattern transition for gas-liquid flow in a vertical rod bundle

Experimental and analytical studies are presented directed to the prediction of flow pattern transitions in a vertical rod bundle array. The test section used consists of a 24 rod matrix on a square pitch in a cylindrical shell. Analytical models are given based on physical interpretation of the transition mechanisms. Data on rise velocities of small and large bubbles in rod bundles are also presented.

Turbulence Modeling of Axial Flow in a Bare Rod Bundle

Temperature distribution within the rod bundle of a nuclear reactor is of major importance in nuclear reactor design. However temperature information presupposes knowledge of the hydrodynamic behavior of the coolant which is the most difficult part of the problem due to the complexity of the turbulence phenomena. In the present work a two equation turbulence model—a strong candidate for analyzing actual three dimensional turbulent flows—has been used to predict fully developed flow of infinite bare rod bundle of various aspect ratios (P/D). The model has been modified to take into account anisotropic effects of eddy viscosity. Secondary flow calculations have been also performed although the model seems to be too rough to predict the secondary flow correctly. Heat transfer calculations have been performed to confirm the importance of anisotropic viscosity in temperature predictions. Experimental measurements of the distribution of axial velocity, turbulent axial velocity, turbulent kinetic energy and radial Reynolds stresses were performed in the developing and fully developed regions. A two channel Laser Doppler Anemometer working in the reference mode with forward scattering was used to perform the measurements in a simulated interior subchannel of a triangular rod array with P/D = 1.124. Comparisons between the analytical results and the results of this experiment as well as other experimental data in rod bundle arrays available in the literature are presented. The predictions are in good agreement with the results for high Reynolds numbers.

SIMPLE-2: A computer code for calculation of steady-state thermal behavior of rod bundles with flow sweeping

A computer code has been developed for use in making single-phase thermal hydraulic calculations in rod bundle arrays with flow sweeping due to spiral wraps as the predominant crossflow mixing effect. This code, called SIMPLE-2, makes the assumption that the axial pressure gradient is identical for each subchannel over a given axial increment, and is unique in that no empirical coefficients must be specified for its use. Results from this code have been favorably compared with experimental data for both uniform and highly nonuniform power distributions. Typical calculations for various bundle sizes applicable to the LMFBR program are also included.

Single Phase Turbulent Mixing in Simulated Rod Bundle Geometries

Single-phase air and water turbulent mixing rates, between two square array subchannels simulating parallel flow areas in a reactor rod bundle array, were measured employing tracer techniques. Mixing rates increased with Reynolds number over the range 1,300 to 38,000. Variation of the interconnection gap spacing from 0.015 to 0.080 in. indicated that the mixing mechanism was both gap spacing and Reynolds number dependent. At low Reynolds number and low gap spacings, molecular diffusion plays an important role in the mixing mechanism. However, generation and penetration of secondary flows are believed to be an important mode of mixing in the intermediate Reynolds number range---the largest secondary flows and minimum penetration being encountered for the lowest pitch to rod diameter ratios. (auth)