Investigation on the nonintrusive multi-fidelity reduced-order modeling for PWR rod bundles
3
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Performing high-fidelity computational fluid dynamics (HF-CFD) to predict the flow and heat transferstate of the coolant in the reactor core is expensive, especially in scenarios that require extensiveparameter search, such as uncertainty analysis and design optimization. This work investigated theperformance of utilizing a multi-fidelity reduced-order model (MF-ROM) in PWR rod bundles simulation. Firstly, basis vectors and basis vector coefficients of high-fidelity and low-fidelity CFD results areextracted separately by the proper orthogonal decomposition (POD) approach. Secondly, a surrogatemodel is trained to map the relationship between the extracted coefficients from different fidelity results. In the prediction stage, the coefficients of the low-fidelity data under the new operating conditions areextracted by using the obtained POD basis vectors. Then, the trained surrogate model uses the lowfidelity coefficients to regress the high-fidelity coefficients. The predicted high-fidelity data is reconstructed from the product of extracted basis vectors and the regression coefficients. The effectiveness ofthe MF-ROM is evaluated on a flow and heat transfer problem in PWR fuel rod bundles. Two data-drivenalgorithms, the Kriging and artificial neural network (ANN), are trained as surrogate models for the MFROM to reconstruct the complex flow and heat transfer field downstream of the mixing vanes. The results show good agreements between the data reconstructed with the trained MF-ROM and the highfidelity CFD simulation result, while the former only requires to taken the computational burden oflow-fidelity simulation. The results also show that the performance of the ANN model is slightly betterthan the Kriging model when using a high number of POD basis vectors for regression. Moreover, theresult presented in this paper demonstrates the suitability of the proposed MF-ROM for high-fidelityfixed value initialization to accelerate complex simulation
Abstract
Performing high-fidelity computational fluid dynamics (HF-CFD) to predict the flow and heat transferstate of the coolant in the reactor core is expensive, especially in scenarios that require extensiveparameter search, such as uncertainty analysis and design optimization. This work investigated theperformance of utilizing a multi-fidelity reduced-order model (MF-ROM) in PWR rod bundles simulation. Firstly, basis vectors and basis vector coefficients of high-fidelity and low-fidelity CFD results areextracted separately by the proper orthogonal decomposition (POD) approach. Secondly, a surrogatemodel is trained to map the relationship between the extracted coefficients from different fidelity results. In the prediction stage, the coefficients of the low-fidelity data under the new operating conditions areextracted by using the obtained POD basis vectors. Then, the trained surrogate model uses the lowfidelity coefficients to regress the high-fidelity coefficients. The predicted high-fidelity data is reconstructed from the product of extracted basis vectors and the regression coefficients. The effectiveness ofthe MF-ROM is evaluated on a flow and heat transfer problem in PWR fuel rod bundles. Two data-drivenalgorithms, the Kriging and artificial neural network (ANN), are trained as surrogate models for the MFROM to reconstruct the complex flow and heat transfer field downstream of the mixing vanes. The results show good agreements between the data reconstructed with the trained MF-ROM and the highfidelity CFD simulation result, while the former only requires to taken the computational burden oflow-fidelity simulation. The results also show that the performance of the ANN model is slightly betterthan the Kriging model when using a high number of POD basis vectors for regression. Moreover, theresult presented in this paper demonstrates the suitability of the proposed MF-ROM for high-fidelityfixed value initialization to accelerate complex simulation
참고문헌 (31)
- G. Chen, "CFD simulation for the optimal design and utilization of experiment to research the flow process in PWR", Ann. Nucl. Energy, 94, pp.1~9 (2016)
- G. Chen, "Design of a CFD scheme using multiple RANS models for PWR", Ann. Nucl. Energy, 102, pp.349~358 (2017)
- M. Conner, "Hydraulic benchmark data for PWR mixing vane grid", Nucl. Eng. Des, 264, pp.97~102 (2013)
- S. Chang, "Phenomenological investigations on the turbulent flow structures in a rod bundle array with mixing devices", Nuclear Engineering and Design - NUCL ENG DES, pp.238~ (2008)
- S. Bhattacharjee, "Comparative study of the contribution of various PWR spacer grid components to hydrodynamic and wall pressure characteristics", Nucl. Eng. Des, 317, pp.22~43 (2017)
- B. Liu, "Sub-channel CFD for nuclear fuel bundles", Nucl. Eng. Des, 355, pp.110318~ (2019)
- M. Wang, "Recent progress of CFD applications in PWR thermal hydraulics study and future directions", Ann. Nucl. Energy, 150, pp.107836~ (2021)
- B. -W. Yang, "Recent challenges in subchannel thermalhydraulics-CFD modeling, subchannel analysis, CHF experiments, and CHF prediction", Nucl. Eng. Des, 354, pp.110236~ (2019)
- "Review of multi-fidelity models, arXiv preprint arXiv:1609.07196" (2016)[URL]
- M. Zarei, "On a reduced order modeling of the nuclear reactor dynamics", Appl. Math. Comput, pp.393~ (2020)
- Y. Sun, "A POD reduced-order model for resolving the neutron transport problems of nuclear reactor", Ann. Nucl. Energy, 149, pp.107799~ (2020)
- S. Lorenzi, "POD-Galerkin method for finite volume approximation of Navier-Stokes and RANS equations", Comput. Methods Appl. Mech. Eng, pp.311~ (2016)
- L. Vergari, "Reduced order modeling approach for parametrized thermal-hydraulics problems : inclusion of the energy equation in the POD-FV-ROM method", Prog. Nucl. Energy, 118, pp.103071~ (2020)
- D. Alonso, "Robust reduced order modeling of heat transfer in a back step flow", Int. J. Heat Mass Tran, 52, pp.1149~1157 (2009)
- R. Chen, "An Effective Learning Procedure for Multi-Fidelity Simulation Optimization with Ordinal Transformation", ~ (2015)
- A. Forrester, "Engineering Design via Surrogate Modelling, A Practical Guide", ~ (2008)
- E. Minisci, "Robust design of a reentry unmanned space vehicle by multifidelity evolution control", AIAA J, 51, pp.1284~1295 (2013)
- M. Fossati, "Multiparameter analysis of aero-icing problems using proper orthogonal decomposition and multidimensional interpolation", AIAA J, 51, pp.946~960 (2013)
- F. Alsayyari, "A nonintrusive adaptive reduced order modeling approach for a molten salt reactor system", Ann. Nucl. Energy, 141 (2020)
- X. Wang, "Multi-fidelity surrogate reduced-order modeling of steady flow estimation", Int. J. Numer. Methods Fluid, pp.92~ (2020)
- M. J. Mifsud, "A Variable-Fidelity Aerodynamic Model Using Proper Orthogonal Decomposition", ~ (2016)
- B. Noack, "From snapshots to modal expansions e bridging low residuals and pure frequencies", J. Fluid Mech, 802, pp.1~4 (2016)
- Z. Karoutas, "3-D flow analyses for design of nuclear fuel spacer", Proceedings of the Seventh International Meeting on Nuclear Reactor Thermal-Hydraulics, pp.3153~3174 (1995)
- Fabian Wiltschko, "Validation of RANS models and Large Eddy simulation for predicting crossfl ow induced by mixing vanes in rod bundle", Nuclear Engineering and Technology, 53(11), pp.3625~3634 (2021), 한국원자력학회[DOI]
- C. C. Liu, "CFD evaluation of turbulence models for flow simulation of the fuel rod bundle with a spacer assembly", Appl. Therm. Eng, 40, pp.389~396 (2012)
- M. Holloway, "Investigation of swirling flow in rod bundle subchannels using computational fluid dynamics, international conference on nuclear engineering", Proceedings, ICONE, pp.2006~ (2006)
- L. Xiaochang, "Methods of simulating large-scale rod bundle and application to a 17"17 fuel assembly with mixing vane spacer grid", Nucl. Eng. Des, 267, pp.10~22 (2014)
- M. D. McKay, "Comparison of three methods for selecting values of input variables in the analysis of output from a computer code", Technometrics, 21, pp.266~294 (2000)
- ANSYS Inc, "ANSYS Fluent Customization Manual", ~ (2016)
- A. Ayodeji, "Causal augmented ConvNet: a temporal memory dilated convolution model for long-sequence time series prediction", ISA Trans (2021)
- G. Chen, "Distributed-parallel CFD computation for all fuel assemblies in PWR core", Ann. Nucl. Energy, 141, pp.107340~ (2020)