Preliminary Simulation Results of VERA Problems 6 and 7 Using NECP-X / SUBSC

Jun Chen,Liangzhi Cao,Xingjian Wen,Zhouyu Liu
2018-01-01
Abstract:With the improved understanding of the governing physical mechanism and the steady growth of computing capability, especially the development of massively parallel processing system, the coupling of high-fidelity neutronics and thermal-hydraulics becomes more and more popular worldwide. Serpent 2/SUBCHANFLOW [1], MC21/CTF [2] and RMC/CTF [3] are several famous Monte Carlo methods based coupling codes. All of them are capable of dealing with large-scale problems, including the BEAVRS benchmark [4] and VERA #7 [5]. Still, the computational burden of the Monte Carlo methods is too large compared to the deterministic based methods. For example, MC21 required 6.8 days using 1024 cores while MPACT only took 57 minutes using 464 cores for one Picard iteration [2]. Therefore, the deterministic based coupling codes, including nTRACER/MATRA [6], MPACT/CTF [2] and NECPX/SUBSC [7] are much more competitive from efficiency viewpoint. This paper summarizes the recent work about applying NECP-X/SUBSC to VERA #6 and VERA #7.
What problem does this paper attempt to address?