To analyze the aerosol dynamics in severe accidents of LMFBR, a new computer code has been developed. The code entitled MCAD (Multicomponent Aerosol Dynamics) can treat two component aerosol system in accident scenarios. Coagulation and removal mechanisms are included in this model. In order to see the effect of particle geometry, the code makes use of the concept of shape factors or density correction factor. In this code, they are utilized simultaneously and their effects are explained. To simplify the process of multicomponent interactions, geometric constraints are considered in calculation of coagulation coefficient.
The code is verified using the experimental result of NSPP-300 series and compared to other code. At present, it fits the result of experiment very well and agrees to the existing code. But, it is very complicate and cumbersome to treat containment circumstance. Hence, several factors have to be studied for accurate analysis.
The input variables included are very uncertain. Therefore uncertainty and sensitivity analysis is essential as a supplement to code development. In this analysis, Latin hypercube sampling and experimental design method are compared. 14 variables are selected to analyze the input uncertainty. The analytical result gives an insight to which variables are significant as time elapse and their reasonable ranges. Therefore, these procedures are repeated until a good fitting is achieved.
중대사고시 LMFBR의 에어로졸(aerosol) 동특성을 살피기 위해 전산코드인 MCAD(MultiComponent Aerosol Dynamics) 가 개발되었다.
Brownian 확산 및 중력 작용에 의한 결합 및 제거 과정을 고려했으며, 입자형태를 고려하기 위하여 밀도보정과 형태요소(shape factor)를 동시에 고려하였다.
ORNL의 NSPP-300 계열 실험 자료를 MCAD의 입증에 이용하였다. 그 결과 MCAD의 계산치와 실험치가 잘 일치함을 보여준다.
또한, 여러 입력자료의 불확실한 값들을 정의하고 그들 값의 한계를 설정하기 위하여 불확실성및 민감도해석을 수행하였다. 14개의 입력자료를 선택하여 실험계획법과 Latin Hypercube Sampling에 의한 입력자료를 조합하여 그 회귀 (regression) 정도를 구하였다. 각 변수들의 중요성 및 그들의 등위를 결정하기 위해 단계식 회기방법(stepwise regression metho) 을 고려했으며, Monte Carlo Method에 의해 위 방법들을 보완 하였다.