【摘要】 用蒙特卡罗方法模拟点通量深穿透问题时 ,存在计算费用高、计算效率低的缺陷 [1 ] .针对阵列中子测井这个点通量深穿透问题 ,除了合理利用 MCNP3.0 B提供的降低方差技巧外 ,提出了效率卷积方法来降低方差。效率卷积方法就是把直接跟踪中子历史的过程分解为两个步骤来完成 :第一步计算从源出发 ,达到探测器镉片处的中子面流 ;第二步计算探测器效率 ;最后用面流和探测器效率进行卷积得到探测器的响应。计算结果表明效率卷积方法在降低方差方面非常有效 ,该方法已成功地应用于阵列中子测井的探测器响应计算和仪器优化设计。更多还原

【Abstract】 The deficiencies in simulation of deep penetration and point flux with Monte Carlo method are high cost and low efficiency. In addition to variance reduction technique by means of MCNP, efficiency convolution is used to reduce variance so as to solve the problem of point flux and deep penetration in array neutron log. In the efficiency convolution method the direct neutron-tracking process is devided into two steps. First, the surface flow from the source to the Gd flake of the detector is simulated. Second, the detector efficiency is simulated. The detector response is calculated by convoluting the surface flow with the detector efficiency. The result indicates that the efficiency convolution method is very effective for variance reduction. And this method has been successfully applied to the simulation of the detector response of array neutron log and the tool optimum design.更多还原