Study on the Performance of the Neutron Diffractometer (HIPD at CARR) by Monte Carlo Simulation and Convolution Methods

Wenyun Yang,Hui Zhao,Youfang Lai,Guanyi Qiao,Yuanhua Xia,Honglin Du,Jingzhi Han,Changsheng Wang,Shunquan Liu,Yingchang Yang,Yanglong Hou,Jinbo Yang
DOI: https://doi.org/10.1109/tns.2018.2833510
IF: 1.703
2018-01-01
IEEE Transactions on Nuclear Science
Abstract:The compromise between intensity and resolution is the core issue for the powder neutron diffraction techniques. An investigation of the optimum configuration of a neutron instrument is necessary to meet the requirement for both the intensity and resolution before the final construction. In this paper, the performance of the high-intensity powder neutron diffractometer (HIPD) at China Advanced Research Reactor has been studied by Monte Carlo simulation and convolution methods. The virtual detector Monitor-nD with super fine recording elements was used to record the diffraction patterns of the simulated HIPD. In order to introduce the spatial resolution effect of the detector in the diffraction pattern, the convolution method was adopted. The diffraction peaks from the diffraction patterns were then used to calculate the minimum interplanar spacing difference Delta d that the HIPD can distinguish. It is obtained that the minimum resolution of the HIPD is 2.08 x 10(-3) (Delta d/d) for the current design. The resolution of the HIPD calculated above is consistent with the resolution actually measured according to the Ni sample diffraction experiment. This indicates that the above simulation method has high reliability. Finally, a movable 20' Soller collimator and sample cans with 3 and 6 mm diameters are recommended to be used to optimize the ND experiment. After installing the 20' Soller collimator, the minimum resolution can be 1.88 x 10(-3) (Delta d/d). The diffraction intensity of the 6-mm-diameter sample will be almost 3 times higher than that of the 3-mm-diameter sample.
What problem does this paper attempt to address?