目的 探讨抽球模型预测粉体最小取样量的可行性。方法 将粉体粒子抽象为小球，建立抽球模型，推导出能够代表所测粉体的最小取样量；并以微晶纤维素（MCC）、淀粉2种不同粒径的药用辅料为模型粉体，测定取样量分别为5、10、100 mg时的粒径分布，用粒径分布的重合性判断取样的均匀性。结果 根据抽球模型推算出MCC的最小取样量大于5 mg，淀粉的最小取样量小于5 mg，与实验结果一致；当取样量为100 mg时，MCC的样本也能代表总体。结论 用抽球模型研究粉体的最小取样量可行，粉体粒径越小，最小取样量越小。

Objective To explore the feasibility of using drawing ball model to predict the minimum sampling mass of powders. Methods A drawing ball model was established by abstracting the particle of powders as small balls. Then the minimum sampling mass which representing the measured powders was calculated by the model. Microcrystalline cellulose (MCC) and starch were used as model powders and then the particle size distribution curves of them were determined at different sampling mass of 5, 10, and 100 mg. The homogeneity of samples was judged by the coincidence of the particle size distribution curves. Results According to the drawing ball model, the minimum sampling mass of MCC was more than 5 mg and that of the starch was less than 5 mg, which were consistent with the results of verification experiment. When the sampling mass was 100 mg, it could also represent the overall MCC. Conclusion Drawing ball model is feasible to investigate the minimum sampling mass of powders. The smaller the particle size of powders is, the less the minimum sampling mass would be.

国家自然科学基金资助项目（81160522）;赣鄱英才555工程领军人才培养计划（赣财教指 [2013] 296号）;江西省教育厅科研项目（GJJ13584）;江西中医药大学重点学科青年教师培养计划（2012jzzdxk020）