中文摘要：

为了研究采样频率和样本容量对明渠紊流统计特征值的影响，利用粒子图像测速技术，对明渠恒定均匀水流的二维流速分布进行测量，获得了包含340535个样本的流场时间序列。对该大容量时间序列进行抽样，得到不同采样频率和容量条件下的一系列子序列。对比分析各子序列的时均流速、紊动强度及雷诺应力的变化特点，发现存在临界的采样频率和样本容量，紊动统计特征参数在临界值以下波动较大，而在临界值以上基本趋于稳定；要满足同一测量精度，时均流速对采样频率和样本容量的要求最低，紊动强度次之，雷诺应力最高。

英文摘要：

The influences of sampling frequency and sample size on the accuracy of turbulence measurements are experimen- tally investigated. A set of 340 535 velocity data are first sampled at a frequency of 600 Hz with the particle image velocimetry to measure the turbulent parameters of the flow. A series of statistical samples corresponding to various sampling frequencies and sample sizes are then achieved by extracting data from the original data set. Comparison of the mean velocity, turbulent in- tensity, and Reynolds stress obtained from these samples reveals that sampling frequency and sample size both show a threshold value above which the measured turbulence characteristics become relatively stable, and that, to achieve the same level of the experimental accuracy, the required sampling frequencies and sample sizes increase progressively in evaluating mean velocity, turbulence intensity and Reynolds stress.