Data distribution analysis on the reversibility in standard torque measurement

Abstract In this paper, the influence of reversibility on the uncertainty of torque measurement is studied. The reversibility distribution of the standard torque sensor is not a known common distribution, so its coverage factor k cannot be determined like other well-defined distribution. In order to explore the reversibility characteristics of the torque sensor, the original definition of coverage factor will be used for calculation and analysis. By using discrete and continuous analysis methods, we discuss the actual distribution characteristics of torque applied in increasing and decreasing steps, and then calculate the coverage interval, and finally determine the coverage factor k. By kernel density estimation, the discrete random variable is transformed into continuous random variable, and the probability density function of unknown distribution is obtained. The center method and the shortest coverage interval are used for the discussion. Finally, the coverage factors in 95% and 99% confidence intervals are 1.794 and 2.439, respectively.