Abstract
Due to industrial development, the importance of GD&T tolerance is growing day by day. Roundness measurement means the size deviated from the ideal circle. Roundness evaluation methods include LSCI, MZCI, MCCI, and MICI. Generally, A is used a lot at industrial evaluation. In this experiment, we studied the variations in table velocity, filter values, and detector angles, which can cause errors in roundness measurements. The measurement conditions were table speeds of 10, 30 and 60 mm/s, probe angles of 10, 20 and 30 degrees and frequency filter settings of 15, 150 and 500 upr and The number of experiments was measured 30 and the average value was chosen as a representative value. The hypothesis test showed that the p-value for the frequency filter was greater than 0.05, and the experiment rejected the null hypothesis and adopted the alternative hypothesis.