Research on the Evaluation of Uncertainty in Roundness Measurement

Accurate evaluation of measurement uncertainty is crucial for precision manufacturing. This paper proposes a two-stage Bayesian-Monte Carlo method for assessing roundness measurement uncertainty in online inspection. The method separates machining errors from measurement system errors by first establishing a prior distribution via calibration with a standard artifact and then updating it with workpiece measurements. To validate the method, measurements were conducted on a certified roundness standard, showing close agreement with the reference value. The method demonstrates effective uncertainty quantification with small sample sizes and provides a foundation for intelligent evaluation in dimensional metrology.