[SUBMITTED] A Complete Re-evaluation of the Accuracy of the Standardised Roundness Evaluation Methods

Roundness evaluation is the process of measuring how well an object conforms to a circle, a ubiquitous process in the manufacture of any rotating part, and the evaluation of rounded objects throughout science and engineering. This evaluation is a standardised process, with four available methods recommended by standardisation bodies since before the advent of digital measurement: Minimum Circumscribed Circle (MCC), Maximum Inscribed Circle (MIC), Minimum Zone Circle (MZC), and Least Squares Circle (LSC). The age of these methods, however, limited the available technologies to assess their validity during their inception, nor allowed for other practical alternatives. This assessment is finally visited in this work, using modern capabilities to challenges some long-standing assumptions about their applicability and accuracy. This is achieved using virtual instrumentation to replicate physical measurements in a controlled manner, while precisely knowing the initial roundness parameters, a process which was unavailable during these standards creation. Through this analysis, this work expands upon previous unquantified observations, showing that the majority of the evaluation methods actually underestimate the roundness deviation. In particular, this underestimation is shown to be highest with the commonly-preferred MZC method, which was often believed to be the most precise method. The uncertainty sources each method will contribute to a measurement are quantified, and systematic errors are highlighted, showing less functional difference between the evaluation methods than expected.