Robust optimization of multi-probe roundness measurement probe angles

Tuomas Tiainen,Raine Viitala

doi:10.1016/j.measurement.2020.108146

Abstract

Positional errors of a multi-probe roundness measurement frame will result in the probe angles deviating from the nominal angles. The actual probe angles can be accurately determined after placing the measurement frame using cross correlation or other methods. However, since multi-probe methods suffer from harmonic suppression, determining the angles is not enough for accurate reconstruction of the roundness profile, if the actual angles are situated in an area with poor harmonic characteristics. To find suitable areas for the probe angles to allow for deviations from the nominal angles, this paper presents a robust optimization for probe angles to avoid harmonic suppression regardless of errors in the probe angles. Suggestions for optimal probe angles are presented for the three-point and the four-point redundant diameter method.

Highlights

Roundness is an important concept in many fields of engineering
From at least three probe signals obtained from a rotating workpiece (Fig. 1), the methods aim to separate the center point movement and the roundness profile, which both contribute to the probe signals
The results show that using the incorrect angles in the calculation of the roundness profile leads an error in the produced profile, which is demonstrated in Fig. 7 for a simulated case and in Fig. 8 the measurement of the actual profile

Summary

Introduction

Roundness is an important concept in many fields of engineering. Relevant roundness parameters are defined in ISO 12181-1 [1] and standard filters in ISO 12181-2 [2]. When determining roundness profiles of cross sections of large flexible rotors such as paper machine rolls, there are several conditions which restrict the available roundness measurement methods. Rotors can be too large to be placed onto precision spindles, and under rotation the cross section center point movement can be unpredictable and unrepeatable. Under these conditions, multi-probe methods can be used to accurately determine the roundness profile of a cross section. From at least three probe signals obtained from a rotating workpiece (Fig. 1), the methods aim to separate the center point movement ( called spindle error motion [3]) and the roundness profile, which both contribute to the probe signals. Several multi-probe roundness measurement methods have been published [4,5,6,7,8,9]

Methods

Results

Discussion

Conclusion