Skip to main navigation Skip to search Skip to main content

Precise freeform surface form measurement based on centering deviation calibration and compensation

  • Beijing Institute of Technology

Research output: Contribution to journal › Article › peer-review

Abstract

Spiral scanning profilometry is a mainstream method for freeform surface measurement. However, its measurement accuracy is affected by the centering deviation between the probe optical axis and the rotation spindle. This paper proposes a centering deviation compensation method integrating static calibration and dynamic self-calibration to improve the measurement accuracy of freeform surfaces. This method employs an inclined planar reference object for static calibration of the initial system centering deviation. A dynamic self-calibration strategy is then employed to identify and compensate for the residual centering deviation caused by axis homing, repositioning, and other nonrepeatable effects. Two overlapping datasets with opposite radial coordinates are acquired through center-passing spiral scanning and mutually registered through iterative optimization, thereby suppressing centering-induced reconstruction errors without direct matching to the nominal surface model. Experimental results show that the PV deviation from the reference value is reduced by 60% after compensation. The PV repeatability is better than 35 nm (3σ), and the RMS repeatability is better than 13 nm (3σ). The proposed method provides a feasible technical route for high-accuracy freeform surface form measurement.

Original languageEnglish
Article number123475
JournalMeasurement: Journal of the International Measurement Confederation
Volume292
DOIs
Publication statusPublished - 15 Dec 2026
Externally publishedYes

Keywords

  • Centering deviation
  • Differential confocal
  • Error compensation
  • Freeform surface measurement

Fingerprint

Dive into the research topics of 'Precise freeform surface form measurement based on centering deviation calibration and compensation'. Together they form a unique fingerprint.

Cite this