High-resolution surface inspection in restricted spaces: a confocal chromatic solution

Abstract The physical and mechanical properties of manufactured surfaces play a crucial role in determining the performance of machine components. Surface roughness is a key indicator of surface quality and significantly affects machine parts and product functionality. Despite advanced real-time, high-resolution profile measurements, confocal chromatic sensors remain limited in inspecting areas with restricted reflective spaces. This work introduces a confocal chromatic measurement system to inspect surface roughness within confined small-bore fittings. The system incorporates median and Gaussian filters alongside a characterization stage that computes nineteen parameters for comprehensive surface analysis. The proposed system was validated using Flexbar’s Surface Roughness Standards Set calibration plate and as-machined, post-machined, and rejected small-bore fittings. Notably, statistical moments, peak prominence, the autocorrelation function, and the total power percentage between 100 and 150 Hz were found to be particularly effective for surface roughness characterization and estimation. Four inspection models were developed and evaluated. The mathematical equations, R 2 values, norms of residuals, and average errors for R a estimation were presented for each model. Three of the models demonstrated an error of less than 10% in R a estimation for surfaces with an R a value of 0.1 µm or lower.