Inherent to the somewhat uncontrolled nature of the additive process, the surfaces of metal powder bed fusion additively manufactured components tend to be very rough. Large isolated ‘bumps’, as one of the major defect features, are often present due to partially melted particles attached to the surface. An enhanced watershed segmentation method is proposed to separate these ‘bump’ features from the underlying surface texture such that the ‘bumps’ and underlying surface can be quantitatively analysed. The results show that the amplitude roughness parameters of the underlying surface are significantly less than the un-segmented surface and spatial roughness parameters differ between two surfaces. Characterising the extracted underlying surface and ‘bumps’ independently allows better correlation between surface measurements and additive system performance and hence aids in process optimization.
Downloads
Downloads per month over past year