Tribo-acoustic informatics of FDM printed PLA surfaces under dry sliding: the coupled influence of printing parameters and surface structures

Abstract

Fused deposition modelling (FDM) produces polymer surfaces whose layer-wise morphology, together with printing parameters, modulates frictional contact and tribo-acoustic emissions. This paper develops a data-driven tribo-acoustic workflow for FDM printed PLA under dry sliding and derives predictive rules for low-noise surface design within the tested process domain. Across five surface types, Plain, Triangle, Hexagonal, Straight Brush, and Curve Brush, the coefficient of friction (CoF) spanned 0.13 to 0.22 and the RMS sound pressure spanned 4.31 to 13.65 Pa. Analysis of variance shows that layer thickness is the dominant factor for both CoF and RMS noise, while printing speed is marginal for CoF and extrusion width is weak. Contrary to common intuition, periodic surface textures do not guarantee noise reduction and can amplify tonal components under unfavourable parameter combinations. Quadratic response surface models achieved R-squared values of 0.826 for coefficient of friction and 0.932 for RMS sound pressure, and validation tests yielded average prediction errors of 7.40% and 7.43%, confirming sub 10% accuracy in practice. Taguchi signal-to-noise ratio analysis for the Plain surface identifies an optimal parameter set for minimising friction and noise, with a layer thickness of 0.2 mm, a printing speed of 40 mm per second, and an extrusion width of 0.6 mm. The curated dataset and predictive models provide a basis for tribo-informatics-oriented optimisation of process settings to mitigate friction induced noise in additively manufactured polymer interfaces.