Experimental evaluation of post-processing techniques for fdm 3d-printed parts: solvent smoothing vs. laser smoothing, vibratory tumbling, and filler-primer coating

Abstract This study directly compared four established surface-finishing methods for Fused Deposition Modeling (FDM) parts: solvent-based chemical smoothing, CO₂ laser smoothing, vibratory tumbling, and filler-primer coating. The full comparison was performed on ABS and PLA specimens with surface inclinations of 10°–50°, while PVB-based PolySmooth and PolyCast were investigated only as supplementary reference materials for alcohol-based solvent smoothing and were excluded from the four-method comparison and ANOVA. Surface quality was evaluated using optical interferometry and ISO 25,178 areal surface-texture parameters. The as-printed S a ranges were 10.24–18.92 μm for ABS and 17.88–25.40 μm for PLA. Solvent smoothing provided the greatest roughness reduction on compatible polymers, reducing S a by 98.6–99.2% for ABS and 89.0–98.8% for PLA, while the supplementary PVB tests achieved average reductions of up to 99.56%. Surface conditions differed significantly at every investigated inclination for both S a and S t (Welch ANOVA, all Holm-adjusted p < 0.001). Pairwise Welch tests with Holm correction confirmed lower S a after chemical smoothing than after every other surface condition at all inclinations for both polymers. A comparative cumulative-cost model estimated average variable costs excluding labor and equipment investment of approximately 1.70, 0.64, 0.57, and 0.25 CZK per part for solvent smoothing, CO₂ laser smoothing, filler-primer coating, and vibratory tumbling, respectively, under the stated nominal batch assumptions. The results demonstrate that process selection requires balancing achievable surface quality against material compatibility, dimensional preservation, production volume, safety requirements, and economic constraints. The findings are limited to the investigated materials, specimen geometry, and selected post-processing conditions, without systematic process-parameter optimization or quantitative evaluation of dimensional and mechanical changes.