Scholarlink

<p>Fused deposition modeling (FDM) is a fascinating 3D printing method that produces complicated items at low cost and with less wasteful material usage. An intrinsic feature of FDM parts is their poor surface smoothness, which must be managed to increase the industrial use of FDM. The current study adopts Taguchi's L18 orthogonal array to optimize six different FDM parameters, including layer height, extrusion temperature, bed temperature, print speed, raster angle and part orientation, with an objective of minimizing both surface roughness and printing time. Cuboid-shaped samples with circular hole at center in accordance with ASTM standard D-2240 have been considered as experimental sample prepared from nylon filaments reinforced with 8% aramid composites. The measured values of roughness average (Ra), roughness height (Rz) and printing time (PT) have been analyzed through techniques, such as signal-to-noise ratio method, ANOVA, CRITIC-CoCoSo and regression machine learning algorithms. Through CRITIC-CoCoSo, the optimal parameter level which enhances the appraisal score is 0.22&#xa0;mm layer thickness, 235&#xa0;°C extrusion temperature, 100&#xa0;°C print bed temperature, 40&#xa0;mm/s print speed, 90° raster angle and upright positioned printing. Adaboost algorithm outperformed other regression algorithms with higher R-sq value of 0.99 representing superior performance and decision tree structure developed through has good correlation with ANOVA outcomes and response Table rankings. ANOVA statistical analysis highlights part orientation and print speed as significant parameter for roughness average with 33.76% and 22.29% contribution, part orientation with significant contribution of 39.69% over roughness height, printing time affected by layer thickness and print speed by 62.34% and 28.16%, respectively. Both part orientation and printing speed are effective over final appraisal score with 43.88% and 18.26%, respectively. Sensitivity analysis performed by varying criteria weight values and techniques, such as TOPSIS, MABAC and WASPAS, has represented a strong positive correlation ranging between 0.84 and 0.99 for alternative ranking. Surface and structural examination through FESEM ensure the presence of voids, pores, semi-solidified material and randomly oriented aramid fibers in printed samples.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Surface Roughness and Printing Time Minimization in 3D Printed Aramid Fiber Reinforced Polyamide Parts through Taguchi-CoCoSo-Machine Learning Techniques

  • 期刊论文

  • 发表日期：

    2025年5月8日

查看全文
  • N. Mohammed Raffic,
  • K. Ganesh Babu,
  • S. Dharani Kumar,
  • B. K. Parrthipan

摘要

Fused deposition modeling (FDM) is a fascinating 3D printing method that produces complicated items at low cost and with less wasteful material usage. An intrinsic feature of FDM parts is their poor surface smoothness, which must be managed to increase the industrial use of FDM. The current study adopts Taguchi's L18 orthogonal array to optimize six different FDM parameters, including layer height, extrusion temperature, bed temperature, print speed, raster angle and part orientation, with an objective of minimizing both surface roughness and printing time. Cuboid-shaped samples with circular hole at center in accordance with ASTM standard D-2240 have been considered as experimental sample prepared from nylon filaments reinforced with 8% aramid composites. The measured values of roughness average (Ra), roughness height (Rz) and printing time (PT) have been analyzed through techniques, such as signal-to-noise ratio method, ANOVA, CRITIC-CoCoSo and regression machine learning algorithms. Through CRITIC-CoCoSo, the optimal parameter level which enhances the appraisal score is 0.22 mm layer thickness, 235 °C extrusion temperature, 100 °C print bed temperature, 40 mm/s print speed, 90° raster angle and upright positioned printing. Adaboost algorithm outperformed other regression algorithms with higher R-sq value of 0.99 representing superior performance and decision tree structure developed through has good correlation with ANOVA outcomes and response Table rankings. ANOVA statistical analysis highlights part orientation and print speed as significant parameter for roughness average with 33.76% and 22.29% contribution, part orientation with significant contribution of 39.69% over roughness height, printing time affected by layer thickness and print speed by 62.34% and 28.16%, respectively. Both part orientation and printing speed are effective over final appraisal score with 43.88% and 18.26%, respectively. Sensitivity analysis performed by varying criteria weight values and techniques, such as TOPSIS, MABAC and WASPAS, has represented a strong positive correlation ranging between 0.84 and 0.99 for alternative ranking. Surface and structural examination through FESEM ensure the presence of voids, pores, semi-solidified material and randomly oriented aramid fibers in printed samples.