Experimental investigation and prediction of surface roughness in abrasive flow finishing of additive manufactured pure copper

期刊论文
发表日期：

2024年8月17日

查看全文

Mahaboob Basha Shaik,
Ravi Sankar Mamilla,
Venkaiah Nasina

摘要

Atomic diffusion additive manufacturing (ADAM) is a newly emerged metal extrusion-based additive manufacturing (AM) process where a thermoplastic filament embedded with tiny metallic particles is used to build the part. Similar to other metal AM processes, the parts built using the ADAM process also suffer from high surface roughness, which necessitates the deployment of further post-processing to enhance the surface finish of the parts before their intended application. Abrasive flow finishing (AFF) is an advanced finishing technique that involves reciprocating a visco-elastic polymer abrasive medium across surfaces to achieve the desired finish. In this work, an indigenously developed natural polymer-based abrasive medium was employed to reduce the surface roughness of ADAM pure copper using the AFF process. The influence of extrusion pressure, number of finishing cycles, concentration, and size of the abrasive particles on the percentage change in surface roughness in the longitudinal ( $\% \, \Delta R_{a}^{{{\mathbf{||}}}}$ $% Δ R_{a}^{| |}$ ) and lateral directions ( $\% \, \Delta R_{a}^{{\mathbf{ \bot }}}$ $% Δ R_{a}^{⊥}$ ) was explored and modelled using a second-order linear regression equation. Artificial neural networks (ANN) were also employed to forecast the $\% \, \Delta R_{a}^{{{\mathbf{||}}}}$ $% Δ R_{a}^{| |}$ and $\% \, \Delta R_{a}^{{\mathbf{ \bot }}}$ $% Δ R_{a}^{⊥}$ of the finished part and found that the prediction accuracy of trained ANN was better than that of the modelled regression equations. In addition, the analysis of variance revealed that the number of cycles is the most contributing parameter for $\% \, \Delta R_{a}^{{{\mathbf{||}}}}$ $% Δ R_{a}^{| |}$ and $\% \, \Delta R_{a}^{{\mathbf{ \bot }}}$ $% Δ R_{a}^{⊥}$ followed by concentration, abrasive particle size, and extrusion pressure for $\% \, \Delta R_{a}^{{{\mathbf{||}}}}$ $% Δ R_{a}^{| |}$ and abrasive particle size, extrusion pressure, and concentration for $\% \, \Delta R_{a}^{{\mathbf{ \bot }}}$ $% Δ R_{a}^{⊥}$ , respectively. A maximum $\% \, \Delta R_{a}^{{{\mathbf{||}}}}$ $% Δ R_{a}^{| |}$ of 88.15% and $\% \, \Delta R_{a}^{{\mathbf{ \bot }}}$ $% Δ R_{a}^{⊥}$ of 63.05% was obtained when the parts were subjected to 300 AFF cycles using an abrasive medium containing 50 wt% concentration of 255 µm sized abrasive particles at 8 MPa extrusion pressure.

Graphical abstract