Document Type

Thesis - Open Access

Award Date

2026

Degree Name

Master of Science (MS)

Department / School

Mechanical Engineering

First Advisor

Todd Letcher

Abstract

Lost PLA investment casting offers a low tooling route for producing complex aluminum parts, but repeatability depends on material use, burnout behavior, geometric process limits, and dimensional response. This thesis evaluated Lost PLA casting as a resource efficient manufacturing method for complex A356 aluminum components in resource constrained settings through four linked objectives: metal reusability, burnout and PLA selection, process limitations, and feature based dimensional shrinkage. A metal mass balance was performed on six casting trials using salvaged A356. Ash residual behavior was studied through a single tray screening test of eight commercial filaments and three invested mold burnout trials using the selected low residue PLA. Process capability and dimensional response were evaluated using three coupon versions containing open, blind, internal, external, thin wall, cylindrical, and thickness-controlled features. Supporting thermal interpretation was provided through a MATLAB burnout schedule tool, thermocouple measurements, and STAR-CCM+ simulations. The mass balance results showed that most charged aluminum remained in a recoverable solid form after each trial, with high pour yield, low dross fraction, and mass balance closure near 100 percent. In the filament screening study, Overture Easy PLA Natural produced the lowest visible residue among the tested PLA materials, while PolyCast provided the cleanest overall burnout as a specialty casting filament. Invested mold trials using Overture Easy PLA Natural and a staged schedule at 300 °F (149 °C), 700 °F (371 °C), and 1350 °F (732 °C) produced generally intact molds with retained feature definition and only limited residual ash. Thermocouple results showed that the chamber air tracked the programmed furnace schedule most closely, while the steel flask and intruded mold exhibited substantial thermal lag, confirming that the generated MATLAB profile should be interpreted as a furnace schedule rather than a direct intruded mold temperature prediction. The feature-based results showed that open and externally accessible geometries were reproduced more reliably than blind or enclosed features. Blind recesses, thin intruded walls, and small intruded cylinders were the most restrictive geometries, and the first consistently survivable intruded wall thickness in this study was 0.625 mm. Dimensional response was not uniform across feature families. Intruded features and thickness-controlled regions showed greater dimensional sensitivity and variability than comparable extruded or outline features. Overall, the study demonstrates that Lost PLA casting can function as a materially efficient and practically useful process when low residue PLA is paired with controlled staged burnout, favorable feature openness, and conservative expectations for thin enclosed geometry. These results provide practical guidance for burnout selection, geometry design, and dimensional interpretation in Lost PLA casting workflows.

Publisher

South Dakota State University

Download

Share

COinS
 

Rights Statement

In Copyright