Characterization of Direct Metal Deposition Printed Copper-Iron Alloys

Abstract

Additive Manufacturing (AM) has been hailed as a disruptive technology, capable of altering significant portions of the manufacturing sector. Its low lead time and geometric flexibility have already made it a key part of prototyping, and an increasing number of industries are looking to integrate AM to lower cost and improve performance. Key to the greater adoption of AM is an increase in the number of materials which can be deposited, and the properties of those materials. With the advent of high-powered lasers, metals AM has become increasingly common, based on the melting of metal powders. This work is an exploration of the effect laser-based AM by the Direct Metal Deposition (DMD) process has on the material properties of Copper (Cu) – Iron (Fe) deposits, across a range of compositions and printing parameters. The Cu-Fe system is notable for its lack of intermediate phases across all temperatures and lack of solubility below 600°C. This dissertation is an exploration into the effects the high cooling rates in laser-based AM (>1000 K/s) has on this material system, and if in-situ sensing methods such as optical spectroscopy can be used for its characterization. Scanning electron microscopy (SEM) shows that even at these high cooling rates, the Cu and Fe samples continued to segregate into a dendritic structure, with a Cu matrix and Fe dendrites. Through the calculation of cooling rates based on an Eagar-Tsai heating and cooling model, it was found that the primary dendrite spacing could be predicted based on the processing conditions, including through changes of deposition speed, composition, and laser power, and followed an inverse square root relationship. Transmission electron microscopy (TEM) of samples at Cu25Fe75 and Cu50Fe50 showed the presence of nanoscale precipitates within the copper matrix and iron dendrites. Both coherent BCC and incoherent FCC nanoprecipitates of Cu were found within the Fe dendrites of the Cu25Fe75 sample, as well as coherent FCC Fe nanoprecipitates within the Cu matrix. Micropillar compression tests show a significant increase in yield strength above the Hall-Petch predicted hardness. This increase in compressive strength could only be accounted for by using a combination of Hall-Petch, shear modulus differences between the coherent precipitates and the bulk Fe or Cu, and the Orowan mechanism between the Fe dendrites and the incoherent FCC Cu precipitates. The BCC Cu precipitates in the Cu25Fe75 sample, which were not present in the Cu50Fe50 sample, led to significant strain hardening, caused by a transformation of the BCC Cu to FCC Cu under compressive strain AM adoption in industry has been slow due to quality control issues, and several in-situ sensing techniques have been explored to sense and improve build quality. In-situ spectroscopy is the only technique that can reasonably collect composition information during a print, as the spectra contains characteristic peaks associated with the composition of the deposited material. Using peak intensity comparison, between the Cu and Fe peaks in the near UV, it is shown that composition can be detected and predicted up to 25% Cu. Above this Cu composition the use of spectroscopy is more limited, as significant self-absorption in the Cu peaks leads to the reversing of the Cu-spectral peaks, and a plateau or reversal of the peak intensity ratios.