Additive manufacturing (AM) or 3D printing of metals is capable of dramatically reducing material waste, improving production times, and forming complex components and structures. However, to move beyond prototyping to critical load bearing applications, validated and verified material components must be reliably manufactured. The proposed research aims to establish a methodology for characterizing direct energy deposited metals by linking processing variables to the resulting microstructure and subsequent material properties. Fundamentally, predicted material properties, damage growth, and life assessment for the AM of structurally critical components must be comparable to those produced using traditional methods. In order to achieve a suitable predictive model, a large sample size is required in each material for accurate statistical modeling due to the plethora of key variables influencing the build. Reported variability in AM materials is a function of raw material properties, processing conditions, and build geometry, but especially of the thermal cycling during the build process. An in-house metal deposition machine has been developed at Lehigh University, where process monitoring captures real-time input power, temperature, travel speed, and wire feed rates at each location within the deposited material. In turn, the proposed research will assess and characterize the build microstructure and material properties of the resulting builds. Specifically, the properties of interest in the printed materials are yield strength, fracture toughness, and wear properties. After these are adequately characterized, an extensive experimental program for fatigue life is planned. Quality metrics, given the geometric location, will be related to the processing conditions and resultant microstructure. Ultimately, a process-microstructure-property model is envisioned (and validated with probabilistic modeling) for reliability and life prediction of printed materials. The proposed research will initially use Stainless Steel (e.g., ER308L) and Carbon Steel (e.g., ER70S-6). The expectation is that sufficient insight will be acquired such that a limited characterization will be required for subsequent materials considered, e.g., other alloy steels, nickel alloys, or aluminums.
Compared to powder based metal AM, wire fed metal deposition using arc welding or laser deposition is still in the initial stages of development, and basic fundamental research is still in its infancy. Because wire fed AM systems offer rapid deposition rates, rely on commercially available wire material stocks, have low risk of contamination, and offer reduced base material loss, they have tremendous potential for certain applications (e.g., traditionally large and complex cast parts and inventory reduction problems). Although the use of welding-related processes for AM is still relatively new, there exists a tremendous body of knowledge from research conducted on fusion welding and similar processes that can be exploited in order to advance AM processes. The research has advanced to the point in which sophisticated numerical codes are now commercially available to model the heat and fluid flow that controls the thermal cycles (e.g., Sysweld®) along with the complex phase transformations (both solidification and solid state phase transformations) which in turn control the microstructure and resultant properties. Examples of microstructure models that are available include ThermoCalc®, Dictra®, and MatCalc®, to name a few. These codes are capable of predicting a wide range of microstructural features associated with solidification of the initial melt pool as well as subsequent solid state transformations that may occur after solidification and re-heating of subsequent weld passes. Lehigh has used these codes extensively for prediction and control of the microstructure and properties of fusion welds and AM builds made with Laser Engineered Net Shaping®, and is well-positioned to continue this effort for the arc welding based system currently under development. The structure of the AM system being developed is shown in Fig. 1. The underlying design was based on open source specifications, but has been modified for big metal applications. Figure 2 shows a single bead being deposited on a plate. The system incorporates a variety of sensors to monitor variables such as power, wire feed rate, and thermal history, for example, for the deposition process in real-time. Consequently, the output yields large datasets for processing conditions, depending on the data capture rates. The proposed experimental program will produce a large number of specimens machined from single and triple weld bead wall geometries. The specimens will be prepared according to the ASTM Standard E8-04 for tensile testing. The builds will be single and triple bead walls approximately 0.3 in wide for the single bead or 3×0.3 in wide for the triple bead, 7 in high, and 27 in long. Figure 3 shows an early proof of concept for a single bead wall that demonstrates the large build sizes that are readily accommodated with this system. Similarly, Figures 4 – 6 show preliminary tests to develop an understanding of the printed material. Figure 4 is a thermal image for temperature profile monitoring. The thermal history acquired with these measurements can be used for validation and refinement of heat and fluid flow codes for predicting and controlling heating and cooling rates that are essential for microstructure control. Figure 5 shows an x-ray of the build to investigate porosity, overlaid with a schematic of tensile specimen locations within the build. Figure 6 is a light optical photomicrograph from a stainless steel build that shows a grain boundary at adjacent passes within the build. The microstructure exhibits a cellular substructure that consists of an austenite matrix with ferrite particles (dark phase). The substructural orientation changes at the grain boundary due to epitaxial growth from the underlying weld passes. The type and amount of phases, substructure scale and orientation, and grain size will all be controlled by the process parameters and resultant thermal history. Further, micrographs of focus-ion-beam-milled cross sections beneath wear scars on AM stainless steel samples reveal grain refinement of the surface and subsurface microstructures. A direct link between subsurface microstructure, microstructural evolution and wear can be exploited by the unique ability of GMAW AM to control microstructure and texture and design surfaces for improved wear performance. Preliminary wear results suggest an improvement in wear performance for end-of-line prints.These data indicate that further extensive investigation is needed and warranted. Consequently, the proposed research will include the following: extensive x-ray imaging for estimation of porosity magnitude and location; tribology testing for directional wear resistance; optical and scanning electron microscopy for microstructural characterization; microhardness array testing; fracture toughness; yield and ultimate strength from stress-strain testing; and fatigue characterization.