The quality of powder bed fusion (PBF) built parts is highly correlated to the melt pool characteristics. Camera-based coaxial melt pool monitoring (MPM) is widely applied today because it provides high-resolution monitoring on the time and length scales necessary for deep PBF process understanding, in-process defect detection, and real-time control. For such functions, MPM data has to be registered correctly to a well-defined coordinate system. This paper presents methods for camera-based coaxial melt pool monitoring (MPM) data registration using the build volume coordinate system defined in ISO/ASTM52921, for both open architecture AM systems and 3rd party MPM augmented closed commercial systems. Uncertainties are evaluated for the proposed methods and case studies provided to demonstrate the effectiveness of the methods.

In laser powder bed fusion metal additive manufacturing, insufficient shield gas flow allows accumulation of condensate and ejecta above the build plane and in the beam path. These process byproducts are associated with beam obstruction, attenuation, and thermal lensing, which then lead to lack of fusion and other defects. Furthermore, lack of gas flow can allow excessive amounts of ejecta to redeposit onto the build surface or powder bed, causing further part defects. The current investigation was a preliminary study on how gas flow velocity and direction affect laser delivery to a bare substrate of Nickel Alloy 625 (IN625) in the National Institute of Standards and Technology (NIST) Additive Manufacturing Metrology Testbed (AMMT). Melt tracks were formed under several gas flow speeds, gas flow directions, and energy densities. The tracks were then cross-sectioned and measured. The melt track aspect ratio and aspect ratio coefficient of variation (CV) were reported as a function of gas flow speed and direction. It was found that a mean gas flow velocity of 6.7 m/s from a nozzle 6.35 mm in diameter was sufficient to reduce meltpool aspect ratio CV to less than 15 %. Real-time inline hotspot area and its CV were evaluated as a process monitoring signature for identifying poor laser delivery due to inadequate gas flow. It was found that inline hotspot size could be used to distinguish between conduction mode and transition mode processes, but became diminishingly sensitive as applied laser energy density increased toward keyhole mode. Increased hotspot size CV (associated with inadequate gas flow) was associated with an increased meltpool aspect ratio CV. Finally, it was found that use of the inline hotspot CV showed a bias toward higher CV values when the laser was scanned nominally toward the gas flow, which indicates that this bias must be considered in order to use hotspot area CV as a process monitoring signature. This study concludes that gas flow speed and direction have important ramifications for both laser delivery and process monitoring.

Melt pool size is a critical intermediate measure that reflects the outcome of a laser powder bed fusion process setting. Reliable melt pool predictions prior to builds can help users to evaluate potential part defects such as lack of fusion and over melting. This paper develops a layer-wise Neighboring-Effect Modeling (L-NBEM) method to predict melt pool size for 3D builds. The proposed method employs a feedforward neural network model with ten layer-wise and track-wise input variables. An experimental build using a spiral concentrating scan pattern with varying laser power was conducted on the Additive Manufacturing Metrology Testbed at the National Institute of Standards and Technology. Training and validation data were collected from 21 completed layers of the build, with 6,192,495 digital commands and 118,928 in-situ melt pool coaxial images. The L-NBEM model using the neural network approach demonstrates a better performance of average predictive error (12.12%) by leave-one-out cross-validation method, which is lower than the benchmark NBEM model (15.23%), and the traditional power-velocity model (19.41%).

The quality of AM built parts is highly correlated to the melt pool characteristics. Hence melt pool monitoring and control can potentially improve AM part quality. This paper presents a neighboring-effect modeling method (NBEM) that uses scan strategy to predict melt pool size. The prediction model can further assist in scan strategy optimization and real-time process control. The structure of the proposed model is formulated based on the physical principles of melt pool formation, while experimental data is used to identify the optimal coefficients. Compared to the traditional power-velocity prediction models, NBEM model introduces the cumulative neighboring-effect factors as additional input variables. These factors represent the neighborhood impact of scan path on the focal point melt pool formation from time and distance perspective. Two experiments use different scan strategies to collect in-situ measurements of melt pool size for model construction and validation. By introducing the neighboring-effect factors, the global Normalized Root Mean Square Error (NRMSE) is improved from around 0.10 to 0.08. More importantly, the local NRMSE of irregular melt pool area prediction is improved to around 0.15 for more than 50% improvement. The case studies verify that the proposed method can predict the melt pool variations by seamlessly integrating scan strategy considerations.

A study has been made of deformation, forces and energy in modulation-assisted machining (MAM), wherein chip formation occurs in the presence of a controlled, low-frequency modulation superimposed on to the machining. A unique feature of the study is the use of high speed in situ imaging and image analysis to map material flow in the chip formation zone at high resolution; and simultaneous measurements of tool motions and forces, such that the instantaneous forces can be overlaid onto the chip formation process. The measurements show that the observed significant reductions in specific energy in MAM relative to conventional machining, when cutting ductile metals such as copper and Al 6061T6, are a consequence of chip formation with reduced strain levels in MAM. Additional insights into the chip formation are obtained by examining the effects of a chip aspect ratio parameter.

The controlled application of low-frequency modulation to machining — Modulation Assisted Machining (MAM) — effects discrete chip formation and disrupts the severe contact condition at the tool-chip interface. The role of modulation in reducing the specific energy of machining with ductile alloys is demonstrated using direct force measurements. The observed changes in energy dissipation are analyzed and explained, based on the mechanics of chip formation.