Search for: All records

This work concerns process monitoring in the laser powder bed fusion additive manufacturing process. In this work, we developed and applied a novel in-situ solution for process stability monitoring and flaw detection using acoustic emission sensing. Current process monitoring methods in laser powder bed fusion only focus on the top surface of the deposition process, using an array of sensors to capture data on a layer-by-layer basis. Common sensors used for in-situ monitoring of the laser powder bed fusion process are optical, infrared, and highspeed imaging cameras along with pyrometers and photodiodes. A critical flaw with traditional top surface monitoring methodologies is that they are unable to reliably monitor the subsurface phenomena that occur in the laser powder bed fusion process. These subsurface effects are caused by the meltpool penetrating multiple layers below the top surface, leading to the re-solidification of the microstructure and potentially generating keyhole porosity. By only monitoring the top surface of the laser powder bed fusion process, the meltpool depth aspects and effects are ignored. To overcome the limitations of current in-situ monitoring of subsurface effects, this work utilizes four passive acoustic emission sensors attached to the build plate. These acoustic emission sensors monitor the energy emissions generated from the surface-level laser material interactions. Moreover, the acoustic emission signals are capable of traveling through the previously deposited layers, through the build plate, and to the sensors. Therefore, the acoustic waveform generated by the laser can capture process phenomena ranging from the crystallographic level to the macro-scale layer level which are at the root of flaw formation inside the deposited part. Hence, acoustic emission monitoring has the ability to monitor the subsurface effects in the laser powder bed fusion process. To monitor and analyze this acoustic waveform, novel wavelet-based decomposition is combined with heterogeneous sensor fusion to not only capture the acoustic waveform in time, but also in locational space on the build plate. Locational acoustic emission data enables the ability to determine the source of the generated acoustic waveform which is advantageous when the location of flaws is desired. This extracted spatially placed acoustic waveform data is able to detect the effect of processing parameters with a statistical fidelity of 99%. The proposed locational acoustic waveform monitoring method correlates to the resulting surface roughness of manufactured samples with a fidelity of 86%. Additionally, we show that acoustic waveform monitoring detects the onset of part failure, recoater crashes, and warpage prior a priori to the actual failure point. 

The long-term goal of this work is to predict and control the microstructure evolution in metal additive manufacturing processes. In pursuit of this goal, we developed and applied an approach which combines physics-based thermal modeling with data-driven machine learning to predict two important microstructure-related characteristics, namely, the meltpool depth and primary dendritic arm spacing in Nickel Alloy 718 parts made using the laser powder bed fusion (LPBF) process. Microstructure characteristics are critical determinants of functional physical properties, e.g., yield strength and fatigue life. Currently, the microstructure of LPBF parts is optimized through a cumbersome build-and-characterize empirical approach. Rapid and accurate models for predicting microstructure evolution are therefore valuable to reduce process development time and achieve consistent properties. However, owing to their computational complexity, existing physics-based models for predicting the microstructure evolution are limited to a few layers, and are challenging to scale to practical parts. This paper addresses the aforementioned research gap via a novel physics and data integrated modeling approach. The approach consists of two steps. First, a rapid, part-level computational thermal model was used to predict the temperature distribution and cooling rate in the entire part before it was printed. Second, the foregoing physics-based thermal history quantifiers were used as inputs to a simple machine learning model (support vector machine) trained to predict the meltpool depth and primary dendritic arm spacing based on empirical materials characterization data. As an example of its efficacy, when tested on a separate set of samples from a different build, the approach predicted the primary dendritic arm spacing with root mean squared error ≈ 110 nm. This work thus presents an avenue for future physics-based optimization and control of microstructure evolution in LPBF. 

The objective of this work is to predict a type of thermal-induced process failure called recoater crash that occurs frequently during laser powder bed fusion (LPBF) additive manufacturing. Rapid and accurate thermomechanical simulations are valuable for LPBF practitioners to identify and correct potential issues in the part design and processing conditions that may cause recoater crashes. In this work, to predict the likelihood of a recoater crash (recoater contact or impact) we develop and apply a computationally efficient thermomechanical modeling approach based on graph theory. The accuracy and computational efficiency of the approach is demonstrated by comparison with both non-proprietary finite element analysis (Abaqus), and a proprietary LPBF simulation software (Autodesk Netfabb). Based on both numerical (verification) and experimental (validation) studies, the proposed approach is found to be 5 to 6 times faster than the non-proprietary finite element modeling and has the same order of computational time as a commercial simulation software (Netfabb) without sacrificing prediction accuracy.