Advanced Machines, Monitoring, and Control for AM

Summary

Metal Laser Powder Bed Fusion Additive Manufacturing (AM) has revolutionized the manufacturing industry by directly producing functional metal components from three-dimensional digital designs in a single process. As a true enabler of for Industry 4.0, it seamlessly integrates with the concept of digital transformation of manufacturing through smart and autonomous systems fueled by data and machine learning. Despite its potential, widespread adoption of AM technology faces two major obstacles - inconsistent part quality and low production efficiency. This project will address these challenges by developing and implementing advanced AM control and monitoring methods and demonstrate their positive impact on enhancing part quality and efficiency by integrating these innovations into AM machines/testbeds.

Description

Objective
To develop, integrate, and implement real-time feedback and other cutting-edge process control techniques using advanced monitoring tools such as high-speed imaging and diffraction for single and multi-laser AM systems.

Technical Idea
The current methods for AM control predominantly rely on machine tool control technology, which is fundamentally different from AM's additive nature. Unlike AM, machining is a subtractive process that begins with a homogeneous bulk material, making it relatively insensitive to machine parameter changes such as spindle speed or feed. Consequently, machine tools utilize line-wise control, defining machining parameters through G-code and M-code with constant feed and spindle speed. However, this approach lacks the ability to adapt continuously to local thermal conditions, thus limiting the development of advanced scan strategies and hindering quality improvements.

To address this challenge, NIST has developed a pointwise AM control method that allows for continuous variation in laser power and speed. This pointwise control delves one level deeper, transitioning from one-dimensional (1D) lines to zero-dimensional (0D) points through different power and path modes, along with motion control. The user-accessible pointwise control file can be further processed by analytical, physics, and machine learning models, creating a seamless connection to industry 4.0's digital transformation. 

An essential aspect of the pointwise control is its dual role as both an advanced laser control and an advanced measurement platform. The point-by-point data registration enables synchronization of data from machine input commands to in-situ process monitoring sensors, facilitating comprehensive study of the AM process and optimizing process control. This novel method unlocks new possibilities for refining the AM process and achieving exceptional outcomes. Its versatility and precision pave the way for significant advancements in the additive manufacturing domain.

Pointwise control was initially implemented on the NIST Additive Manufacturing Metrology Testbed (AMMT). It enables the creation of numerous high-quality datasets, including AM-Bench 2022. Subsequently, the pointwise control was applied in the development of the Laser Processing Diffraction Testbed (LPDT), a portable testbed capable of conducting AM experiments under Synchrotron X-ray. As the pointwise control framework matures through these developments and implementations, along with established control algorithms and technology, we are now ready to apply this pointwise AM control to address the various issues described in the Problem Statement.  The products/deliverables from this project includes three major categories: 

1. Enhanced AM Research Platforms: This will include AMMT 2.0 and advancements to LPDT. AMMT 2.0, which is an evolved version of the existing AMMT system, will be equipped with real-time feedback control and dual laser capabilities. Meanwhile, three-dimensional (3D) building and coaxial imaging functionalities will be added to LPDT.
2. Innovative Control/Monitoring Methods: The focus here will be on the production of dimensionally accurate and defect-free components with functional graded properties. By having the laser power dynamically adjusted according to local thermal conditions—either through a future model-based scan strategy (feedforward) or real-time feedback control—issues like lack of fusion defects and keyholing pores will be effectively circumvented. Flexibility in scanning paths and patterns will be provided, allowing for the fabrication of distinctive structures such as honeycombs. Furthermore, the ideal temperature gradient and cooling rate will be established, fostering the growth of preferred phases/microstructures.
3. High-quality Data Production: Data generated from the current AM control studies has been recognized by stakeholders as a valuable resource. This data serves as a foundation for further research, optimization, model development, and demonstrations of AM data management. The initiatives described in sections 1 and 2 will be integral to the consistent generation of such data as a natural outcome of the project.

Research Plan
The Advanced Machines, Monitoring, and Control (AMMC) project emphasizes on the development of novel control and monitoring capabilities, and their subsequent validation. The core of the AMMC project revolves around the development and validation of cutting-edge process control techniques, utilizing advanced monitoring tools such as high-speed imaging and diffraction for both single and multi-laser AM systems. Key research activities will revolve around the following areas, with practical demonstrations conducted on the AMMT 2.0 and LPDT platforms.

1. Real-Time Feedback AM Control Mechanisms: The project proposes the integration of two distinct real-time feedback mechanisms on the AMMT 2.0 platform. The first mechanism, in-line melt pool control, leverages in-situ processed information from melt pool images to adjust laser power in real-time, ensuring rapid response times. The second mechanism, layer-wise feedforward control, utilizes layer-wise surface images to inform laser power adjustments for subsequent layers. These advanced control mechanisms are expected to lead to more consistent process variables and improved part quality.
2.   In-Situ High-Energy X-Rays Measurement Capabilities:  By utilizing high-energy X-rays for in-situ monitoring, the project aims to gain insights into the internal changes occurring during the AM process, which can significantly influence process design. Two primary X-ray measurement capabilities will be integrated within the LPDT framework: high-contrast X-ray imaging for assessing melt pool morphology and X-ray diffraction analysis for identifying phase transitions. These techniques will enable a deeper understanding of the AM process and contribute to the development of more effective scan control strategies.
3. Temperature Field Control Scan Strategies: The project seeks to develop innovative scan strategies to create desired temperature fields during the AM process. This includes exploring unique scan patterns and paths, such as elliptical and Hilbert curves, to achieve localized temperature modulation. Additionally, the project will investigate synchronized multi-laser collaboration, focusing on the creation of distinctive heating/cooling profiles through the combined use of multiple laser sources.
4. Microstructure Control: Aiming to enhance material properties, the project plans to control the microstructure of printed components through systematic modifications of the solidification process. This effort will be supported by in-situ Synchrotron X-ray diffraction for measuring phase transitions, alongside traditional ex-situ methods for microstructure analysis. The ultimate goal is to consistently produce preferred microstructures and explore the effects of additives like graphene on microstructure and residual stress.
5. Transient State Study for The Laser Melting Process: Recognizing that most challenges in AM arise during transient states, the project will explore the interactions between laser and powder during these critical periods. This study will leverage advanced controls capable of managing a broad spectrum of transient states, aiming to understand their impact on melt pool emissions and overall part quality.
6. Defect Prediction from In-Situ Monitoring: Predicting build quality through in-situ monitoring is a crucial aspect of the project. This will involve introducing process-induced defects under controlled conditions and using machine learning models to predict defects based on monitoring data. This approach is expected to enable comprehensive oversight of the building process, enhancing overall part quality.
7. Enhancing Build Efficiency: The project also focuses on enhancing build efficiency by exploring alternative scan strategies that can expand the power-velocity (P-V) range. For instance, elliptical scan patterns may allow for building in the keyhole mode without the formation of pores, optimizing the process and increasing efficiency.
8. Surface Roughness Monitoring and Control: Understanding the factors influencing surface finish is another key area of research. The project will investigate the impact of various scan strategies on surface roughness, utilizing temperature field control scan strategies and X-ray imaging to improve component surface finish.
9. Powder Denudation Reduction: Reducing powder loss during the AM process is vital for more efficient powder usage and minimizing defects. The project will explore techniques such as pre-sintering of powder and innovative scan patterns to achieve this goal, employing in-situ monitoring methods for assessment.
10. Metrology for Advanced Machine Characterization: Accurate characterization of machine performance is essential for all research based on AM machines. The project will focus on laser power density distribution metrology, measurement of laser-galvo synchronization, beam positioning accuracy, and cross-comparisons of instruments. These efforts are expected to contribute to the development of best practices and standards for the AM industry.

The logical progression of these research efforts will lead to a comprehensive solution for AM control and monitoring, effectively addressing the challenges and achieving the project's objectives. The outcomes of this research will contribute significantly to the AM community and industry, empowering the advancement of additive manufacturing technologies.