A Hierarchical V-Network Framework for Part Qualification in Metal Additive Manufacturing

Date

2022

Authors

Roh, Byeong-Min
Yang, Hui
Simpson, Timothy W.
Jones, Albert T.
Witherell, Paul

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Advances in metal additive manufacturing (AM) technologies have enabled greater design freedoms than subtractive manufacturing has afforded. The design freedoms and flexibilities offered by a metal AM system, however, dramatically increase process uncertainties that may also increase part-quality variabilities. Any metal AM part must be tested, validated, and verified to meet quality, safety, and performance requirements. Common qualification methodologies rely on destructive testing, which is neither cost-effective nor efficient. Much research, which has been conducted on sensing-based, part qualification in AM systems, attempts to maximize the reduction of destructive testing by closely monitoring the fabrication process in real-time. So much “big data” is generated by this increased use of sensors and available measurement sources. However, the use of the data is still hindered by 1) scale & size and 2) uniformity. We propose a hierarchical, V-network framework of quality assurance with the corresponding translation from ex-situ to in-situ part qualifications. This framework offers an innovative, Cyber-Physical System (CPS) that accurately ties models, processes, and measurements together to interpret the sensor data. The framework also supports and guides translation from ex-situ to in-situ quality measurements, thereby providing a systematic structure and focusing on interrelationships between key observations that influence AM part quality. Ultimately the sensor data can support the detection of process anomalies, thus providing a more streamlined and more efficient qualification process than is otherwise possible.

Description

LCSH Subject Headings

Citation