Functional Requirements of Data Analytic Tools and Software for Metal Additive Manufacturing

Various software tools for the measurement, modeling, simulation, and management of additive manufacturing (AM) processes and parts are increasingly becoming available to users and researchers. The capabilities and accessibility of these tools vary greatly, particularly for AM data management. Tools with the capability of managing AM data range from point solutions to legacy product lifecycle management tools. Advances in AM production capabilities have been driven by increase in in-situ and ex-situ measurements of processes and parts, led by a rapid increase in the volume, variety, velocity, and value of data. As a result, there has been an increase in need for AM data analytics capabilities. These analytic capabilities need to be supported by tools with ability to process a variety of measured data needed to understand defects formation, geometric variation, surface roughness, and progress in the layer-by-layer laser scanning process.

There are three main challenges relate to data management tools can be found in metal AM (MAM) data-related workshops. First, there is a lack of integrated suite of tools for data management, analysis, monitoring, and control of MAM processes, including modeling and simulation of heating, melting, and solidification, microstructure analysis, and material properties prediction in MAM. Second, there is a lack of software tools for process planners to determine process planning allowable for MAM processes. Third, there is a lack of appropriate software tools that can enable users to correlate data from different sensors for measuring materials, processes, and parts. As a result, there are few tools available to meet these challenges when industrial users need to 1) specify design rules and allowable to ensure manufacturability, 2) monitor and control the processes, and 3) analyze production lines to optimize future part production scenarios.

To better understand what capabilities are currently available for AM data analytics, and what capabilities could be further developed, this paper provides a landscape of software tools for AM product lifecycle, specifically, design, design analysis, process planning, process monitoring, process modeling, simulation, and production management. We start by identifying software functional requirements for MAM product data-driven analytic tools for ensuring the quality of 3D printed parts. We then identify software tools that have the capability to satisfy these requirements to some extent. The identified requirements for MAM software tools are in the following areas: design modeling, design analysis, material selection, process planning, in-situ process monitoring planning, process analysis, process modeling and simulation, microstructural analysis, part property analysis, data registration, material management, product lifecycle management, production management, and part validation. Next, we select a sample set of tools and characterize them how well they are capable of meeting the identified software functional requirements in a tabular form. Based on this characterization, a software tool can be categorized under one of the three capabilities: highly capable, moderately capable, or somewhat capable based on the level of meeting the identified requirements. Some requirements cannot be matched with available tools.

Finally, we identify opportunities to develop new tools for data analytics to improve the product quality and reduce production time. These opportunities are as follows: 1) design allowable tool for exploring various options in design, 2) in-situ and ex-situ monitoring methods, 3) MAM process and property analytics, 4) MAM data fusion tools for aligning data from different sensors or sources, and 5) a suite of integrated tools to valid materials, processes, and parts. For future work, complex input-output data objects, such as powder material property properties, pore growth model, scanning strategy, tessellated model, grain growth model, and XCT model, must be defined and modeled for capturing physical phenomena. Complex physics-based models and software tools can then be developed for new data analytic capabilities for MAM technology users.