May 2023 Volume 5

AUTOMATION

Digital Process Twins - A New Paradigm for the Forging Industry By Julius Schoop, Ph.D. & Rollie Dutton, Ph.D.

Digitally-enabled manufacturing processes increase global competitiveness, offer opportunities for continuous improvement, and are potentially amajor enabler for the surge production of critical components by the Defense Industrial Base. In this article, we will present a case for howDigital Process Twins can help manufacturers adopt robust digital technologies that go beyond automation. Digital Process Twins: A Key Enabler for Agile Process Development Realizing the promise of process modeling to reduce the time and cost required to get to a viable first article has been a topic of intensive research and development for over a century. A well known example, CNC machining, has revolutionized productivity in the manufacturing industry since it was first introduced in 1952. However, despite decades of progress during the 20th century, most will agree that 21st current industry practice for critical processes like casting, machining, and forging remains stubbornly empirical, and overly reliant on human experts. Experts who are becoming increasingly difficult to find.

Integrated Materials and Processing Intelligence (IMPI) is an emerging framework to enable model-based manufacturing process development and optimization. At its core, the IMPI approach consists of intelligent integration of process-specific materials science with a production engineering focus on cost, quality, and throughput metrics. It is data-driven, relying on process-specific material behavior to rapidly define Digital Process Twins (DPTs). These DPTs are carefully calibrated and validated fit-for-purpose models that prioritize computational efficiency (speed) by focusing specifically on cost and quality-relevant metrics. In this way, DPTs quickly and reliably deliver process-relevant information to engineers, technicians, and craftsmen to enable agile process modeling during production, even in high mix/low volume scenarios. The four key aspects of the IMPI approach, centered on properly calibrated and validated DPTs, may be applied to any thermo-mechanical manufacturing process such as forging, machining, and additive manufacturing. These are illustrated in Figure 1.

Figure 1: The IMPI paradigm is a new approach for generating fast-acting Digital Process Twins.

The IMPI method fuses historical process data, current process specific in situ data (from both manufacturing tools and materials), and fast-acting DPTs to generate “smart” and “explainable” data that is more compact and meaningful than so-called big data. Traditional ‘black box’ data analytics and machine learning (ML) approaches, working from a combination of historical data and some in-situ data from generic sensors, typically lack the ability to

achieve robust performance, especially outside of the context within which the data was originally obtained (i.e., a lack of transferability and generalizability). Our approach instead lets ML algorithms draw on preexisting data describing manufacturing best practices, augmented by data specific to the process-of-interest: an intelligent combination of physics-based (fast, semi-analytical) DPT model outputs, and in-situ data from process-specific sensors.

FIA MAGAZINE | MAY 2023 36