1
Review of Processing Techniques in Large-Format Additive Manufacturing of Polymers and Composites

Ahmed Arabi Hassen1, Alex Roschli1, Daniel Moreno2, Vlastimil Kunc1, and Brian Post1

1Manufacturing Science Division (MSD), Oak Ridge National Laboratory (ORNL), Knoxville, TN, 37932, USA
2University of Cadiz, School of Engineering, Department of Mechanical Engineering and Industrial Design, Río San Pedro Campus, Cádiz, 11510-Puerto Real, Spain

1.1 Introduction

Large-format additive manufacturing (LFAM) evolved from the principles of desktop and small-format fused filament fabrication (FFF). FFF systems are rooted in the technology pioneered back in the 1980s, which was commercialized by Stratasys [1]. The key patent for FFF technology, US5121329A, expired in 2009 [2], opening the door for broader innovation and competition in extrusion-based 3D printing.

The leverage of this patent has permitted the growth of this technology as hundreds of companies have started producing their own proprietary systems. This growth has permitted its development and improvement, creating more reliable systems that have generated the conditions for this technology to flourish not only in market share but also in build size, leading to the specific approach of LFAM. The adaptation of pellet screw extruders was crucial in this development, mainly coming from the conventional polymer transformation industry, not without the work of Bellini et al. [3], who in 2005 proposed a precision screw system with continuous feeding. Also, it is important to highlight Volpato et al. [4], who used a piston-driven extrusion in 2015 for polypropylene. Wang et al. [5] in 2016 presented a system called fused pellet modeling (FPM) that was integrated immediately into additive manufacturing. It is important to highlight the research done on screw performance, geometry, and configuration of the systems [6, 7], identifying the optimal shape and configuration to achieve the technical requirements of this approach.

Oak Ridge National Laboratory (ORNL) is widely recognized for pioneering the LFAM platform. Their initial system, known as the "Blue Gantry" due to its distinct, blue-painted weldment, became operational in 2013 (Figure 1.1) [8]. Originally developed to test concrete extrusion devices, the gantry platform was later adapted for polymer-based LFAM applications. This milestone marked the beginning of a new era in extrusion AM, driving advancements in build size and scalability.

Figure 1.1 Early configuration of the ORNL's "Blue Gantry" LFAM system.

Source: Oak Ridge National Laboratory / Public Domain.

Regarding scalability, desktop FFF systems typically use filament as feedstock, often operate with heated chambers, and are controlled by microcontrollers. In contrast, LFAM systems are designed for industrial-scale production, functioning in open-air environments, and employing programmable logic controllers (PLCs) or other industrial-grade controllers for enhanced precision and durability. ORNL, in developing this technology, has consistently emphasized three core goals: increasing the size ("bigger"), increasing the speed ("faster"), and reducing the cost ("cheaper"). By "bigger," the aim is to expand the physical size and scalability of LFAM systems, enabling the production of large-scale parts and structures that were previously impractical with traditional 3D printing methods. The goal of "faster" reflects a commitment to improving deposition rates and reducing production times, resulting in LFAM being viable for high-throughput manufacturing environments. Finally, "cheaper" focuses on reducing material costs, optimizing energy usage, and enhancing system efficiency to ensure that LFAM becomes an economically viable solution for industries ranging from automotive and aerospace to construction and energy. Achieving these objectives democratizes access to LFAM technology, pushes the boundaries of AM capabilities, and establishes the technology as a transformative tool in modern industrial processes.

When ORNL began developing the LFAM system in 2013, the largest commercially available FFF system was the Stratasys Fortus 900, which offered a build volume of 914.4 mm × 609.6 mm × 914.4 mm (36 in × 24 in × 36 in) [9]. In comparison, common desktop printers like the MakerBot Replicator had much smaller build volumes, typically around 226 mm × 145 mm × 150 mm (8.9 in × 5.7 in × 5.9 in) [10]. While other systems were emerging as new companies entered the market, none at the time offered a printable dimension exceeding 1 m (3 ft) in any direction. For perspective, the total build volume of the MakerBot Replicator was 4.9 mm3 (0.000299 in3), while the Fortus 900 provided 0.51 m3 (31.104 in3). In contrast, ORNL's Blue Gantry system introduced a build volume of 2438 mm × 2438 mm × 1829 mm (8 ft. × 8 ft. × 6 ft), offering a total printable volume of 10.87 m3 (384.5 ft3). This capacity was more than three orders of magnitude larger than the MakerBot system and an entire order of magnitude greater than the Fortus 900, setting a new benchmark for scalability in AM. The initial implementation of the Blue Gantry, as illustrated in Figure 1.1, featured a fixed plywood vacuum table as the build platform, utilizing build sheets adapted from a Stratasys printer. The end effector was designed to move dynamically across all three axes (-x, -y, and -z), enabling precise deposition across the significantly expanded build area.

To achieve scalability in LFAM systems, filament-based feedstock was replaced with pellet-based materials (Figure 1.2). This shift enables significantly higher throughput, allowing for faster printing speeds necessary for large-scale manufacturing. Unlike traditional systems that use a single heating element in the nozzle, LFAM incorporates advanced extruders with multiple heating elements (Figure 1.2). The extruder system consists of a hopper, where polymer pellets or powder are fed into a heated barrel containing a rotating screw. The screw plays a crucial role in conveying, melting, and homogenizing the polymer as it moves through different screw zones. The feed zone at the start of the screw picks up the raw material from the hopper and begins transporting it forward. The compression (or transition) zone follows, where the screw channels become narrower, increasing pressure and compacting the material while aiding in melting. Finally, the metering zone ensures complete melting and uniform mixing, preparing the polymer for extrusion through the nozzle. The motor and gearbox drive the screw, while heaters and cooling systems regulate the temperature to ensure proper melting.

Figure 1.2 Advancing scalability in AM by transitioning from filament-based to pelletized feedstock materials and upgrading from a single heating element nozzle to a multielement pellet extruder.

Source: Oak Ridge National Laboratory / Public Domain.

These improvements ensure consistent melting and flow of the pelletized material, providing better control over deposition rates and improved material properties. The extruder used in ORNL's Blue Gantry system (Figure 1.2) was an adaptation of a plastic welder from Dohle, repurposed for 3D printing applications. This extruder represented a significant departure from conventional filament-based systems, both in design and functionality. One key feature of the extruder was its swappable nozzles, which threaded into the end of a single-heat-zone barrel. The standard nozzle size was 7.6 mm (0.3 in), a contrast to the much smaller nozzle sizes commonly used in filament-based 3D printers, which typically range from 0.2 mm (0.0079 in) to 0.5 mm (0.0196 in). This increase in nozzle size, more than an order of magnitude, was a critical factor in achieving the higher deposition rates and throughput necessary for large-scale printing. The larger nozzle enabled the extruder to lay down significantly more material in a single pass, reducing printing times while still maintaining sufficient precision for large-format applications. The single-heat-zone barrel, while simpler than modern multizone systems, was designed to effectively melt the plastic pellets used as feedstock, ensuring consistent material flow through the extruder.

Speed has also been a metric that has evolved since 2013, when filament-based 3D printers typically operated at print speeds between 40 and 60 mm/s (approximately 2 in/s). High-speed printers could achieve speeds up to 200 mm/s (approximately 8 in/s), though their reliability was often limited. This was primarily due to poor extruder designs and the absence of advanced control algorithms such as input shaping and linear advance, which are now commonly used to enhance extrusion system performance at high speeds. Initially, these systems operated at an average speed of 125 mm/s (5 in/s) and could reach 250 mm/s (10 in/s) when equipped with command-shaping controls. Figure 1.3 demonstrates a large part being printed at high speed.

While the print speed of these novel developments was...