Northrop Grumman Corp. (NGC), a pioneer in additive manufacturing (AM), recognized early on that integrating continuous carbon fiber during the process is critical to producing load-bearing composite structures. For the last seven years, NGC has been developing advanced continuous-fiber-AM (CFAM) capabilities with increasing levels of performance and complexity. One example of this is the scalable composite robotic AM (SCRAM), an agile, flexible system for fabricating aerospace-grade, complex composite structures developed in partnership with Electroimpact Inc.
Traditional 3D-printing processes should more accurately be described as “2.5D” printing. Here, the material is deposited successively in flat slices that, when stacked together, create a 3D object. SCRAM, on the other hand, is a true 3D system that utilizes robotics and a seven-axis deposition system to precisely place material in the X, Y and Z axes. When placed along load paths, the continuous carbon fiber material creates a structurally optimized ply architecture, making SCRAM an ideal solution for aerospace-grade composite structures.
SCRAM (not to be confused with the Scramjet, an airbreathing engine designed for Mach 6 and higher hypersonic flight) is a cost-effective production cell that supports build-on-demand requirements and can scale as needed. This in-situ additive process does not require capital equipment such as an autoclave or other post-consolidation methods. SCRAM uses commercially available thermoplastic tape in various widths to provide scaling and design flexibility.
Another feature that no other CFAM system offers is an integrated, 3D dual-continuous-fiber-plus, fused-filament-fabrication (FFF) deposition capability (patented by NGC) that allows SCRAM to optionally print a soluble tool in-situ, print the part on top of the tool and dissolve the tool away once the part is complete. SCRAM also integrates milling to subtract material from printed parts, eliminating the need for secondary operations such as drilling fastener holes, cutouts and final part routing. These uniquely combined attributes enable a reliable and agile manufacturing process with the added ability to design a topology-optimized part.
The SCRAM fabrication cell eliminates the costs and lead times associated with hard tooling, removes the need for autoclaves, provides revolutionary design freedoms, enables multi-part integration, facilitates rapid response time to design changes and provides a digital-twin part representation. All of these support the A&D community’s goals of rapid prototyping for quicker time to market, as well as shortened air vehicle qualification and certification timelines.
The ability to design, analyze and optimize air vehicles for CFAM remains one of NGC’s core competencies. NGC has been awarded seven patents related to its SCRAM technology, with accompanying patents pending for additional capabilities that, when issued, will widen the envelope of aerospace and defense applications. The issued patents include the use of a multi-material and multi-process end effector to additively manufacture composite structures.
NGC also holds patents in composite structures made of a hybrid layup architecture using CFAM and AM. Electroimpact licenses these patents, leveraging them in the SCRAM system it builds and sells to the open market. Electroimpact was selected as NGC’s hardware platform provider due to its robotics and automated fiber placement (AFP) capabilities. The interest in making SCRAM cells available for purchase is driven by the foreseen need for a future robust supply chain for CFAM-fabricated hardware across numerous industry segments.
The robot-mounted end effector combines two manufacturing technologies, FFF and AFP. The system currently utilizes high-performance, continuous-carbon-fiber (CCF) material for the AFP module and discontinuous-carbon fiber (DCF) for the FFF module. The end effector is also equipped with a second FFF module that prints tooling in-situ with a soluble “neat” polymer that can be washed away when the part comes off the build plate.
When manufacturing structures, the system uses its multiple deposition heads to perform both AFP deposition and material extrusion of neat or particulate-filled polymers in a true 3D-deposition process. The latter serves multiple purposes, including fabricating temporary support structures and manufacturing in-situ, honeycomb-like structures between continuous-fiber layers to manufacture unitized core-stiffened structures in one production process.
Combining these technologies enables the production of complex composite assemblies into a single build/print process. Figure 1 shows a highly complex panel manufactured in a single setup using SCRAM. The octagonal pads are solid laminate insets with three densities of core stiffener to transition between the hard points and panel edges, while structural optimization was used to develop the variable density core. Making such a part via traditional composite manufacturing techniques would require multiple core materials and machining processes, multiple tools, cure cycles, trimming and bonding operations. SCRAM eliminates all this.
Once a part has been designed, structurally analyzed and sized to required loads, the next critical step before printing is to write the numerically controlled (NC) program that defines the robot motion, often referred to as “toolpaths” or G-code. At this stage in the value stream, the NC programmer selects a manufacturable coverage strategy that accounts for variables such as tow lap, tow gap, steering/wrinkling, angular deviation, deposition sequence, compaction, course/tow spacing, roller crush and much more.
The toolpathing of CFAM versus typical 2.5D-type printing is orders of magnitude more complex because of the kinematics behind a seven-axis system versus a three-axis machine. Offline NC programming is an area often overlooked and thus has proven to be the Achilles heel for those in the CFAM arena. NGC has long recognized the need to develop software tools for slicing the designed part into a series of true 3D toolpaths that, when combined, create a structurally optimized composite coverage strategy.
This software development has been a major thrust, arguably just as important as developing the hardware platform itself, and was a key enabler to the success of the SCRAM system. Furthermore, because the coverage of the entire composite structure is now fully digitized, this creates the foundation needed to establish a digital twin of the “as-built” condition. Figure 2 shows these digital tools in use when manufacturing a 3-ft (0.91-m) complex duct in a digital simulation environment prior to being physically printed.
Because SCRAM is a highly digital environment, as well as a physical capability, NGC is working to build the link between the process simulation and physics-based integrated computational materials engineering models. The linking of these simulations/models will allow for a robust predictive performance toolset to capture the “as-built” condition and correlate resulting structural behavior. Having this digital toolset should result in rapid trade studies during the design/manufacturing iterations and can accelerate the certification process with reduced empirical testing needed for the insertion of new materials and/or process parameter variation.
In March, NGC integrated its first SCRAM system into a production environment. This marks a significant milestone because it is the first non-development system ever produced by Electroimpact, indicating a level of maturity that allows NGC to begin transitioning away from development and into low-rate initial production.
Three additional cells are being prepared for NGC facilities. The first task is to execute on a recent contract to produce a topology-optimized structure for a customer’s attritable (a new class of low cost, reusable unmanned aircraft with limited life) vehicle platform. The first flight of a SCRAM-fabricated component is expected in early 2024.
Connect With Us