You are here

Printing continuous carbon fibre in true 3D

News International-French

14 Oct 2020

Electroimpact developed a continuous fibre-reinforced thermoplastic 3D printing system featuring six degrees of freedom. This system enables the tool-less rapid fabrication of aerospacegrade composite parts while providing new design freedom not possible with existing manufacturing methods.

Ever since Markforged burst onto the scene in 2014 with the Mark One, the first commercially-available continuous fibre 3D printer, engineers around the world have dreamt of a system that could 3D print large-scale continuous fibre-reinforced parts. The ability to purely additively produce complex parts with a strength several times that of most metals, but with the density of plastics, and without expensive layup tools or autoclaves, has proved an elusive challenge. Several companies, both seasoned veterans and startups, have entered this space attempting to carve out a portion of the broader multibillion-dollar additive manufacturing (AM) market. Electroimpact has been working with aerospace partners to develop this technology for OEM applications near term.

Bay door – cutaway shows a combination of printed honeycomb structure and continuous fibre attachment points

Bay door – cutaway shows a combination of printed honeycomb structure and continuous fibre attachment points

The dream
Many in the aerospace industry will correctly argue that large-scale continuous fibre 3D printing already exists in the form of automated fibre placement (AFP) and automated tape layup (ATL). Both are AM technologies whereby composite parts are produced by depositing subsequent layers of material. However, these technologies require a layup tool in the shape of the finished part to deposit the material onto. A true continuous fibre 3D printer as envisioned does not require layup tools or vacuum bagging, nor should it require substantial ancillary equipment or secondary processing steps such as an autoclave. Rather, it is a generic manufacturing platform that attempts to remove as many constraints as possible from the system, enabling the end user to create a wide variety of parts, and to create geometries that are not producible by traditional means. The result will be the creation of new composite designs not previously imagined possible, including substantially more integrated structures that reduce the number of fasteners and adhesives needed to join assemblies.

There are three fundamental characteristics to any high-quality composite part, regardless of the specific fibre and matrix material selections. These are: 1) fibre volume fraction (ratio of fibre to matrix material), 2) porosity, and 3) straightness of the fibre. Electroimpact is not alone in the pursuit of a continuous fibre 3D printing system. Other teams have taken a couple of fundamentally different approaches to this challenge. Some have endeavoured to combine a traditional fused filament fabrication (FFF) 3D printer with a mechanism that introduces the continuous fibre in the molten thermoplastic stream, embedding it in the printed part. Others have chosen to use UV-catalyzed thermoset resins, mixing their continuous fibre reinforcement with the resin right at the tool point, and initiating the snap-cure resin with UV radiation in-process.

Electroimpact Continuous Fiber 3D Printing

Electroimpact continuous fiber 3D printing

These “coextrusion” processes attempt to combine too many steps into a single system. First, a consistent fibre volume fraction approaching aerospace grade (50%+) is difficult to achieve. The tool point of the system depositing the material must speed up and slow down as it traverses the programmed toolpaths of the part in space. Thus, the infusion process of the liquid or molten matrix into the fibre must also accelerate and decelerate in sync with the tool point.

As any seasoned composite material producer can attest, the most uniform material is produced when the entire process is at steady state, not when it is subject to large transients. When this approach is taken to high fibre volumes, it is even more difficult to achieve uniform wetting of the fibre. This often results in dry fibre patches and poor fibre distribution in the matrix. The impregnation step is best performed offline in dedicated production equipment running at steady state. This equipment is designed specifically to produce high-quality pre-impregnated material with a uniform fibre distribution and volume fraction.

Next is the issue of porosity, voids do not carry load. Gold-standard, aerospace-grade autoclaved primary structure composites achieve well under 1% porosity by curing the laminate under vacuum in an autoclave while applying several atmospheres of pressure. The forces are immense, but the result is that nearly all voids are removed from the laminate. Out-of-autoclave material systems are finding more use in aerospace, and commonly achieve under 3% porosity. They eliminate the costly autoclave, but still require vacuum to consolidate the laminate and remove voids, and an oven to cure. Systems that do not consolidate the laminate in some manner will never achieve the low levels of porosity necessary to produce quality parts.

Finally, the physics of depositing continuous fibre require that it be deposited under some amount of tension. A lesson learned early on is that you cannot push a rope, you must lay a rope. If the process requires the fibre to be pushed at any point while the matrix material is softened, it will result in bunched fibres. Fibre that is not straight does not carry a load until it is straightened out. For composite parts, bunched fibres mean the load is solely borne by the much lower-strength matrix material.

SCRAM technology
Electroimpact is developing a new technology based on old technologies. Named SCRAM, or Scalable Composite Robotic Additive Manufacturing, the system is an integration of an FFF 3D printer and a thermoplastic AFP machine. The system is comprised of an accurate robot, a rotating build platform, and a climate-controlled build chamber. The end effector carries multiple material systems to print a soluble support material (the “tool”), a continuous fibre tape, and a chopped fibre material. Each print starts with the robot depositing the support material on the build platform. It then automatically switches to printing continuous fibre and chopped fibre-reinforced material to produce the part. The continuous fibre is deposited using “in-situ consolidation”, in which the tape is laser-welded to the substrate material and compacted in process. The resulting continuous fibre-reinforced parts are on the order of out-of-autoclave levels of porosity.

Bay door print in process

Bay door print in process

The chopped fibre material system was incorporated to complement the continuous fibre. High-fibre-volume-fraction continuous fibre introduces a degree of geometric constraint that omnidirectional FFF processes do not have. Often there are features of greater complexity than purely continuous fibre tape can produce. In these situations, the designer can use chopped fibre material to produce the desired feature. Once the print is complete, the support material is dissolved away, leaving a finished part. The material systems are all thermoplastic, so no subsequent autoclave or oven cycle is needed to cure the parts.

The SCRAM process differs from traditional FFF 3D printing in that it uses true six-axis toolpaths to produce parts. Most AM systems such as FFF, SLA, and SLS, are so-called “2.5D”, in which flat 2D layers are stacked on top of each other to produce 3D shapes. In contrast, SCRAM is a true 3D process where the end effector deposits material in true six-degree-of-freedom space. This is especially important when depositing continuous fibre in order to tailor the fibre orientation to the load path and to produce quasi-isotropic layups out of plane from the build platform.

The challenges
There are four major challenges to producing this complex technology: 1) the material system, 2) the printing hardware, 3) the controls system, and 4) the part programming. Of the many teams attempting to develop this technology, some have made impressive progress in two or three of these areas, but none seem to have conquered all four.

The material system itself is the most fundamental challenge to overcome. There are plenty of polymers to choose from, but few remain once extreme requirements are added, such as elevated temperature use, chemical resistance, and smoke and toxicity requirements. Plus, they can be quite challenging to process. Next, since the fibre carries most of the load, one wants to use as high of a fibre volume as practical while remaining able to reliably deposit the material and achieve a good bond. For all these reasons, and after much experimentation, Electroimpact settled on PAEK-based thermoplastics and a fibre volume between 50-60%.

Transition duct – single-wall transitioning to a double-wall duct with internal support structure

Transition duct – single-wall transitioning to a double-wall duct with internal support structure

The printing hardware turns out to be surprisingly complex and nuanced. The FFF aspect of the technology is familiar to engineers and hobbyists alike for its simplicity. Even AFP for flat charge layups has been refined and simplified to the point that it is now accessible in modest lab environments. However, once one starts printing in six degrees of freedom and incorporating continuous fibre, accuracy becomes extremely critical. The complexity of the parts that can be produced depends directly on the dexterity and accuracy of the system. Every component on the end effector adds bulk and presents a possible limitation to the geometry that can be produced. To control such a complex mechanical system, you need nothing less than an industrial CNC. The complex kinematics, frame calculations, tool definitions, and accuracy demand it.

Finally, the CAM software is an undertaking as complex as the system itself. Users of commercially-available 3D printers are accustomed to slicing software that imports a solid model of the part and automatically generates the toolpaths for the printer. When working in 2.5D, it is relatively straightforward to algorithmically generate tool paths, since there are only two degrees of freedom for a given layer. When you remove this simplifying constraint and allow for material deposition in all six degrees of freedom, the problem of path generation becomes an order of magnitude more difficult.

Printing continuous carbon fibre in true 3D

Named SCRAM (Scalable Composite Robotic Additive Manufacturing), the system is an integration of an
FFF 3D printer and a thermoplastic AFP machine

Rather than try to conquer all four aspects on their own, Electroimpact has chosen to focus on their core strengths: the hardware and control systems. The company partnered with peers from industry best suited for creating the material systems and CAM software. The resulting integrated system represents significant progress in overcoming all four challenges identified above, and as a result, demonstrates capabilities not yet seen elsewhere.

The future
While more work remains to be done in the maturation of SCRAM technology, its development continues to proceed at a rapid pace, and the level of enthusiasm and interest surrounding it will continue to propel it forward. Today, SCRAM already enables the production of part geometries that were previously impossible to fabricate. Further development will focus on hardening it for industrial production and improving performance metrics across the board. The wait is over, true six-degree-of-freedom continuous fibre-reinforced 3D printing has finally arrived.

This article has been edited with the participation of Cody Brown, Project Manager, Electroimpact Inc..

Want to discover more about Composite?
Join the worldwide composites community, and register now to the JEC Composites Magazine!

JEC Composites Magazine N°137, featuring Automotive