Additive manufacturing (AM) is joining materials to make objects from 3D model data, usually layer upon layer, instead of subtractive manufacturing. It gives engineers far more freedom to produce components with complex shapes, more than what is possible using subtractive methods. Its attributes give it the potential to impact more significantly on technology, manufacturing, and retail for the industry than almost any other technology to date.
Motorsport is a constant improvement exercise within rapid timeframes, and this advanced disruptive manufacturing technology is crucial in implementing new improvements. ‘With AM, freedoms are simply realized in hitherto un-manufacturable geometries,’ highlights Michael Fuller, CEO and Founder of Conflux, a company specializing in AM heat exchangers. ‘The sweet spot for AM lies where designers want to create a component with only the required structures for its application, optimizing its size and weight. For example, designing hollow tubes and I-beams inside constructions can help create a stiff internal structure with minimal weight.’
The application’s load case determines where material needs to be present. Often the centre of a structure doesn’t require any material, as the outside of the part carries the stress. ‘The part’s application dictates the technology, material, and any post-processing requirements to ensure it is fit for purpose,’ says Samuel Persaud, Head of Projects at Graphite AM. Fuller adds, ‘Design for AM allows far greater freedom than traditional manufacturing. Limitations of imagination and creativity are common and, as with any industrialized manufacturing technology, training in design for AM is a critical factor for the successful rollout of the technology.’
Subtractive manufacturing and surface finishing can produce tolerances of as little as 0.2 microns and a Ra 0.2, respectively. This level of refinement is only possible with AM with severe post-processing of the part. Even then, it’s unlikely to be that precise. ‘AM surface finish typically falls within +/- .125mm in the x, y or z direction,’ notes Michael Littrell, CEO of CIDEAS. ‘It’s not uncommon to build a part, measure it and scale areas of the part file to dial in tighter tolerances against the AM part.’
Kevin Lambourne, Managing Director of Graphite AM, says, ‘The tolerance and accuracy is technology and material specific. Ultra-high-resolution AM machines can build to tolerances of 25 microns, but these machines are limited in materials and are only suitable for small components. So, there are still plenty of components that must be manufactured using more traditional methods.’ Fuller added, ‘Surface roughness aside, the microstructure across bulk geometries (>0.5mm) can be consistent, and this is achievable and measurable. In the case of thin walls and microfluidic channels, surface roughness can be the same as the geometric features themselves.
Nathan Rawlings, UK Manager at EOS, surmises that there are now so many types of AM that there is likely a solution for most applications. ‘The market can appear confusing to someone with no previous experience, and it depends on what you try to achieve,’ says Rawlings. ‘Some technologies are ideal for wind tunnel testing components, but you wouldn’t make a functional part with that technology.
There are four key technologies which are the most used and most established – SLA (stereolithography), which is curing a moulding resin in a vat with UV light; SLS (selective laser sintering), which is sintering polymer powders into 3D shapes, DMLS (direct metal laser sintering) which is melting metal powders in into a 3D shape using high power lasers and FDM (fusion deposition modelling) which is a printing method but in a structural way.
By units sold, FDM is the most popular technology worldwide due to its low cost and ease of use. Due to speed, build volume and repeatability, SLS is becoming increasingly popular because of their potential for low to medium-volume production applications. DLMS and FDM technologies are often used for the most demanding motorsport applications; however, Tim Chapman, Head of AM at McLaren Racing, emphasizes, ‘Stereolithography technology and the materials have evolved so much recently that it is changing the way we use it. We do not just manufacture prototypes with this technology anymore; we now produce many full-scale components and full-size tooling. Wind tunnel testing is still the gold standard when assessing how every surface works together, either as an assembly or as a complete car. Our [Stratasys] Neo series of AM machines have helped us dramatically reduce the lead times of our aerodynamic wind tunnel and other components.’
Software
AM software is constantly improving. A significant challenge facing component designers who want to manufacture parts using AM is defining the properties of the layered construction. The problem is that CAD and FEA software cannot define layered material properties because CAD programs work out each structure as an idealization. The structure as it appears out of the AM machine is not ideal – it has very rough surfaces and other imperfections, and the geometry at a microscopic level often diverges significantly from the idealization. Fuller says there have been huge strides across all disciplines of AM when it comes to software for defining the properties of AM construction, but it still needs some discretion.
‘Design automation and topology optimization tools are now very well suited to AM,’ he remarks. ‘Multi-physics modelling and simulation are utilized to predict manufacturing phenomena (akin to mould flow or weld analysis) and, critically, the advent of in-process monitoring supports a very high level of real-time QA though there is some way to go.’
Most parts produced using AM are designed in 3D CAD and run through a program that turns them into 2D sections that are then fed into the AM machine software as instructions for the build. ‘Conventional CAD/FEM software is developing increasingly in the direction of AM design with new features,’ comments Rawlings. ‘A typical example is thread modelling, and for a few years now, a thread can be modelled and manufactured using the thread feature with just a few clicks. ‘Before that, the thread was only hinted at for a drawing, but there was no physical thread.’
For powder-based technologies, the AM software automatically chooses the laser power and the speed at which it passes across the powder bed. ‘There are standard parameters for which these values are already specified,’ continues Rawlings. ‘Often, however, these parameters do not get the maximum out of the parts. With appropriate knowledge and experience, you can adapt or optimize the parameters for a specific part to achieve the best possible part quality and reduce the printing time.’
AM operation
The general principle of Powder Bed Fusion is to print parts by solidifying the cross-section of the CAD model into a fine layer of powder (usually a derivative of Nylon). The un-solidified powder acts as a support structure enabling complex geometries to be built in one piece quickly. It’s very efficient and quick and can quickly yield a high volume of plastic parts. Most importantly, it allows for stacking parts above one another in the Z plane.
Fused Filament or FDM technologies feed a plastic filament from a spool into a computer-controlled head which deposits hair-thin strands of melted plastic layer by layer. It is a very linear process, and build speeds are somewhat slow, but substantial, durable parts in several exotic plastics can be achieved. In this technology, parts can be built solid or large cross sections can be built with a supported interior to reduce material usage and improve throughput. This process is ideal for jigs, fixtures, and production aids.
Digital Light Processing (DLP) technologies have gained a lot of interest due to their small footprint, high speeds and detailed results. Although production sizes are limited, complex resins that are very close to traditional production-grade plastics have been developed recently. Due to the material capabilities and fine feature resolution, DLP parts are finding their way into customized components such low production vehicle components, connectors, fasteners and more.
On the metal AM side, SLS and MJF offer the most design flexibility though restrictions with minimum wall thickness and getting access to remove unwanted material from inside components must be considered. Metal material science has been a key development path for AM, more so than incremental hardware changes. It can seem that material innovation in the market is fairly slow. Still, looking at the bigger picture, AM materials have developed to a price and performance level where they are more widely accepted for production parts. The demand for adopting AM in recent years has increased the number of available performance materials, such as traceability and unique features such as flame, smoke, toxicity ratings, ESD, chemical resistance, UV resistance, durability and other exotic characteristics starting to show up. The most notable area of growth in recent years is within photopolymers, inching them closer to traditional thermoplastics.
Hardware
AM hardware development has been a process of continuous improvement where each new system gets quicker, produces better quality parts, and features more integrated quality controls. The number of new machine manufacturers and start-ups joining makes it look like the evolution is happening over the months. Still, at the production end of the market, development is much slower, with the occasional jump. F1 teams use additive equipment 24/7, and full-scale production for AM parts has emerged only in the past few years.
“We find the high-definition components from our [Stratasys] Neo machines require minimal hand finishing now, which allows much faster throughput to the wind tunnel,’ says Chapman. ‘In addition to speed, we can now produce wind tunnel parts with supreme accuracy, detail and surface finish, enabling our team to enhance testing and find innovative new ideas to improve performance. I cannot overstate how important these benefits are in F1, with super tight deadlines to deliver cars to the next race and where the smallest design change can make all the difference between winning, losing or making up positions in the race.’
Major machine manufacturers have demonstrated automated factory solutions that promise to dramatically improve productivity across the complete AM process chain, from CAD/CAM file preparation to part build, heat treatment, other postproduction, and build plate removal. These systems remove most of the manual processes inherent in the AM part production and automate them into one system. These AM production systems require the user to load the build plate, and everything runs automatically until the part is complete. Some are modular, so they can be built up to each user’s requirements to complete a part from CAD/CAM file upload to a serviceable component.
Even multiple build chambers can be used to manufacture numerous parts simultaneously. Because the powder handling is integrated into each module, the system allows the operator to use different materials in each build area. The postproduction and heat treatment modules are considered critical to part performance and must be carefully controlled to maintain the part quality. The AM process usually causes tension, and if it is allowed to cool to room temperature before it is put through a heat treatment process, it can warp and be ruined.
Finish
The resolution of each layer at the layer’s boundary and its orientation determines the finish characteristics. If the surface takes the bulk of the load, it must be tailored differently from the interior section to create a load-bearing structure. Generally, FDM is the AM technology where hollow parts are common practice; nearly all the other technologies will build solid parts if they are modelled this way.
Increasing the laser’s power at the surface increases the resolution of the layers in that area. If the users understand the particular material and have the freedom in the machine to change the laser’s power and speed, they can get down to feature sizes of between 100 and 200 microns. Heat exchangers and other components requiring a high surface area to volume ratio can be made much more efficient if the machine is tailored to the specific project. The exacting nature of heat exchanger geometries necessitates bespoke processes, including but not limited to laser power and speed.
Surface finish is a function of geometry, orientation, process parameters, laser spot size, powder particle size distribution, gas flow, build plate geometry, material density, layer height, material quality and laser calibration. The trade-off between speed and resolution is straightforward – the higher the speed of the machine, the lower the resolution. SLA is resin based, offering a better surface finish over SLS or MJF, which requires bead blasting to remove the surface powder. Vapour smoothing can improve SLS and MJF by removing layer lines and leaving a smooth, glossy surface finish.
After the AM part is complete, post-processing such as de-powdering, finishing the component and cutting off the build plate must be considered. Post-processing has evolved a long way in recent years. Some AM firms utilize a Synchrotron light source to acquire a significant amount of high-resolution CT scan data on parts. These large CT data sets (>15TB) are analyzed within a software suite to measure 3D pore density distribution and morphology, surface roughness, wall thickness, and feature distances and identify critical defects.
The use of AM in motorsport shows no signs of abating, and a much higher percentage of components will likely be manufactured using this technology. We will undoubtedly see the emergence of larger, faster, more precise machines and many more materials.
