Jonathan McDonough explains how 3D printing will benefit the chemical engineering discipline over the next 5–10 years
I WORK in the Process Intensification Group (PIG) at Newcastle University, UK. My research intersects the areas of reaction engineering, flow chemistry, fluid mechanics, process intensification, fluidisation, and 3D printing. 3D printing is heavily embedded in most of my research activities, and I see this technology as one of the key enablers for realising next-generation processes that are less energy-intensive, highly-sustainable, and high-performance. I will share my thoughts on the areas where 3D printing will benefit chemical engineering within the next few years.
Additive manufacturing, also called 3D printing, refers to the fabrication of three-dimensional objects where material is successively added over time rather than subtracted or formed, hence its name. Under the joint ISO/ASTM standard (52900-15), there are actually seven general categories of 3D printing: vat polymerisation (VP); material extrusion (ME); material jetting (MJ); powder bed fusion (PBF); binder jetting (BJ); directed energy deposition (DED); and sheet lamination (SL). Through this diversity of techniques (Figure 1) comes diversity of materials: metals, ceramics, plastics, and many more are now readily printable, and we are seeing increased uptake across numerous applications within the academic and industrial sectors.
We classify 3D printing as a digital technology, because it enables the seamless transition between a digital CAD model and its physical counterpart. Whilst not new, there has been an explosion in popularity within the last few years due to improvements in the technology, improvements in reliability, increased access to a wider range of materials, and the main factor – lower costs.
3D printing, like many technologies, has its niche. Formative manufacturing involves the use of stress to deform the material into the desired shape and is the rapid high-volume approach. This includes techniques like injection moulding and products like the humble Lego brick. Subtractive manufacturing, which starts with a solid block of material that is machined into the final form by removing material, is the high-precision approach. This would be the go-to for high-performance applications because of the very high resolutions and tolerances; those usually required in the aerospace, transport or motorsport industries. Where formative enables high volumes and subtractive enables high precision, additive manufacturing can be regarded as the “high-freedom” approach. By removing many of the geometric constraints of the formative and subtractive techniques, we can now readily fabricate previously impossible structures, and customise with almost no time penalty. Figure 2 illustrates these three manufacturing methods.
By removing many of the formative and subtractive manufacturing constraints, particularly for interior structures, we now have access to a large complex design space. I hypothesise that an optimal structure exists within this space for most applications. To find it, we can either employ CFD-optimisation algorithms, use simple trial-and-error searches (with advantages of rapid build speeds of 3D printing), or look to nature for inspiration. For example, the Centre for Nature-Inspired Engineering at UCL looked at hierarchical transport networks in nature – such as the structures in trees and lungs – to design more intricate 3D printing-ready distributor designs (see Figure 3)1.
Looking forward, I predict a paradigm shift in design philosophies because of this enhanced design freedom. Currently, the geometry manipulation tools available in most computer-aided design (CAD) software packages only mimic subtractive manufacturing operations. As CAD software embraces the capabilities of 3D printing, we are likely to see increased prevalence of out-of-the-box ideas and a breakaway from the conventional.
3D printing can fabricate multiple components as a single piece. Eliminating joints reduces the amount of material required and leads to lighter-weight components that have higher space-efficiencies. The aerospace and motorsport sectors in particular benefit from this concept of “lightweighting”. For example, two of the collaborators of PIG, Aavid Thermacore and HiETA, can recover the development costs of their advanced heat exchanger concepts and compete with the higher resolutions of subtractive manufacturing.
Typically, 3D printing produces less waste than subtractive manufacturing, where discarded material is difficult to recover and recycle. 3D printing starts with nothing and just adds the material we need. This might reduce the cost of using more expensive high-performance materials for new applications. However, the “less waste” argument rarely considers the full lifecycle of the process, where waste can take the form of support structures, ancillary processing, and the energy consumption of the process itself.
D printing enables a much faster transition between an idea and a testable prototype. In many of our projects at PIG we have reduced the lead time from weeks to hours. For example, using a simple trial-and-error approach, one of my undergraduate students rapidly designed, fabricated, and tested a range of fluidic devices. Within the time it would take to “conventionally” fabricate a single device, we had already experimentally tested the behaviours of 20+ devices, enabling us to propose an ideal design within the variables considered2. More recently I have collaborated with industry to speed up research. For example, with Torftech to miniaturise its toroidal fluidised bed concept (Figure 4), which would have been almost impossible via conventional manufacturing3. With this knowledge we now have a rapid-response small-scale platform for sorbent testing that minimises the need for expensive pilot-scale testing.
Imperfections often exist between layers, that produce roughness at scales of tens to hundreds of microns that may require post-processing treatments. As chemical engineers, we can exploit this roughness for a range of applications. Roughness disturbs the boundary layer, and whilst this may increase pressure drop, we also observe the promotion of mixing for reactor engineering and heat transfer applications.
The status quo for many commercial chemical reactor systems is to optimise the chemistry to fit the specific reactor geometry available. Of course, this inherently makes the chemistry sub-optimal because the reaction is being run in a way that makes it fit into the equipment. This limits the capability of achieving truly optimised processes and imposes time-consuming constraints that, for example, don’t meet the needs of pharma’s shift towards small volume manufacturing for tailored drug design. I predict a transformative approach: use 3D printing to tailor the reactor to the specific chemistry.
Many have speculated why pharma has been slow to adopt potentially more efficient flow chemistry technologies over “antiquated” batch processing. One major reason is the vast existing infrastructure, both on the technology and regulatory sides. However, attitudes are changing – 90% of companies view continuous as important4, and the director of the FDA has called for the sector to embrace new innovations like flow chemistry5. With 3D printing we could manufacture single-use flow chemistry reactors that would promote a more gradual transition to flow. These single-use reactors could be easily customisable, compliant with traceability and good manufacturing practice, and eliminate the need for cleaning between runs.
Whilst there are several specific engineering challenges (eg limited resolutions, high investment costs for metal), I would like to focus on four more general caveats that all pose potential roadblocks for our discipline.
Intellectual property (IP) underpins business. However, this creates a potential challenge for collaboration. Remember that 3D printing is a digital technology; once we have the CAD model we can easily fabricate then replicate it. Numerous software packages enable the manipulation and copying of these files without much need for CAD literacy. With insufficient oversight or regulatory frameworks, companies risk losing control over their IP once they share a CAD model with their collaborators. Most available protections (like copyrights, trade secrets, or trademarks) are ill-equipped to tackle the problem. It’s also possible that we may see a shift away from IP altogether, where branding and reputation will become the de facto success factors for a given speciality.
Earlier I suggested that customisation comes without a time penalty. This is because 3D printing takes approximately the same time for equivalently-sized parts regardless of the geometric differences. However, it’s likely that this might shift the cost of customisation onto certification. For high-end applications in particular, parts must undergo rigorous testing to ensure they comply with gas tightness, tolerances, pressure rating, safety ratings, and standards. Bespoke parts necessitate time-consuming bespoke testing.
At present, in terms of speed and cost, 3D printing is probably best geared for prototyping and research, and potentially for high-end performance applications provided that clear performance benefits exist. The scaleup challenge, both on the hardware and software sides, explains this limitation. In hardware terms, it is not simply a case of making larger machines. The different rates at which volume and surface area change upon scaleup makes it difficult to control the fabrication process as printers get bigger. Put simply, larger printers increase the likelihood for defects because it is difficult to control the stresses and temperatures inside larger objects. On the software side, as CAD models become larger, so too do their file sizes. Even small-scale geometries produce handling problems; typical desktop computers often struggle to render the millions of surface coordinates that quickly proliferate when using copy-pasting operations. This is exemplified in Figure 5. For reference, this file crashed twice whilst I was trying to take a screenshot to include in this article!
It is clear to me that we need to see the growth of more multi-disciplinary teams, particularly between materials scientists, engineering practitioners, chemists, modellers, and 3D-printing specialists in order to actually deliver the next-generation of process engineering concepts
Finally, materials are a key enabler for the uptake of 3D printing. However, there are very few examples of materials standardisation in the 3D printing field. Most materials are proprietary, which reduces the designer’s control. Here we see a potential restriction to the design freedom, especially if the rate-limiting step resides within the performance of the material itself (such as the thermal conductivity for heat transfer applications). 3D printing might also enable us to manufacture complex composite materials, including new possibilities like graded transitions between two different materials, dispersing secondary materials in a primary material matrix, or functionalising polymer backbones for chemistry. However, CAD software lacks the ability to design and investigate these possibilities, potentially restricting access to true-optimal geometries.
The workflows of 3D printing extend far beyond fabrication into design, post-processing, certification, and procurement, for example, which all likely contain hidden bottlenecks. Additionally, there is often a significant knowledge gap between 3D printing experts, usually the early adopters, and beginner users – something I see when training new students. Unfortunately this knowledge gap exists across all organisations that use 3D printing. Here, the deep understanding of the nuances and informal design rules are typically concentrated within single individuals or small teams, leading to unconnected pockets of knowledge. This can create problems in terms of cascade training, especially if these experts move on without sharing their knowledge. In some cases, the experts are probably not even aware of what the beginners don’t know. Thus, undoubtedly one of the key skills for the next generation of engineers will be 3D-printing literacy. It is also clear to me that we need to see the growth of more multi-disciplinary teams, particularly between materials scientists, engineering practitioners, chemists, modellers, and 3D-printing specialists in order to actually deliver the next-generation of process engineering concepts.
I hope this article has inspired some debate over the possibilities and potential pitfalls. For a more in-depth overview of the ideas presented here, you can either read my more detailed review paper6 or watch my recorded webinar hosted by the UK National Heat Transfer Committee7. You can also get in touch with me directly.
1. Bethapudi et al, Energy Conversion and Management 220, 2020, 113083.
2. McDonough et al, Chemical Engineering Research and Design 117, 2017, 228–239.
3. McDonough et al, Chemical Engineering Research and Design 160, 2020, 129–140.
4. Pollington, Johnson Matthey Technol Rev 63(3), 2019, 157–165.
5. https://bit.ly/35GJG2K, 2015.
6. McDonough, Thermal Science and Engineering Progress 19, 2020, 100594.
7. www.uknhtc.org/webinars, 2021.
Catch up on the latest news, views and jobs from The Chemical Engineer. Below are the four latest issues. View a wider selection of the archive from within the Magazine section of this site.