Alloyed, a University of Oxford spin-out founded in 2017 as OxMet Technologies, is gaining traction with its vertically integrated approach to metallic component design, driven by computational alloy innovation and digital manufacturing tools.
Speaking to 3D Printing Industry, CEO Michael Holmes explained the shift in the company’s emphasis.
Originally focused on computational alloy design, Alloyed found its real market potential not just in creating new alloys, but in optimizing the entire production stack. “We could see that the same underlying metallurgical expertise that we use to design alloys computationally would be applicable downstream in optimizing alloy components,” Holmes said. “Although metallurgy is relevant to everything we do, less than 30% of the components we work on are made from bespoke alloys.” While some clients approach Alloyed in search of new alloys, Holmes noted that often the optimal solution lies in maximizing existing materials first. “Using a bespoke material is often the last thing rather than the first thing a company should do,” he said.
The company’s platform combines four core tools: Alloys by Design (alloy design), Architect (component and build design), a build processor that directly programs laser activity down to the voxel level, and a data management system for process control and feedback loops. Together, these enable what Holmes calls “full stack optimization,” a systemic approach from alloy formulation to in-process data feedback.
Alloyed’s strategy has also proven appealing to investors. Following a recent £37 million Series B round, the company has raised a total of £78 million from backers including Oxford Science Enterprises and the Development Bank of Japan. Holmes was quick to downplay fundraising hype in favor of fundamentals: “The better measure of traction is revenue. We turn over £20 million or so, and that’s growing at 50 to 100% a year.”
Join AM experts on July 10th at Additive Manufacturing Advantage: Aerospace, Space & Defense. Spaces are limited for this free online event. Register now.
Alloyed’s global optimisation engine rewrites metal design assumptions
The company’s alloy design work, though now a minority share of its operations, still drives high-value technical projects. Using a platform based on predictive modelling, Alloyed navigates complex trade-offs between properties such as high-temperature strength and printability, particularly in alloys for additive manufacturing.
“Traditional cast alloys are often unprintable because the mechanisms that make them strong also make them crack during additive manufacturing,” Holmes explained. To overcome this, Alloyed deploys models that relate alloy composition not just to mechanical performance (such as strength, creep resistance, and fatigue) but to processability, including two distinct crack mechanisms in nickel alloys: liquation cracking during printing and strain-age cracking during post-process heat treatment.
Key examples are Alloyed’s ABD-900AM and ABD-1000AM nickel-based superalloys. Instead of increasing gamma prime content, which improves strength but leads to printability issues, Alloyed’s optimization exploits interface boundary energy, a less conventional strengthening mechanism that reduces the likelihood of cracking.
“What we do is apply both processability models and performance models simultaneously to hundreds of thousands of possible alloy compositions,” Holmes said. The objective is not a slightly improved material, or trade off between characteristics, “It’s global optimisation, not local.” He contrasted this with traditional industry methods. “A lone metallurgist might tweak an existing alloy based on intuition. That can get you to a local optimum, but nickel alloys involve 8–10 major elements and minor additions. The design space is enormous and full of local peaks. Most people just climb one and stop.”
Unexpected findings often emerge from Alloyed’s computational process, though Holmes cautioned that surprise outcomes may signal either innovation or oversight. “You think you’ve invented something, but sometimes it’s just a sign you didn’t include the right variables,” he said. For example, fatigue performance may deteriorate despite gains in strength and creep resistance, due to unmodelled corrosion interactions.
Alloyed’s recent work with Anglo American on a novel platinum alloy yielded similarly non-obvious results. “Within the 5% of permitted non-platinum elements, we ended up with constituents quite different from what we anticipated at the outset,” Holmes said.
Alloyed pushes additive manufacturing into magnesium and grain-optimised alloys
Alloyed has expanded its materials portfolio into magnesium alloys and is advancing additive manufacturing capabilities at the grain and microstructure level, according to Holmes.
While most of Alloyed’s projects remain under NDA, the CEO confirmed the company is developing proprietary magnesium alloys for laser powder bed fusion (LPBF), positioning Alloyed as “one of the very few” organisations globally, and the only one in Europe to his knowledge, actively printing high-performance magnesium parts. “Magnesium is notoriously difficult to process,” Holmes said. “It has a very narrow processing window. There’s not much difference between the melting and boiling point, and it’s highly emissive, which disrupts the laser path.”
Processing existing alloys has also required Alloyed to implement customised toolpath algorithms, particularly to manage multi-laser interference in high-emission materials. Holmes noted that the flexibility of the Renishaw platform has been advantageous.
The firm’s work in nickel alloys has reached a new milestone with the upcoming release of ABD-1000AM. This alloy operates in what Holmes described as a “rarefied atmosphere,” requiring fine-tuned management of grain structure, interstitial content, and post-processing parameters.
“ABD-1000 has a higher gamma prime content than ABD-900, but also derives strength from interface boundary effects,” Holmes explained. “To avoid cracking and control grain growth, we’ve had to build models that account for interstitials like boron, silicon, and oxygen at very fine scales.”
This focus on grain boundary engineering is particularly critical in rotating components, where requirements for creep resistance and fatigue durability vary with grain size. “In turbine applications, you want larger grains for creep performance, which is counter to what you usually get from AM,” Holmes said. Alloyed’s modelling allows it to influence grain growth during both alloy design and post-processing.
Defect mitigation, including hot cracking and lack of fusion, is approached primarily through process-aware alloy design but also through scan strategy optimisation when necessary. “Solving problems lower in the stack, at the material level, gives more freedom up the chain,” said Holmes. “But in practice, we often solve them at the processing level.”
On the potential market size for magnesium additive manufacturing, Holmes was pragmatic. “Whether the market is one billion or ten billion is less relevant to a £20 million company than whether we can do it right,” he said. Still, he acknowledged the potential: “If we succeed with magnesium, it could unlock application areas as valuable as the entire current additive manufacturing market.”
Looking more broadly, Holmes identified three persistent constraints holding additive manufacturing back: limited materials, cost-performance imbalance, and lack of scalable certification pathways. “If you can push those last two, optimising both print economics and system-level qualification, then AM parts don’t need to cost 10x milled or stamped equivalents,” he said. “That’s when entirely new markets become viable.”
Digital qualification and supply chain resilience in aerospace materials strategy
Holmes confirmed that Alloyed is collaborating with Boeing across multiple programmes, including a high-profile initiative funded in part by the UK’s Aerospace Technology Institute (ATI) to accelerate component certification. “Qualifying a new aerospace material can cost up to $100 million,” Holmes said. “With additive, the irony is that a fully digital process should simplify qualification, but the industry still treats it as a complication.”
The ATI-backed programme aims to change this. “Within two to three years, we expect to have models that relate alloy composition and processing parameters to defect rates and fatigue performance. The goal is to reduce the number of physical tests required by using statistically validated predictive models,” said Holmes. If successful, this could allow qualification of new materials in aerospace timeframes measured in weeks for space, months for military applications, and under three years for civil aviation.
Beyond qualification, Alloyed is also contributing to supply chain resilience through targeted alloy design. One example involved creating a chassis-grade aluminium alloy that tolerates higher levels of copper and iron (impurities found in recycled US feedstock) allowing an automotive manufacturer to reduce virgin aluminium use. “This is sustainability, but also cost and resilience,” Holmes said.
Rare earths have received renewed attention amid geopolitical tensions. Alloyed has observed an unintended benefit from its alloy design work: some rare earths have been engineered out entirely. “It wasn’t the goal, but it’s a positive side effect,” Holmes said. While he acknowledged rare earth supply is currently dominated by China, he argued that the long-term solution lies in reactivating historical mines elsewhere. “The supply issue is real, but solvable outside of Alloyed’s remit. What we can do is reduce dependency through smarter material choices.”
The company also fabricates certified parts for Boeing’s space division at its Seattle facility, though Holmes said most technology development remains bilateral and confidential. Still, the strategy is clear: qualify faster, produce smarter, and lower unit costs for AM across both high- and low-margin segments.
Holmes is sceptical of industry efforts that focus on scaling through brute-force hardware upgrades, such as adding more lasers. “If your solution to the travelling salesman problem in a build chamber is just to hire more salesmen, you’re going to hit thermal constraints and inefficiencies,” he said. “You need optimisation of pathing, scanning strategies, and sometimes smaller, more application-specific machines.”
Rejecting the replicator logic that additive can make anything for anyone, he sees parallels with CNC machine evolution, where platform specialisation led to leaps in efficiency. “Right now, additive systems are too generic. Once the use cases are clearer, we’ll see gains from machines tailored to specific classes of part and material,” Holmes added.
For certain low-cost, high-volume applications, automotive components, for instance, Holmes believes the optimal number of lasers might top out at eight. “Twenty lasers might be useful in specific cases, but not across the board. You risk heat management problems that cancel out your productivity gains,” he said. The implication is that cost reduction in AM is as much a function of intelligent process design as it is of brute hardware scaling.
Alloyed eyes mass-scale additive applications with physics-led machine learning
Holmes underscored that machine learning at Alloyed is embedded not as a generic AI overlay but as a calibrated companion to physical modelling. “We have a physical bias. We want to understand the mechanisms driving strength, crack propagation, or overheating,” he said. For poorly understood or computationally intractable phenomena, such as electrical conductivity in copper alloys, the company uses pure Gaussian process models, trained on historical data, to complement physics-based insights.
The Alloyed CEO highlighted two key pitfalls: blind extrapolation beyond training data, and excessive computational cost. “You risk getting nonsense predictions if you don’t understand the physics and try to use a model outside its trained regime,” Holmes said. “And there are plenty of cases where advanced machine learning adds no value over more efficient algorithmic methods.”
Looking ahead, Holmes expects Alloyed’s largest growth areas to emerge not in existing defense or aerospace contracts but in sectors currently untouched by AM economics. “Additive can be five to ten times cheaper than it is today,” he said. “That unlocks entirely new consumer and industrial applications that no one’s seriously addressed yet.”
“Some of our most promising applications haven’t been announced yet,” he said. “We think some of our biggest wins will come from those.”
Holmes concluded by reframing the perception of metallurgy. “People hear ‘metallurgy’ and think new alloys. But it’s also taking a standard material like 316L or Inconel 718 and pushing it to its real performance and economic limits. That’s where the value is.”
Join AM experts on July 10th at Additive Manufacturing Advantage: Aerospace, Space & Defense. Spaces are limited for this free online event. Register now.
How is the future of 3D printing shaping up?
To stay up to date with the latest 3D printing news, don’t forget to subscribe to the 3D Printing Industry newsletter.
Featured image shows Alloyed Lattices 3D printed in metal. Photo via Alloyed.
