Design for Additive Manufacturing at CDFAM: Part 1

Why CDFAM Was Founded: Computational Design Thinking for Additive Manufacturing (and Beyond)

Design for Additive Manufacturing (DfAM), and the lack of a dedicated platform for its exploration and discussion, was the impetus behind the inaugural CDFAM event held in New York City in 2023.

Over the preceding decade, more than $10 billion had been invested in the development of machines and materials that have advanced 3D printing from a rapid prototyping tool to a viable method for production. However, the evolution of software to optimize designs for additive processes has received only a fraction of that financial investment and focus from both academia and industry.

The tools that enable designers to create geometries suited for additive production, ones that justify the cost by improving part performance, often rely on a data-driven approach that can only be effectively executed using computational design thinking.

We are not just talking about modifying for additive manufacturability, your 45 degree overhang constraint, and minimum wall thickness, but true Design for Additive Manufacturing…

CDFAM was originally conceived as a response to this gap, adding the “C” for computational to DfAM to address the design challenges and opportunities emerging at the intersection of computational engineering and advanced manufacturing.

While the series has since expanded to encompass a broader range of manufacturing methods and scales, agnostic of andy given process, many of the presentations at the first CDFAM, and those that followed, continue to focus on the unique demands and innovations in design for additive manufacturing.

CDFAM NYC 2023: The Foundations of Computational DfAM

The inaugural CDFAM symposium provided a comprehensive and critical view into how computational design methods are transforming the additive manufacturing landscape.

Experts from aerospace, architecture, healthcare, and materials science presented diverse approaches to applying DfAM, not only as a generative methodology but also as a practical framework constrained by machine capabilities and production demands (yes, including overhang constraints).

The following presentations showcased a wide range of strategies in design for additive manufacturing.

From engineering lattice structures to fine-tune material properties to applying topology optimization for structural efficiency, speakers demonstrated how software tools, simulation workflows and perhaps more importantly, computational thinking, are reshaping how parts are designed for performance.

These sessions highlighted the value of aligning geometric design with strength, weight, function, cost, and the specific requirements of additive processes.

Equally important was the focus on optimizing the additive process itself.

Talks addressed build strategies at the toolpath level, where raster orientation, bead overlap, and deposition sequence are treated as critical variables.

In some cases, these principles were scaled up to architectural applications, where robotic systems adapt fabrication methods in real time.

The result is a view of DfAM that extends beyond surface geometry to embed performance, feedback, and process control into every layer of the build.

Together, these talks painted a comprehensive picture of DfAM’s current state: a rapidly evolving design discipline grounded in physics and simulation, shaped by data-rich workflows, and essential for realizing the full potential of 3D printing technologies.

Expanding the Boundaries of Design for Additive Manufacturing

Vision and Reality: Foundational Voices on Design for Additive Manufacturing

The keynote from Neil Gershenfeld set the intellectual foundation for the event.

Drawing from his experience leading MIT’s Center for Bits and Atoms, he presented a compelling vision of digital fabrication that links computational intent directly to physical output.

From the early development of fab labs to current research in self-assembling robotic systems, Gershenfeld emphasized fabrication as an information process rather than a sequence of mechanical operations.

He encouraged attendees to rethink the relationship between design and fabrication, imagining manufacturing systems that are autonomous, distributed, and materially responsive.

Ronald Rael expanded this idea with examples of architectural-scale additive manufacturing, emphasizing the use of the toolpath itself as a design medium, unmediated by mesh or abstraction.

His inspiring presentation placed AM in a broader historical context, linking experimental material practices of the past with emerging capabilities in robotic digital construction.

This theme of using toolpaths not just for production but as a core design element was further developed by Alex Roschli of Oak Ridge National Laboratory, who demonstrated how material properties can be encoded through toolpath control. His presentation focused on the relationship between deposition patterns and resulting mechanical performance, thermal properties, and structural integrity. Roschli showed how additive manufacturing can go beyond shape to become a method for spatially programming performance. The approach has implications for lightweighting, cooling systems, and multifunctional components, all driven directly from toolpath-level decisions.

Real-World Applications and Scalable Workflows in Design for Additive Manufacturing

Ryan McClelland of NASA offered a reality check, demonstrating how topology optimization directly informed aerospace components, while also acknowledging that DfAM often reveals the limitations of additive and the need to pivot to other manufacturing methods when constraints demand it.

From the perspective of industrial design and scalable workflows, Nathan Shirley of HP¹ explored how generative methods can expand additive’s production potential. His presentation highlighted how design automation and optimization routines can reduce iteration time and help scale additive workflows across product lines. He illustrated this with examples of HP’s Multi Jet Fusion process, demonstrating how part families can be rapidly adapted to material and process constraints using field-driven design.

This emphasis on scalable, parametric design was echoed by the team from Mode Lab, who presented strategies for building repeatable, data-driven workflows using computational tools. Their talk emphasized the importance of connecting geometric logic to manufacturing realities, helping organizations bridge design intent with production-ready outcomes.

Carbon also exemplified the application of DfAM at scale, presenting their integrated platform that combines software, hardware, and materials to enable high-volume production with additive manufacturing. Their Design Engine Software allow for complex lattice structures and performance-driven geometries to be tailored directly to the capabilities of Carbon’s Digital Light Synthesis (DLS) process.

This vertical integration has enabled commercial successes like the Adidas Futurecraft 4D footwear, where midsole geometries are optimized for performance, comfort, and printability across millions of units. Carbon’s platform demonstrates how a tightly integrated design-to-production workflow can turn advanced DfAM into scalable consumer products.

Onur Yüce Gün of New Balance delivered the opening keynote that bridged the gap between computational design research and commercial product development. Drawing from his experience at the intersection of architecture, computation, and industrial design, Onur shared how computational workflows are being integrated into the development of high-performance footwear.

He discussed how geometry, material, and manufacturing constraints are evaluated in tandem using parametric modeling and simulation, enabling the creation of responsive, optimized structures tailored to athletic use.

His presentation emphasized the importance of building computational literacy within design teams and illustrated how the ethos of research-driven experimentation can inform scalable commercial outcomes.

Matt Shomper mapped biomimetic forms into manufacturable designs, using custom computational workflows to translate complex natural geometries into print-ready structures. His work emphasized the importance of tailoring internal architecture to function, whether for mechanical performance or material efficiency. This same approach was extended to the design of medical devices, where biological analogs informed internal geometry optimized for both performance and patient-specific constraints.

Adam Wentworth of Mayo Clinic outlined the institutional challenges of bringing personalized medical devices to clinical reality. His presentation focused on the regulatory, logistical, and technical barriers to scaling patient-specific additive manufacturing in a clinical environment.

Ryan Carter of Johns Hopkins APL spoke candidly about the friction between optimized geometries and deployable parts. He emphasized the importance of reconciling high-performance computational designs with operational realities, particularly within aerospace and defense applications.

Chelsea Cummings from The Barnes Global Advisors added to this perspective by making a compelling case for aligning DfAM with Manufacturing for AM (MfAM), arguing that process awareness must come before geometric freedom. Her presentation underscored the importance of understanding machine behavior, process parameters, and post-processing requirements as foundational to the success of computational design workflows.

Meanwhile, experimental work at the edge of biology and computation was presented by Natalie Alima (Biolab + Adidas), who merged robotic systems with living materials in an eclectic presentation that ranged from training robots to nurture the growth of mycelium to the creation of experimental footwear, famously including mushroom-based Yeezys for Kanye.

Her work explored feedback-driven fabrication using living organisms, pushing the boundaries of material agency and design authorship.

Software Tools Driving DfAM Innovation

Marek Moffett of General Lattice discussed how performance data can be embedded into digital material systems, illustrating how geometric lattices can be designed to respond to mechanical stress, energy absorption, or damping requirements. One high-profile example of this approach is the Wilson airless basketball, a collaboration in which General Lattice applied digital material principles to produce a 3D printed ball with equivalent bounce and grip to its inflated counterpart, using lattice engineering alone.

Metafold also contributed to this theme with a presentation demonstrating how their cloud-based geometry engine connects lattice and metamaterial design directly to machine output. By streamlining the translation from parametric structure to printable format, Metafold showcased how performance-optimized geometries can be delivered at scale with minimal manual intervention. Their approach highlights the growing importance of linking computational design to actual machine instructions in an efficient, scalable workflow.

Rhushik Matroja of Cognitive Design Systems showed how machine learning could help designers address manufacturability constraints early in the CAD process. By integrating predictive feedback into the design workflow, his team aims to reduce costly iterations and ensure print-ready results.

Kirill Volchek of Oqton explained hybrid workflows connecting topology, simulation, and toolpath, enabling performance-driven design outputs to translate efficiently into manufacturing-ready instructions. His presentation emphasized the need for seamless integration between design intent, simulation validation, and toolpath planning within a single workflow.

Complementing this, the team from nTop demonstrated how their field-driven design platform extends beyond lattice generation to encompass metamaterials and full-scale topology optimization.

Their presentation highlighted the use of spatially varying fields to drive material distribution, stiffness gradients, and structural performance criteria.

By unifying design and simulation within a single computational environment, nTop showed how engineers can generate parts that are not only optimized for function but inherently manufacturable for a range of additive processes. These concepts will be expanded by nTop in their upcoming presentation at CDFAM Amsterdam.

Educating the Next Generation of DfAM Practitioners

Education and long-term practice were also part of the conversation. John Hart of MIT and Tim Simpson the OG DFAM expert from Penn State each reflected on a decade of teaching and applying DfAM, offering insights from their leadership in two of the most established additive manufacturing programs in the United States.

Hart emphasized the integration of DfAM into engineering curricula and industry training, highlighting the importance of combining theory with practical experience. He described how MIT incorporates hands-on additive projects, material science modules, and simulation tools into its coursework to prepare students for real-world challenges.

By doing so, students not only learn how to use DfAM software but also develop a deeper understanding of the relationships between design, material behavior, and manufacturing constraints. Simpson highlighted case studies where academic research directly informed industrial applications across aerospace, automotive, and healthcare sectors. Their shared perspective emphasized the need to bridge education with real-world deployment through hands-on, multidisciplinary engagement.

Looking Ahead: DfAM at CDFAM Amsterdam 2025

This article has explored how Design for Additive Manufacturing emerged as a central theme at the first CDFAM symposium in 2023, where foundational thinkers, software developers, educators, and engineers outlined both the promise and the complexity of designing for 3D printing.

From pioneering work in topology optimization and field-driven design, to experimental biofabrication and toolpath-level innovation, the event revealed the diversity of approaches shaping the DfAM landscape.

This year, CDFAM Amsterdam 2025 will continue to build on these foundational ideas with a new wave of presentations exploring the frontier of (Computational) Design for Additive Manufacturing will include:

• Alessandro Zomparelli, AiBuild
• Aaron Porterfield, F=F
• Bradley Rothenberg, nTop
• Carbon
• Chris Robinson, Ansys
• Christian Waldvogel, Spherene
• Hamilton Forsythe, Rapid Liquid Print
• Luca Breseghello, DTU
• Mathias Fuchs, McNeel & cydric
• Or Benifla, CustoMED
• Rhushik Matroja, Cognitive Design Systems
• Thomas Rees, ToffeeX
• Tiffany Cheng, Cornell University
• Vanessa Cool, KU Leuven
• Vittoria Laghi, University of Bologna
• Wesley Essink, Altair

Topics will range from toolpath-level performance control, AI-driven manufacturability guidance, field-driven structural optimization, and scalable production of customized parts.

To learn more and take part in these conversations, register to attend CDFAM Amsterdam and connect with the engineers, researchers, and designers shaping the future of computational design and additive manufacturing.

Register to attend CDFAM

^{1 The interview with Nathan was actually the first time that we had used the term CDFAM somewhat as a joke, prior to the foundation of the CDFAM symposium, but, here we are now…}