3D printing processes
3D printing, also known as additive manufacturing, is a transformative fabrication process that creates three-dimensional objects by successively depositing or building up material layers based on a digital model, contrasting with traditional subtractive manufacturing methods that remove material from a solid block.[1][2] This layer-by-layer approach enables the production of complex geometries with minimal waste, using materials such as plastics, metals, ceramics, and even biological tissues, and has applications across industries including aerospace, healthcare, and consumer goods.[1] The process begins with the creation of a digital 3D model using computer-aided design (CAD) software, which is then sliced into thin layers—typically 0.1 mm thick—via specialized software to generate instructions for the printer.[1] The printer interprets these instructions to deposit or solidify material precisely, often requiring post-processing steps like removing support structures, curing, or surface finishing to achieve the final part.[1] Invented in the 1980s by Chuck Hull, who patented stereolithography, 3D printing has evolved from rapid prototyping to full-scale production, reducing material waste by up to 98% and energy consumption by up to 50% compared to conventional methods.[1] Standardized by ASTM International and ISO into seven categories, 3D printing processes vary by material state (liquid, powder, sheet) and energy source (laser, heat, light), allowing customization for specific applications.[1] These processes continue to advance with hybrid systems and new materials, driving innovations in customized manufacturing while addressing challenges like speed, cost, and scalability.[1]Fundamentals
Definition and principles
3D printing, also known as additive manufacturing (AM), is a process that creates physical objects by successively joining material layers based on a three-dimensional digital model, in contrast to subtractive manufacturing methods such as computer numerical control (CNC) machining, which remove material from a solid block to form the desired shape.[4][5] The foundational principles of 3D printing begin with the preparation of a digital model, typically created using computer-aided design (CAD) software and exported in formats like STL (stereolithography) files, which represent the object's surface geometry as a mesh of triangles.[6] This model is then processed by slicing software that divides it into thin horizontal layers and generates machine-readable instructions, often in G-code format, specifying the toolpath for material deposition.[7] During fabrication, the printing system employs a build platform to support the emerging object, while components such as nozzles for extruding material, lasers for selective solidification, or other mechanisms deposit or cure the material layer by layer, adhering each new layer to the previous one through thermal, chemical, or physical bonding.[8] Resolution in 3D printing is primarily determined by layer thickness, which typically ranges from 0.05 to 0.3 mm, influencing surface finish, detail accuracy, and build time, with finer layers enabling higher precision at the cost of longer processing durations.[9] Due to the sequential layering, printed parts often exhibit anisotropy, where mechanical properties such as tensile strength vary directionally—stronger in the plane of the layers (x-y directions) but weaker perpendicular to them (z-direction)—resulting from incomplete fusion or bonding between layers.[10] This additive approach uniquely enables the fabrication of complex internal geometries, such as lattice structures for lightweighting or overhanging features without extensive support tooling, which are challenging or impossible to achieve economically with traditional subtractive or formative manufacturing techniques.[11]Layer-by-layer fabrication
The layer-by-layer fabrication process in additive manufacturing begins with the preparation of a digital 3D model, which is sliced into a series of two-dimensional cross-sections corresponding to individual layers using specialized software; each slice represents the geometry at a specified layer height, typically ranging from 0.01 to 0.3 mm depending on the process resolution requirements.[12] For each layer, material is selectively deposited or cured onto the build platform according to the slice data, forming the desired shape through mechanisms such as extrusion, powder spreading, or photopolymerization, while excess or unsolidified material is managed to ensure precise layer formation.[13] Once a layer is complete, the build platform descends along the Z-axis by the layer height to accommodate the next deposition, allowing the process to iterate sequentially from the base to the top until the full object height is achieved.[14] Interlayer bonding in this iterative process relies on physical and chemical mechanisms to ensure structural integrity across layers, primarily through molecular diffusion where polymer chains or atomic species interpenetrate at the interface when temperatures exceed the material's glass transition point, facilitated by controlled heating to reduce viscosity and promote chain mobility.[15] Mechanical interlocking contributes by creating surface roughness or geometric features that enhance grip between layers, often amplified by process-induced textures during deposition, while chemical reactions such as cross-linking in photopolymers or sintering in metals can form covalent bonds under specific energy inputs.[15] Key factors influencing these bonds include temperature control to maintain the interface above activation thresholds (e.g., 220–260°C for thermoplastics) and cooling rates, where slower cooling extends diffusion time but risks residual stresses, necessitating balanced thermal management to optimize adhesion without defects.[15] Common challenges in layer-by-layer fabrication include warping, caused by uneven thermal contraction during cooling that induces residual stresses and deforms the part, particularly in larger builds or materials with high coefficients of thermal expansion.[16] Delamination arises from weak interlayer adhesion due to rapid cooling or insufficient bonding time, leading to layer separation under mechanical load.[16] Overhangs and bridges exceeding critical angles (typically >45°) require temporary support structures to prevent collapse under gravity or sagging, which are generated during slicing and removed post-fabrication.[16] Basic mitigation strategies encompass using heated build beds to minimize thermal gradients and promote uniform cooling, thereby reducing warping and enhancing adhesion, alongside optimized build orientations to limit support volume and stress concentrations.[16] The build time T for layer-by-layer fabrication can be approximated using volume flow rate principles as T \approx \frac{V}{r}, where V is the total part volume and r is the volumetric deposition rate (in units of volume per unit time, e.g., mm³/s), representing the fundamental time required to deposit the material; this derivation assumes steady-state flow and neglects non-deposition motions, though layer height h indirectly influences r via cross-sectional flow area (e.g., extrusion width times h). Finer layer heights increase the number of layers n = \frac{H}{h} (with H as part height), extending total build time due to repeated per-layer operations and overheads. Practical estimates often include factors for Z-axis movements, travels, and other non-deposition activities.[17][18]Historical Development
Early inventions and prototypes
The conceptual foundations of 3D printing emerged in 1974 when British chemist David E. H. Jones described a layer-by-layer fabrication process in his "Ariadne" column in New Scientist, envisioning the use of a laser to solidify successive layers of photopolymer material to build three-dimensional objects from digital designs.[19] This speculative idea, presented in a science fiction context, highlighted the potential for automated, additive construction but remained theoretical without practical implementation. The transition to prototypes began in the early 1980s with Japanese researcher Hideo Kodama, who in 1980 filed the first patent for a rapid prototyping device using ultraviolet light to cure layers of liquid photopolymer resin, enabling the creation of three-dimensional plastic models from layered solidification.[20] Although Kodama's invention demonstrated the feasibility of light-based curing, it did not lead to commercialization due to funding limitations. In 1984, French engineers Alain Le Méhauté, Olivier de Witte, and Jean Claude André independently developed a similar stereolithography-like process and filed a patent for a device that used UV radiation to polymerize resin layers based on computer-controlled patterns, aimed at producing industrial part models.[21] This effort, conducted at the French General Electric Company, was abandoned shortly after due to perceived lack of commercial viability. Concurrently, American inventor Chuck Hull advanced the technology by inventing stereolithography in 1983, constructing the first functional prototype machine that used a UV laser to selectively cure epoxy-based photopolymers layer by layer, successfully producing simple proof-of-concept parts such as medical models.[22] Hull formalized the process with a patent filed in 1984, describing an apparatus for generating three-dimensional objects through precise control of laser exposure on a vat of liquid resin.[23] Early prototypes faced significant hurdles, including exorbitant costs—often exceeding $300,000 per machine—protracted build times of several hours for small parts, and confinement to photopolymer materials that limited durability and application scope.[24] These constraints restricted the technology to experimental and niche uses, underscoring the need for further refinements before broader adoption.Commercialization and key milestones
The commercialization of 3D printing began in 1986 when Chuck Hull founded 3D Systems to bring stereolithography (SLA) to market, following his patent for the technology.[25] The company released its first commercial 3D printer, the SLA-1, in 1988, marking the initial shift from prototypes to industrial applications primarily in prototyping and tooling.[26] In the 1990s, expansion accelerated with Stratasys developing fused deposition modeling (FDM) in 1989, which was patented in 1992 and enabled the extrusion of thermoplastic materials for durable prototypes.[27] This technology found early adoption in aerospace for rapid part prototyping, allowing engineers to iterate designs faster than traditional methods.[28] The 2000s saw democratization through the RepRap project, launched in 2005 by Adrian Bowyer, which open-sourced FDM designs to create self-replicating, low-cost printers.[29] The expiration of key FDM patents in 2009 further spurred the consumer market, leading to affordable hobbyist printers available by 2010 and widespread accessibility.[30] Key milestones include the publication of the ISO/ASTM 52900 standard in 2015, which standardized terminology for additive manufacturing and was updated in 2021 to reflect evolving practices.[31] The COVID-19 pandemic in 2020 accelerated adoption, with 3D printing used globally to produce personal protective equipment (PPE) like face shields and swabs amid supply shortages.[32] Recent advancements from 2023 to 2025 have focused on multi-material printing for complex, functional parts and AI-optimized designs that enhance efficiency and reduce waste.[33] Market growth has been robust, expanding from approximately $2 billion in 2012 to about $20 billion as of 2025, driven by sectors such as automotive for custom components and healthcare for patient-specific implants.[34][35]Process Classification
ISO/ASTM standard categories
The ISO/ASTM 52900:2021 standard, which revised the original 2015 edition, establishes a standardized terminology and classification framework for additive manufacturing (AM) processes, defining seven core categories based on the initial state of the feedstock material and the primary energy source or mechanism for material consolidation. These categories are binder jetting, directed energy deposition, material extrusion, material jetting, powder bed fusion, sheet lamination, and vat photopolymerization. This system ensures a common language across the AM industry, facilitating global communication, process comparison, and technological advancement.[36][37] The rationale behind this classification emphasizes the fundamental physics and mechanics of material addition and bonding, rather than specific machine implementations or end-use materials. For example, it differentiates processes by how material is deposited—such as in liquid form for jetting or powder for bed fusion—and how it is solidified, through mechanisms like thermal fusion, chemical binding, or photochemical curing. This approach promotes interoperability between equipment vendors, consistency in research protocols, and easier integration of AM into broader manufacturing ecosystems, ultimately supporting scalable industrial adoption.[38][39] Common legacy processes are mapped to these categories for clarity; fused deposition modeling (FDM), a widely used extrusion-based technique involving thermoplastic filament, falls under material extrusion, while selective laser sintering (SLS), which uses a laser to fuse polymer or metal powders, is classified as powder bed fusion. Notably, the core categories exclude hybrid processes that integrate additive manufacturing with subtractive or formative methods, reserving such combinations for specialized applications outside the primary framework.[40][41]Alternative classification approaches
Alternative classification approaches offer flexible frameworks for understanding 3D printing processes beyond the standardized ISO/ASTM categories, enabling analysis through lenses like material properties, energy mechanisms, or end-use contexts to better suit specific research or industrial needs.[42] Material-based classifications organize processes by the feedstock type and state, distinguishing discrete categories such as polymers (e.g., via material extrusion for thermoplastics), metals (e.g., via powder bed fusion), ceramics, and composites from continuous hybrids that blend materials for multifunctional parts. This method underscores how material selection influences outcomes like mechanical strength, thermal conductivity, and biocompatibility, facilitating targeted advancements in sectors like biomedical engineering.[42][43] Energy source classifications group processes according to the bonding mechanism, including thermal energy for melting (e.g., laser or electron beam in directed energy deposition), chemical reactions for curing (e.g., light-induced polymerization in vat photopolymerization), and mechanical adhesion (e.g., ultrasonic welding or lamination in sheet lamination). Hybrid energy applications, combining thermal and chemical inputs, are increasingly noted for improving efficiency in complex builds. This approach highlights the underlying physics, aiding optimization of energy consumption and material integrity.[42][43] Application-oriented classifications differentiate processes by purpose, contrasting rapid prototyping techniques like fused deposition modeling (FDM) for cost-effective, low-resolution models in design iterations with end-use methods such as directed energy deposition (DED) for durable, high-strength components in repairs or aerospace parts. This perspective prioritizes factors like speed, cost, and performance to align processes with practical demands across industries.[44][45] Other systems encompass pre-2015 frameworks from the ASTM F42 committee, which employed process-agnostic groupings based on material states (e.g., liquid, powder, solid) rather than specific techniques, and European efforts like the CECIMO Additive Manufacturing Strategy, which stress metrics such as resolution, build speed, and scalability for industrial adoption. These provide contextual insights but vary in scope.[46][47] Such alternatives, while insightful, suffer from less standardization than ISO/ASTM schemes, fostering confusion in interdisciplinary literature and complicating comparisons. Emerging research is applying artificial intelligence and machine learning for process optimization, defect classification, and quality control in additive manufacturing, potentially aiding in more personalized process selection.[48][49]Core Additive Manufacturing Processes
Vat photopolymerization
Vat photopolymerization is an additive manufacturing process that selectively cures layers of liquid photopolymer resin using ultraviolet (UV) or visible light, typically from a laser or digital projector, within a vat containing the resin. The light initiates a photochemical reaction that solidifies exposed regions of the resin, forming a solid layer that adheres to a build platform, while unexposed resin remains liquid and can be recycled for subsequent layers. This method enables the creation of complex geometries with high precision by building objects incrementally from the bottom up or top down, depending on the system configuration. Prominent variants include stereolithography (SLA), pioneered by Charles Hull in 1984 through a patent filing that described the use of a scanning UV laser to cure resin point by point, achieving lateral resolutions finer than 50 μm. Digital light processing (DLP) adapts projector technology to expose entire layers simultaneously, allowing for faster parallel curing compared to the sequential scanning in SLA. Continuous liquid interface production (CLIP), developed in 2015, advances this by creating a non-adherent "dead zone" at the build window through oxygen inhibition, enabling continuous upward motion of the platform and volumetric curing speeds up to 100 times faster than conventional SLA.[50][51] The resins employed are primarily acrylate- or epoxy-based formulations that undergo free-radical polymerization when exposed to light, converting liquid monomers and oligomers into a crosslinked solid network. Photoinitiators in the resin absorb light to generate reactive species that propagate the chain reaction, with additives controlling viscosity and cure properties. The depth of cure is governed by the working curve model, expressed as C_d = D_p \ln \left( \frac{E}{E_c} \right) where C_d is the cure depth, D_p is the light penetration depth, E is the energy exposure, and E_c is the critical exposure threshold below which no polymerization occurs; this model, derived from Beer-Lambert absorption principles, predicts how exposure parameters influence layer thickness and resolution.[52][53][54] Applications of vat photopolymerization excel in producing high-detail prototypes, intricate jewelry, and dental models, where the process delivers exceptional surface finish with roughness values (Ra) below 1 μm after minimal post-processing. These strengths stem from the photochemical precision, enabling feature sizes down to tens of micrometers and isotropic mechanical properties in some variants. However, the resulting parts often exhibit brittleness due to the inherent limitations of photopolymer networks, which can limit load-bearing applications without reinforcement.[55][56][52]Material extrusion
Material extrusion is an additive manufacturing process defined in the ISO/ASTM 52900 standard as the selective dispensing of material through a nozzle or orifice to build parts layer by layer. In this method, thermoplastic filament or pellets are fed into a heated nozzle, typically operating at temperatures between 200°C and 300°C, where the material melts and is extruded semi-continuously onto a build platform.[1] The extruded material solidifies upon cooling in ambient air, and the platform or print head moves along the Z-axis to deposit successive layers, forming the three-dimensional structure.[57] The most common variant is fused deposition modeling (FDM), also known as fused filament fabrication (FFF), which primarily uses thermoplastic filaments such as polylactic acid (PLA) or acrylonitrile butadiene styrene (ABS) for prototyping and functional parts.[58] Another key variant is direct ink writing (DIW), which employs shear-thinning pastes, gels, or inks suitable for applications like bioprinting of soft tissues or ceramics, where the material maintains flow under pressure but solidifies post-extrusion without heat.[58] These variants enable a range of viscosities and material properties, with FDM/FFF dominating consumer and educational use due to its simplicity.[59] The physics of melt flow in material extrusion is governed by the Hagen-Poiseuille equation, which describes laminar flow through a cylindrical nozzle: \eta = \frac{\pi r^4 \Delta P}{8 L Q} where \eta is the viscosity, r is the nozzle radius, \Delta P is the pressure drop, L is the nozzle length, and Q is the volumetric flow rate.[60] This equation helps predict the pressure required for consistent extrusion, ensuring uniform bead deposition and minimizing defects like under-extrusion.[61] Material extrusion excels in rapid prototyping and educational settings, where low-cost printers priced under $500 make it accessible for hobbyists and institutions.[62] Its advantages include multi-material capability through dual-extruder systems, allowing integration of supports or composites in a single build, and versatility for larger parts up to several meters in industrial setups.[63] However, limitations such as visible layer lines from typical resolutions of 0.1-0.4 mm and reduced mechanical strength (often 20-50% lower than injection-molded equivalents due to anisotropic bonding) restrict its use in high-load applications.[57]Powder bed fusion
Powder bed fusion (PBF) is an additive manufacturing process that selectively fuses regions of a powder bed using a focused energy source, typically a laser or electron beam, to create solid parts layer by layer. A thin layer of powder is evenly spread across a build platform using a recoater blade or roller, forming a uniform bed typically 20–100 μm thick. The energy source then scans the surface according to a digital model, melting or sintering the powder particles in the desired areas to form a cross-section of the part; the unfused powder remains as support for subsequent layers. After each layer, the platform lowers, and a new powder layer is applied, repeating the cycle until the part is complete. Post-processing involves removing the excess powder through sieving or blasting, often followed by heat treatment to relieve stresses.[4][64][65][66][67] Key variants of PBF include selective laser sintering (SLS), primarily for polymers, which originated in the 1980s at the University of Texas under Carl Deckard and uses a CO₂ laser to sinter powders like nylon into functional prototypes and end-use parts without full melting. For metals, selective laser melting (SLM) and direct metal laser sintering (DMLS)—a trademarked term by EOS—employ a high-power fiber laser to fully melt metal powders such as titanium or stainless steel, achieving densities exceeding 99% and enabling high-strength components. Electron beam melting (EBM), developed by Arcam (now GE Additive), uses an electron beam in a high-vacuum environment to melt powders, particularly titanium alloys like Ti-6Al-4V, minimizing oxidation and residual stresses during processing.[68][69][70][71][72][73][74][75] The physics of fusion in PBF relies on thermal input from a Gaussian-profiled beam, where the volumetric energy density E governs melting and porosity reduction, calculated as E = \frac{P}{v \cdot h \cdot d} (in J/mm³), with P as laser power (W), v as scan speed (mm/s), h as hatch spacing (mm), and d as beam spot diameter (mm). Optimal E values, typically 50–200 J/mm³ for metals, ensure complete fusion while avoiding defects like keyhole porosity from excessive heat or lack of fusion from insufficient energy. This parameter balances heat conduction, powder absorptivity, and cooling rates, influencing microstructure such as fine grains and β-phase retention in titanium.[76][77][78] PBF excels in applications requiring complex geometries with isotropic mechanical properties, such as aerospace turbine blades from Inconel alloys and custom medical implants like hip replacements from Ti-6Al-4V, leveraging the powder bed for self-support of overhangs and internal channels. These processes yield parts with near-full density and uniform strength in all directions, unlike directional methods. However, challenges include high equipment costs (often exceeding $500,000 for industrial systems) and thermal stresses from rapid heating-cooling cycles, which can cause warping or cracking without careful parameter tuning or supports.[79][80][81][82][83][84]Binder jetting
Binder jetting is an additive manufacturing process in which a liquid bonding agent is selectively deposited to join powder materials, classified as one of the seven categories in the ISO/ASTM 52900 standard.[4] This method operates at room temperature, enabling multi-material capabilities without thermal fusion, and is particularly suited for rapid prototyping and complex geometries. The process begins with the deposition of a thin layer of powder, typically 50-100 μm thick, spread evenly across a build platform using a roller or blade mechanism. A printhead, similar to those in inkjet printers, then selectively applies the liquid binder to the powder bed according to the digital model, where it infiltrates the particles through capillary action and evaporates to form a "green" part with initial cohesion. This layer-by-layer building continues until the object is complete, followed by post-processing such as depowdering to remove unbound material and sintering to enhance mechanical strength.[85] Key variants of binder jetting include ColorJet Printing (CJP), which uses multi-colored binders for full-color models and sand molds in foundry applications, allowing intricate patterns without additional pigmentation steps. Metal Binder Jetting (MBJ) involves depositing binder onto metal powders followed by infiltration with a secondary material, such as bronze, to achieve higher density and strength for functional parts. In ceramics, binder jetting is applied in dentistry for creating biocompatible crowns and bridges, leveraging the process's precision for patient-specific restorations. These variants expand the technology's versatility across materials like polymers, metals, and ceramics, often achieving build rates up to 200 cm³/h.[85][42][86] The physics of binder jetting relies on capillary action, where the liquid binder wets and penetrates the powder pores driven by surface tension, followed by evaporation that solidifies the bonds between particles. Binder saturation, defined as S = \frac{V_{\text{binder}}}{V_{\text{pores}}}, quantifies the volume of binder relative to the available pore space in the powder bed, typically optimized around 60% to balance green part strength and avoid excess liquid that could cause distortion. This saturation directly influences the initial "green" density and tensile strength, which is crucial before sintering, during which volumetric shrinkage of approximately 15-20% occurs due to particle rearrangement and densification.[85][87] Applications of binder jetting prominently feature full-color models for visualization and design validation, as well as foundry patterns like sand cores and molds that enable complex metal castings with reduced lead times. Its advantages include high-speed production without melting, low energy consumption, and compatibility with a wide range of powders, making it cost-effective for batch manufacturing. However, limitations such as inherent porosity in green parts (often 40-60% void fraction) necessitate post-processing like infiltration or sintering, which can introduce shrinkage and require precise control to maintain dimensional accuracy.[85][88]Material jetting
Material jetting is an additive manufacturing process defined in the ISO/ASTM 52900:2021 standard as one in which droplets of build material are selectively deposited layer by layer and solidified using curing methods such as ultraviolet (UV) light.[89] This technique enables high-resolution printing with layer thicknesses as low as 16 μm, supporting multi-material and full-color fabrication for complex geometries.[90] Unlike immersion-based methods, material jetting operates on a drop-on-demand principle, allowing precise placement of photopolymers or waxes directly onto the build platform.[91] The core mechanism involves piezoelectric printheads that eject tiny droplets of liquid material, typically photopolymers heated to 30–70°C to reduce viscosity, onto a build platform.[90] These droplets solidify almost immediately upon deposition via ultraviolet (UV) light exposure (190–400 nm wavelength), either in-flight or on the platform, forming a cured layer before the platform lowers for the next pass.[90] Support structures, often wax-like materials, are jetted simultaneously to enable overhangs and intricate designs, later removed through methods such as water jetting, sonication, or dissolution in sodium hydroxide.[92] This process mirrors inkjet printing but scales to three dimensions, with printheads moving in coordinated axes to deposit multiple materials in a single layer for gradient or composite parts.[91] Key variants include PolyJet technology developed by Stratasys, which primarily uses photopolymers like Vero series for multi-material prototypes, enabling combinations of rigid and flexible properties in one print.[90] MultiJet Printing (MJP), commercialized by 3D Systems, extends this to wax-based binders for investment casting applications, where patterns are created for metal molding.[91] Emerging approaches incorporate nanoparticle jetting for metals, depositing metallic inks that require post-processing sintering to achieve conductivity and density, expanding beyond polymers.[90] Physically, droplet formation occurs through the Rayleigh-Plateau instability, where surface tension breaks a liquid jet into uniform droplets controlled by printhead nozzle diameter and ejection pressure.[91] Curing kinetics depend on UV intensity I, with gelation time approximated as t_{\text{gel}} \approx \frac{1}{k \cdot I}, where k is a material-specific rate constant, ensuring rapid solidification to prevent sagging in fine features.[91] Photopolymer curing in material jetting follows similar radical polymerization as in vat processes but benefits from localized exposure for sharper interfaces.[90] Applications of material jetting excel in producing realistic prototypes with smooth, glossy surfaces and tolerances down to 15 μm, ideal for functional testing and visual models.[90] In medicine, it fabricates tissue-mimicking phantoms for surgical planning, such as liver models for living-donor transplants using flexible TangoPlus materials.[92] Its multi-material capability supports color and texture integration, enhancing applications in consumer product design and dental aligners.[91] However, limitations include a narrow material palette dominated by photopolymers, restricting use in high-strength or conductive parts without post-processing, alongside elevated costs due to specialized inks and equipment.[90]Directed energy deposition
Directed energy deposition (DED) is an additive manufacturing process that builds three-dimensional objects by focusing a high-energy beam—such as a laser, electron beam, or electric arc—onto a substrate to create a localized melt pool, into which material in the form of powder or wire is simultaneously fed and melted to deposit layers of material.[93] The process typically occurs in a coaxial setup, where the energy source and material feedstock are aligned to enable multi-axis movement of the deposition head relative to the substrate, allowing for freeform building, feature addition, or repair on existing components without the constraints of a pre-defined powder bed.[93] This method is particularly suited for metallic materials and large-scale applications due to its ability to handle high material throughput and integrate with robotic systems for complex geometries.[94] Key variants of DED include laser metal deposition (LMD), which employs a laser beam to melt fine metal powders delivered through a nozzle, enabling precise deposition for applications requiring fine features and low dilution rates.[95] Wire arc additive manufacturing (WAAM) uses an electric arc as the energy source to melt continuous wire feedstock, offering a cost-effective approach for fabricating large structures with centimeter-scale features and deposition rates up to several kilograms per hour, making it ideal for industrial-scale production.[94] Electron beam DED, conducted in a vacuum environment, is advantageous for processing reactive metals like titanium alloys, as the vacuum minimizes oxidation and contamination during melting.[96] The physics of DED revolves around melt pool dynamics, where the energy input governs the pool's size, shape, and solidification behavior, influenced by factors such as surface tension gradients driving Marangoni convection.[93] A foundational model for predicting the temperature distribution in the melt pool is the Rosenthal equation, which approximates the thermal field from a moving point heat source: T(r) = \frac{q}{2\pi k r} \exp\left(-\frac{v r}{2\alpha}\right) + T_0 Here, T(r) is the temperature at distance r from the source, q is the heat input, k is the thermal conductivity, v is the scan speed, \alpha is the thermal diffusivity, and T_0 is the initial temperature; this simplified form helps estimate thermal gradients and cooling rates critical for microstructure control.[97] DED finds prominent applications in tool repair, such as restoring worn turbine blades, and aerospace cladding, where it deposits protective or functional layers on high-value components like Ti-6Al-4V engine parts.[93] Its advantages include high deposition rates (often 1-10 kg/hr depending on the variant), near-net-shape fabrication with minimal material waste compared to subtractive methods, and the flexibility to repair or modify parts in situ.[94] However, challenges persist, including porosity formation due to gas entrapment in the melt pool and rough surface finishes necessitating post-processing, which can affect mechanical integrity if process parameters like heat input and feed rate are not optimized.[93]Sheet lamination
Sheet lamination is an additive manufacturing process defined by the ISO/ASTM 52900 standard, involving the bonding of sheets of material, such as paper, metal foil, or composites, which are then selectively cut to shape and stacked layer by layer to form a three-dimensional object.[98] The process relies on mechanical bonding techniques like adhesives, heat, pressure, or ultrasonic welding, without requiring the material to enter a liquid or powder state, enabling the use of a wide range of pre-formed sheet materials.[99] Cutting is typically performed using a laser, blade, or knife after or during lamination, allowing for precise contouring while excess material is often cross-hatched for easy removal post-build.[99] Key variants include Laminated Object Manufacturing (LOM), developed by Helisys Inc. in 1991, which uses adhesive-coated paper or plastic sheets fed from a roll, bonded with heated rollers, and cut with a CO2 laser to create low-cost prototypes.[100] In LOM, each layer is pressed onto the previous one under heat and pressure, with the adhesive activated to form bonds, making it suitable for rapid production of conceptual models due to its affordability and speed.[101] Another variant is Ultrasonic Additive Manufacturing (UAM), a solid-state process that bonds thin metal foils (typically 100-200 μm thick) using high-frequency ultrasonic vibrations at around 20 kHz, combined with a normal force from a sonotrode and periodic CNC machining for shaping.[102] UAM, pioneered in the early 2000s, avoids melting the material, operating at temperatures below 100-250°C, which preserves material properties and allows embedding of temperature-sensitive components like electronics.[103] The physics of bonding in sheet lamination emphasizes interlayer adhesion through diffusion welding in solid-state variants like UAM, where ultrasonic energy induces severe plastic deformation at the foil interfaces, breaking oxide layers and promoting atomic diffusion via vacancy mechanisms to form metallurgical bonds without bulk heating.[104] For adhesive-based processes like LOM, bond strength is governed by peel resistance, which can be modeled as the peel force P = \sigma \times t \times w, where \sigma is the shear stress of the adhesive layer, t is the adhesive thickness, and w is the bond width; this equation highlights how peel strength scales with the adhesive's shear capacity and contact area.[105] In UAM, bond quality is further quantified by linear weld density, often approaching 100% under optimized parameters such as sonotrode amplitude above 25 μm and welding speeds of 10-30 mm/s, yielding shear strengths exceeding 100 MPa in materials like aluminum-titanium composites.[104] Applications of sheet lamination span architectural and conceptual modeling with LOM, where its ability to produce large-scale, detailed prototypes from inexpensive paper enables visualization in design reviews and pattern making for foundries.[99] UAM excels in functional metal parts, such as orthopedic implants and aerospace components, particularly for embedding electronics or sensors within solid metal matrices due to its low-temperature bonding and compatibility with dissimilar materials like aluminum and copper.[103] Advantages include high material versatility, rapid layer deposition rates (up to several meters per minute in UAM), and cost-effectiveness for large builds, with surplus material often reusable.[98] However, limitations arise from stair-stepping effects on curved surfaces, requiring post-processing for smoothness, generation of waste from uncut sheets, and challenges in achieving uniform bond strength in complex geometries, which can lead to anisotropic mechanical properties.[99]Hybrid and Emerging Processes
Additive friction stir deposition
Additive friction stir deposition (AFSD) is a solid-state additive manufacturing process that deposits metal feedstock through frictional heating and plastic deformation, avoiding full melting to produce defect-free parts with fine microstructures. In this hybrid technique, a rotating tool engages a consumable rod or powder feedstock, generating heat from friction and deformation to soften the material to approximately 80% of its melting point, enabling extrusion and forging onto a substrate or prior layer without forming a melt pool. The deposited material solidifies rapidly through dynamic recrystallization, resulting in strong metallurgical bonding via mechanical interlocking and diffusion at the interface. This process differs from directed energy deposition by relying on mechanical stirring rather than thermal beams, allowing for low-distortion builds with preserved material properties. AFSD was developed in the 2010s by AeroProbe Corporation, building on friction stir welding principles to address limitations in traditional metal additive manufacturing, such as porosity and cracking from melting. Significant research and early demonstrations were conducted at Brigham Young University (BYU) and collaborators, including AeroProbe Corporation, focused on adapting friction stir tools for layer-by-layer deposition, with seminal demonstrations using aluminum alloys to achieve near-bulk density. Variants emerged, such as refill friction stir spot deposition for localized repairs, expanding its utility beyond bulk builds. By 2025, advancements include integration with robotic systems for enhanced scalability, as pursued by collaborations between BYU, Mazak MegaStir, and Bechtel, enabling commercial applications in large-scale manufacturing. As of 2025, NSF-funded projects, such as a $896,186 grant to Southern University (July 2025–June 2028), focus on scaling AFSD for broader applications.[106] The physics of AFSD centers on thermomechanical deformation, where frictional heat generation drives material flow and bonding. Heat input is primarily governed by the equation Q = \mu F \omega r, where Q is the heat generation rate, \mu is the friction coefficient, F is the axial force, \omega is the tool rotation speed, and r is the radius of contact; this softens the feedstock without exceeding the solidus temperature, promoting shear-thinning flow and recrystallization upon cooling. Strain rates during deposition range from 10 to 100 s⁻¹, with temperatures typically between 385–570°C for aluminum and 700–1000°C for titanium, leading to refined grain structures and high-strength interfaces in materials like aluminum and titanium. These conditions minimize residual stresses and distortions compared to fusion-based methods, enabling defect-free parts with tensile strengths approaching wrought equivalents. AFSD finds primary applications in aerospace for component repairs and multi-material layering, where its solid-state nature preserves alloy integrity and allows deposition of dissimilar metals like aluminum onto titanium substrates with minimal distortion. Advantages include high deposition rates (up to several kg/h) and superior mechanical properties, such as yield strengths exceeding 300 MPa in aluminum builds, making it suitable for structural components. Recent integrations with robotics have improved precision and scalability for industrial use, supporting applications in automotive cladding and large-format manufacturing while reducing energy consumption relative to laser-based processes.Computed axial lithography
Computed axial lithography (CAL) is a volumetric additive manufacturing technique that enables the rapid, layerless fabrication of three-dimensional objects by simultaneously curing an entire volume of photosensitive resin using computed light projections. Unlike traditional layer-by-layer methods, CAL rotates a vial containing the resin while projecting synchronized two-dimensional light patterns from multiple angles, allowing light rays to intersect and deliver a precise dose throughout the desired geometry, solidifying it via photopolymerization in seconds to minutes. This approach draws from principles of computed tomography, where the inverse problem of reconstructing a 3D volume from projections is adapted to control light dosage for uniform curing.[107] The mechanism relies on a rotating cylindrical container of photocurable resin, typically illuminated by a digital light processing projector or similar source emitting at wavelengths like 405 nm, which penetrates the material. As the vial rotates—often at rates of 10–30 degrees per second—dynamic binary or grayscale images are projected to modulate light intensity, ensuring that regions intended to solidify receive a cumulative exposure exceeding a critical threshold while uncured areas remain below it. This tomographic illumination avoids the need for mechanical supports or post-processing removal of layers, enabling complex, overhang-free structures with isotropic mechanical properties and smooth surfaces free of stair-stepping artifacts. High-viscosity resins, up to 90,000 centipoise, can be used due to the lack of flow requirements between layers, expanding material options beyond low-viscosity photopolymers.[107] At its core, the physics of CAL involves solving the exponential Radon transform to compute the necessary projection patterns that achieve a target light dose distribution, accounting for light absorption and scattering in the resin. The absorbed dose at a point (r, z) in cylindrical coordinates is given byD(r, z) = \frac{1}{\Omega} \int_0^{2\pi} g(r, \theta, z) e^{-\alpha r} \, d\theta,
where g(r, \theta, z) is the projected intensity at angle \theta, \alpha is the material's absorption coefficient, and \Omega is the rotation rate; the solid object forms where D(r, z) \geq D_c, the critical polymerization dose. This formulation, derived from back-projection algorithms similar to those in medical imaging, minimizes shadowing effects by distributing exposure over the rotation, promoting uniform curing and reducing over- or under-exposure in dense geometries. The process approximates solutions to the Helmholtz equation for light propagation in absorbing media, ensuring intensity distributions that align with the resin's Beer-Lambert absorption law.[107] Developed by researchers at the University of California, Berkeley, including Hayden Taylor and Brett Kelly, CAL was first demonstrated in 2019 through experiments fabricating objects up to 55 mm in height, such as lattice structures and embedded components, in 30–300 seconds—orders of magnitude faster than comparable stereolithography processes. The technique has since evolved, with open-source implementations available and commercial adaptations like xoloGraphy emerging for broader accessibility. By 2024, Berkeley teams tested CAL in microgravity aboard a Virgin Galactic flight, validating its potential for space manufacturing of tools and biomedical parts without supports. As of 2025, advancements include adaptations for inorganic materials, such as nanoparticle-laden resins converted to bioactive glasses and ceramics post-printing.[107][108][109] In applications, CAL excels in fabricating soft, compliant structures for biomedical scaffolds using hydrogels like gelatin methacrylate, which support cell growth for tissue engineering, and elastomers for soft robotics components that require flexibility and rapid prototyping. It also enables optics and microfluidics with transparent multimaterial prints, as well as bone graft scaffolds from bioactive ceramics like β-tricalcium phosphate, achieving feature resolutions down to 45 μm after thermal processing. Key advantages include high build speeds equivalent to 100–500 μm/min axially for cm-scale parts, inherent material isotropy yielding uniform tensile strengths (e.g., 20–40 MPa in acrylates), and the ability to overprint existing objects without disassembly. However, challenges persist in resolution, typically 100–500 μm due to optical diffraction and absorption limits, and material constraints, as resins must remain transparent to the curing wavelength; scattering in nanoparticle suspensions and shrinkage (up to 50% in ceramics) further complicate scaling to high-precision or diverse compositions as of 2025.[107][109][110]