Additive manufacturing has enabled the fabrication of sparse lattice structures with complex geometries tailored to specific uses. These lattice infills provide attractive properties while weighing less than a fully solid part and satisfying the strength and stiffness requirements. Historically, variations in infill lattices have altered either the global geometry or the infill wall thickness to increase strength in high load regions of structures. In this work we expand the generation of these lattices to continuously vary the local infill density at each point within the part volume, allowing for further customization of the mechanical properties. Enabled by a modified penalty-based topology optimization method and an adaptation of frequency modulation, this work continuously optimizes the density of lattice infill geometry while maintaining a constant wall thickness to maximize print performance, mechanical properties, and infill continuity throughout the structure with deposition-based fabrication. These stiffness-optimized lattice structures were successfully printed, tested, and compared to equivalent constant density structures, resulting in stiffness increases from up to 15.8% without impact to material usage or fabrication time. By increasing the stiffness to mass ratio of structures through this infill, it allows for increased adoption of 3D printed components in aerospace, medical, and other applications.

Additive manufacturing (AM) allows for manufacturing of complex three-dimensional geometries not typically realizable with standard manufacturing practices. The internal microstructure of AM components has a significant impact on mechanical, vibrational, and shock properties and permits richer design space when this is controllable. Due to complex interactions of internal geometry of an extrusion-based AM component, it is common practice to assume homogeneous behavior or to perform characterization testing on specific toolpath configurations. To avoid testing or material waste, it is necessary to develop a consistently accurate numerical simulation framework with relevant boundary value problems that can handle the complicated geometry of internal material microstructure present in AM components. Herein, a framework is proposed to directly create computational meshes suitable for finite element analysis (FEA) of the fine-scale features generated from extrusion-based AM tool paths to maintain a strong process-structure-property-performance linkage. This mesh can be manually or automatically analyzed using standard FEA simulations such as quasi-static preloading or modal analysis. The framework allows an in-silico assessment of a target AM geometry where fine-scale features greatly impact quantities of design interest such as in soft elastomeric lattices where toolpath infill can greatly influence the self-contact of a structure in compression, which we use as a motivating exemplar. This approach greatly reduces both time and resource waste present in traditional build and test design cycles for non-intuitive design spaces, and acts as a tool for use in the production of a key component of a digital twin, a mesh suitable for finite element analysis. It also further allows for the exploration of toolpath infill to optimize component properties beyond simple linear properties such as density and stiffness.

Laser direct energy deposition (LDED) is a prominent additive manufacturing (AM) technique due to its high deposition rates, scalability, and material efficiency. However, maintaining consistent part quality remains a significant challenge due to complex thermal cycles inherent to high-deposition processes, where rapid heat accumulation and varying cooling rates lead to heterogeneous microstructures and residual stress. This work introduces a data-driven in-situ process monitoring and control (ISPMC) framework for wire-fed LDED that combines high-speed infrared (IR) tomography, computer vision, and real-time process parameter adjustment to improve print quality. Here, high-speed infrared (IR) thermography was used to capture optical heat mapping and temperature history data, providing real-time insights into heat transfer and microstructural evolution. By integrating the IR data with computer vision, a predictive framework was developed for real-time adaptive control of LDED process parameters. Experimental validation demonstrated significant improvements in thermal stability, visual quality, and microstructural homogeneity across complex geometries. Optical microscopy revealed that No ISPMC samples developed dendritic and cellular grain structures, whereas ISPMC samples produced consistent equiaxed cellular grains throughout the build. Mechanical testing showed up to a 90% reduction in ductility variance and a 49% reduction in ultimate tensile strength variance compared to the No ISPMC samples. Nano-indentation further confirmed reduced hardness and modulus gradients throughout the build height, demonstrating improved thermal and mechanical homogeneity. These results establish ISPMC as an effective closed-loop control framework for reliable production of high-quality metal AM components.

For semicrystalline polyolefin thermoplastics, the balance between interconnected ordered crystalline and disordered amorphous regions is paramount to their performance and processability. However, contemporary manufacturing strategies, from injection molding to three-dimensional (3D) printing, result in monolithic objects, unable to spatially encode crystallinity. We develop a light-based approach for fabricating mechanically robust polyolefin thermoplastics with microscopic control over crystallinity in 3D space. Light dosage governs polymer stereochemistry giving access to a continuum of materials, from strong rigid plastics, such as high-density polyethylene, to more extensible materials akin to low-density polyethylene, all at the flick of a switch. Leveraging this finding in lithographic grayscale 3D printing enables rapid multimaterial fabrication with voxel-level control over optical and mechanical properties, opening avenues in information storage, soft robotics, and energy damping.

Liquid crystal elastomers (LCEs) are anisotropic polymeric smart materials with promise for soft robotic actuators, dampers, and adhesives. Realizing these applications requires precise 3D alignment of the polymer backbone, yet current manufacturing approaches struggle to produce complex spatial patterns in three dimensions. Here, we introduce a digital light processing (DLP) 3D printing strategy that enables voxel-level control of LCE alignment domains. By integrating a rotatable magnetic array with DLP photomasking, we achieve spatially tunable structures at voxel resolutions. This approach allows freeform 180 degrees alignment within individual voxels, generating highly nonlinear shape transformations in both 2D films and 3D architectures. Finite element modeling and inverse design guide the creation of multidomain alignment patterns that could achieve targeted nonlinear deformation behaviors. As a demonstration, multidomain LCE smart valves are fabricated that exhibit up to 70% improved flow control compared to monodomain analogues. This technique dramatically expands the design space of 3D programmable matter and establishes a pathway toward complex, application-ready LCE systems.

Additive manufacturing (AM) processes, like 3D printing, help to facilitate complex and customizable battery geometries which can provide design freedom and enhance volumetric energy density within electronic devices. AM materials must have the thermal and mechanical properties that enable printability, and when used in batteries, AM materials must also be chemically and electrochemically compatible with the battery chemistry. The compatibility between AM materials and the battery is of particular importance for the cell packaging materials which must be inert and are often overlooked. This study systematically studies AM-compatible polymeric materials for use as gaskets in lithium-ion cells. The materials investigated include three thermoplastics suitable for material extrusion printing: polylactic acid (PLA), polycarbonate, and polypropylene/polyethylene copolymer (PPPEC); and two photoresins suitable for vat photopolymerization (VPP) printing: an acrylate-based photoresin and a polyethylene glycol diacrylate photoresin. The AM gasket materials were tested in comparison to a conventional commercial polypropylene gasket. Mechanical testing (swell measurements and material stiffness) and electrochemical testing (linear sweep voltammetry and galvanostatic cycling of full cells) demonstrated that PLA and the VPP polymers were the least compatible with the lithium-ion battery chemistry, despite their prevalent use in studies of AM batteries, and that PPPEC was the most compatible.

ULTEM 9085, a polyetherimide (PEI), is the first 3D printed thermoplastic material to be qualified by the Federal Aviation Administration (FAA) for use as a high-performance aviation component due to its high strength-to-weight ratio, chemical resistance, and flame retardance properties. However, 3D printed ULTEM suffers reduction in mechanical properties compared to the bulk material. Understanding the role of 3D printing parameters on resulting mechanical performance requires lengthy qualification studies which take many years. This study investigated a combined experimental and machine learning (ML) approach for correlating density, surface profile, and mechanical properties using tensile testing and microscopy across various coupon geometries and build orientations. An ML model predicted the relationship between 3D printing parameters and resulting mechanical properties in a broad design space within 1%. This study provides a straightforward framework for rapid, ML-driven evaluation of 3D printed designs, significantly reducing experimental burden needed high-performance component qualification.

Vat-based photopolymerization (VP) 3D printing processes, such as Digital Light Process (DLP), enable rapid fabrication of geometrically complex parts with tunable mechanical properties. However, measuring polymerisation and part quality in real-time is challenging due to complex platform configurations and opaque resin systems. This study introduces an in-situ ultrasonic testing (UT) system that tracks ultrasonic wave propagation through each printed layer to create a digital twin of the part with 0.2 mm geometric accuracy. Simultaneously, ultrasonic waves can be used to estimate material properties in real-time such as Young’s modulus, detecting changes from 65 MPa to 1.4 GPa, depending on the resin system used. To validate these measurements, ex-situ tensile tests were performed, and the resulting Young’s modulus values confirmed the UT estimates with a maximum mean absolute percentage error of 8.97%. Finally, the UT system could detect internal defects as small as 0.249mm2 during printing. This work uses a UT-based system to monitor VP 3D printing in real-time, providing geometric and material property insights to improve reliability for high-precision applications such as biomedical devices or microfluidic components.

Additive manufacturing (AM) enables the fabrication of complex geometries, yet its application to thermosets remains limited by post-processing requirements. Frontal ring-opening metathesis polymerization (FROMP) offers a promising alternative, enabling energy-efficient, in situ curing of freestanding thermoset structures. This study presents a real-time process monitoring and automated control system for direct ink writing (DIW) of FROMP thermosets. By integrating thermochromic leuco dyes and computer vision, we enable real-time polymerization front tracking, allowing autonomous printing parameter adjustments for consistent geometries across resin formulations. The system’s accuracy was validated against manual tracking, demonstrating precise front velocity detection. Its adaptability was confirmed by printing freestanding mechanical springs with different resins, achieving consistent geometries and mechanical properties despite front velocity variations. These findings highlight the potential of automated DIW control for scalable, repeatable, and material-agnostic 3D printing of thermosets.

Architected LCE lattices are fabricated with flow-induced alignment via direct ink writing and systematically characterized their shape morphing, stiffness, and energy absorption behavior across strain rates spanning six orders of magnitude from 10-3 to 103 s-1. It is shown that architected liquid crystal elastomer (LCE) lattices exhibit superior energy absorption compared to their non-mesogenic (silicone) counterparts. Importantly, the LCE-to-silicone energy absorption ratios are up to 18-fold higher at the highest strain rate tested. A finite element model that captures their shape-morphing response is developed, which exhibits excellent agreement with the experimental observations. The work opens new avenues for designing and fabricating LCE lattices with programmable alignment, shape morphing, and mechanics.

Digital Light Processing (DLP) 3D printing is a widely used vat photopolymerization (VPP) technique known for its ability to produce intricate geometries with a broad range of mechanical properties. However, ensuring consistent, defect-free parts remains challenging, primarily due to shrinkage, partial curing, and uneven exposure. In this work, an in-situ ultrasonic monitoring system is introduced which continually assesses each newly formed layer in real-time. By analyzing how ultrasonic waves travel through each layer, internal flaws down to 0.249 mm² were reliably identified, and dimensional accuracy of approximately 0.2 mm was achieved. Identifying defects precisely when they emerge allows immediate corrective actions, significantly reducing material waste and post-processing efforts. This approach provides a promising solution for enhanced reliability and efficiency of DLP manufacturing, particularly in fields such as biomedical engineering and microfluidics, where precision and material integrity are critical.

The exponential growth of the human population has led to a global housing crisis. To solve this problem, additive manufacturing (AM), also known as 3D printing, has become widely used for on-demand infrastructure construction. While 3D printing offers faster build times and greater design flexibility, it is limited by slow-setting concrete, interruptions to install supports, and the massive environmental impact of cement, which accounts for around 8

Designing structures that effectively dissipate energy across a broad range of mechanical loading rates, including those from compression, shock, and vibration, poses a significant engineering challenge. In this study, liquid-crystal elastomers (LCEs), which possess anisotropic properties due to the alignment of their polymer backbone, are explored. As a result, LCEs exhibit a soft elastic response under mechanical loading, making them ideal for energy dissipation. Advances in additive manufacturing (AM) enable simple fabrication of foamlike dissipative structures with complex lattice geometries. Herein, direct ink write 3D printing, an extrusion AM method, is used to fabricate aligned, monodomain LCE lattice structures for broad strain-rate mechanical damping. In this work, it is shown that these structures can dissipate strain energy in quasi-static environments, comparable to traditional elastomeric lattices, and provide improved damping under high strain-rate drop testing due to LCE soft elasticity. Additionally, under dynamic mechanical vibration, monodomain LCE lattices enhance damping at structural natural frequencies and provide high-frequency attenuation. In these findings, a promising method is presented for fabricating mechanical damping devices that effectively dissipate energy across a broad range of loading rates.

Liquid crystalline elastomers (LCEs) are anisotropic soft materials capable of large dimensional changes when subjected to a stimulus. The magnitude and directionality of the stimuli-induced thermomechanical response is associated with the alignment of the LCE. Recent reports detail the preparation of LCEs by additive manufacturing (AM) techniques, predominately using direct ink write printing. Another AM technique, digital light process (DLP) 3D printing, has generated significant interest as it affords LCE free-forms with high fidelity and resolution. However, one challenge of printing LCEs using vat polymerization methods such as DLP is enforcing alignment. Here, we document the preparation of aligned, main-chain LCEs via DLP 3D printing using a 100 mT magnetic field. Systematic examination isolates the contribution of magnetic field strength, alignment time, and build layer thickness on the degree of orientation in 3D printed LCEs. Informed by this fundamental understanding, DLP is used to print complex LCE free-forms with through-thickness variation in both spatial orientations. The hierarchical variation in spatial orientation within LCE free-forms is used to produce objects that exhibit mechanical instabilities upon heating. DLP printing of aligned LCEs opens new opportunities to fabricate stimuli-responsive materials in form factors optimized for functional use in soft robotics and energy absorption.

Fiber reinforced polymer composites (FRPCs) are valued for their high strength and lightweight and have found applications from aerospace to renewable energy. Additive manufacturing (AM, or 3D printing) of FRPCs is of great interest in recent years due to the manufacturing flexibility offered by AM. Direct ink write (DIW) 3D printing is a popular choice for FRPC 3D printing due to its low cost and open framework for broad material choices. However, previously explored techniques for AM of FRPCs are hindered by heavy reliance on fiber orientation dictated by extrusion toolpath and often require specialized hardware to print FRPCs with low viscosity and long cure time thermoset epoxy matrices. In this paper, we introduce a single-stream hybrid DIW AM technique for the creation of mechanically robust FRPC functional structures where the matrix is DIW 3D printed, and pre-epoxy impregnated (prepreg) woven carbon fiber (CF) fabrics are robotically placed. Additionally, functional components, such as conductive elements, can be readily integrated. We investigated the impact of prepreg woven CF reinforcement on a two-stage photo-thermal thermoset resin matrix on the mechanical characteristics of manufactured FRPCs. We also combined these efforts to fabricate functional structures, including strain sensors for in-situ deformation monitoring, heating elements, and embedded light sensors. This study found that the proposed single-stream hybrid DIW AM process could be a facile approach to fabricate high-strength FRPC functional structures.

Material extrusion additive manufacturing (AM) has enabled an elegant fabrication pathway for a vast material library. Nonetheless, each material requires optimization of printing parameters generally determined through significant trial-and-error testing. To eliminate arduous, iteration-based optimization approaches, many researchers have used machine learning (ML) algorithms which provide opportunities for automated process optimization. In this work, we demonstrate the use of an ML-driven approach for real-time material extrusion print-parameter optimization through in-situ monitoring of printed line geometry. To do this, we use deep invertible neural networks (INNs) which can solve both forward and inverse, or optimization, problems using a single network. By combining in-situ computer vision and deep INNs, the printing parameters can be autonomously optimized to print a target line width in 1.2 s. Furthermore, defects that occur during printing can be rapidly identified and corrected autonomously. The methods developed and presented in this work eliminate user-intensive, time-consuming, and iterative parameter discovery approaches that currently limit accelerated implementation of extrusion-based AM processes. Furthermore, the presented approach can be generalized to provide real-time monitoring and optimization pathways for increasingly complex AM environments.

Liquid crystal elastomers (LCEs) are a class of active materials that can generate rapid, reversible mechanical actuation in response to external stimuli. Fabrication methods for LCEs have remained a topic of intense research interest in recent years. One promising approach, termed 4D printing, combines the advantages of 3D printing with responsive materials, such as LCEs, to generate smart structures that not only possess user-defined static shapes but also can change their shape over time. To date, 4D-printed LCE structures have been limited to flat objects, restricting shape complexity and associated actuation for smart structure applications. In this work, we report the development of embedded 4D printing to extrude hydrophobic LCE ink into an aqueous, thixotropic gel matrix to produce free-standing, free-form 3D architectures without sacrificing the mechanical actuation properties. The ability to 4D print complex, free-standing 3D LCE architectures opens new avenues for the design and development of functional and responsive systems, such as reconfigurable metamaterials, soft robotics, or biomedical devices.

Material extrusion 3D printing has enabled an elegant fabrication pathway for a vast material library. Nonetheless, each material requires optimization of printing parameters generally determined through significant trial-and-error testing. To eliminate arduous, iteration-based optimization approaches, many researchers have used machine learning (ML) algorithms which provide opportunities for automated process optimization. In this work, we demonstrate the use of an ML-driven approach for real-time material extrusion print-parameter optimization through in-situ monitoring of printed line geometry. To do this, we use deep invertible neural networks (INNs) which can solve both forward and inverse, or optimization, problems using a single network. By combining in-situ computer vision and deep INNs, the printing parameters can be autonomously optimized to print a target line width in a matter of seconds. Furthermore, defects that occur during printing can be rapidly identified and corrected autonomously. The methods developed and presented in this paper eliminate time-consuming, iterative parameter discovery approaches that currently limit accelerated implementation of extrusion-based additive manufacturing processes.