Dynamic ductile fracture in metallic structures involves large deformation, strain localization, crack initiation, and progressive material separation, making mesh-objective simulation difficult. In explicit finite element analyses, local damage models combined with element deletion may produce mesh-dependent crack paths and energy dissipation. This study develops an explicit nonlocal Lemaitre damage framework in which integral-type Gaussian averaging is coupled with an objective large-deformation Kirchhoff stress update. The coupled nonlocal constitutive problem is integrated without using a preceding-step proportional-factor approximation. Final material separation is controlled by a local equivalent-plastic-strain threshold numerically derived from reference local damage simulations rather than by the spatially averaged damage variable, thereby separating constitutive regularization from element removal and avoiding discontinuous failed regions and artificial residual ligaments. The underlying elastoplastic damage formulation is validated against tensile tests of DC04 steel up to fracture initiation, showing good agreement with the measured force–displacement response and necking-zone thickness reduction. Tensile mesh-sensitivity analyses produce smoother damage fields and more consistent post-peak responses than the local model. For the single-edge-notched plate, the mesh-induced variation in work to failure decreases from 40.38% to 10.14%, while the variations in peak force, peak displacement, and failure displacement decrease to 0.54%, 2.96%, and 2.30%, respectively. Perforated-plate simulations further demonstrate ligament-controlled crack coalescence and loading-direction-dependent fracture patterns. Energy histories remain bounded during damage evolution and element deletion, with negligible artificial energy and no mass scaling. The proposed framework provides a practical and less mesh-sensitive approach for explicit dynamic ductile-fracture simulation.

This review presents the evolution of strain sensor technologies from traditional bonded wire models to next-generation flexible designs enabled by advanced materials and additive manufacturing through 3D printing. It highlights breakthroughs in fabrication techniques particularly fused filament fabrication (FFF), direct ink writing (DIW), and vat photopolymerization (VPP) that address the limitations of conventional approaches, including complex multi-step processing and alignment challenges. The incorporation of novel materials such as conductive polymers and hybrid composites is shown to significantly enhance key performance parameters like sensitivity, mechanical durability, and strain sensing range. The convergence of cutting-edge manufacturing, material science, and computational modeling signals a paradigm shift in strain sensing, with broad implications for emerging applications in aerospace, biomedicine, soft robotics, and beyond. This work underscores the importance of continued interdisciplinary collaboration to fully realize the potential of flexible strain sensor technologies in adaptive, high-performance systems.

This study presents an innovative machine learning (ML) approach that can make rapid property predictions for AZ31 alloy based on temperature and deformation conditions in the defined range. Additionally, the approach can inversely suggest a set of processing parameters to achieve target properties. The training dataset is generated using a recently developed field fluctuations visco-plastic self-consistent (FF-VPSC) model capable of simulating anisotropic deformation in the dynamic recrystallization regime. The FF-VPSC model incorporates a temperature-sensitive dislocation density-based hardening law, an advanced composite grain model for handling primary and secondary twinning, a grain fragmentation model, and a recrystallization model. The model was validated to capture the evolution of thermo-mechanical response, including stress-strain relationship, texture, twin volume fraction, extent of recrystallization, grain size, and relative activities of slip and twinning modes, given an initial texture and an imposed deformation from room temperature to 200 °C. An artificial neural network (ANN) model is first shown to accurately reproduce the stress-strain curves and π-plane polycrystal yield surface (PCYS) projections for the alloy based on given processing parameters, including temperature, pre-strain levels, and pre-strain paths. A genetic algorithm (GA)-based inverse design methodology is then developed to find optimized processing parameters based on predefined mechanical responses. A detailed discussion is provided on the conditions leading to successful inverse design and the factors that may cause failure.

3D-printed catalysts with open-celled structures can significantly reduce mass and heat transfer resistance compared with traditional pellet counterparts. However, current 3D printing techniques typically achieve only about 70 wt% zeolite loading. While higher loading is desirable to improve catalyst performance, it also increases slurry viscosity, leading to printing challenges such as nozzle clogging and extrusion difficulties. In this work, we present a new direct ink writing (DIW) slurry formulation that increases the zeolite loading limit to 90.5 wt% while maintaining good printability. We found that incorporating a nonionic dispersant, such as hydroxypropyl methylcellulose (HPMC), is more effective in preventing particle aggregation and sedimentation than using an anionic dispersant like carboxymethyl cellulose (CMC). Optimized preprocessing parameters (slurry aging, printing speed, extrusion rate, and nozzle size) along with post-processing conditions (calcination temperature) are reported to yield high-quality 3D printed zeolite catalyst monoliths. Adequate calcination at 600 °C results in the highest accessible specific surface area (SSA) and excellent structural integrity, sustaining compressive strength up to 1.77 MPa. The mechanical strength is over 200% higher than that of existing 3D-printed zeolite catalysts.

Biobased composites, which consist of natural fibers and biobased polymer binders, are gaining traction due to their renewability, low carbon footprint, lightweight nature, multifunctionality, and potential recycling capabilities. Despite their promise, these materials face challenges such as moisture sensitivity, thermal degradation, and limited durability, often due to weak fiber-matrix interfaces. Addressing these challenges and advancing their development requires a comprehensive understanding of material constituents, interfacial behavior, processing techniques, and lifecycle performance. However, existing reviews remain scattered, typically focusing on a single aspect, such as material development and fabrication methods. In this paper, we provide an integrated overview of the biobased composites across the full lifecycle, from raw material selection and interface treatments to scalable manufacturing, recycling, and end-use applications. We summarize various surface treatment methods used to enhance the mechanical properties, durability, and functionality of biocomposites, and systematically compare their performance, cost, and biodegradability. Scalable production techniques that affect the structure and properties of biocomposite products are compared. End-of-life management routes, including mechanical, chemical, and thermal recycling, are evaluated with respect to cost and efficiency. The industrial applications and future research directions are also explored to promote the wider adoption of biobased composites across key sectors such as automotive, aerospace, and construction.

Metamaterials offer transformative potential across various engineering domains. Despite significant progress in building databases linking metamaterial architectures to their properties, conventional machine learning (ML) approaches still face notable limitations. They cannot be directly applied to predict the same properties when the geometry or size scale of metamaterials changes, nor can they be readily extended to predict new properties. This is because a substantial amount of new training data is typically required to retrain the model, which is essentially equivalent to restarting the learning process from scratch. We present a transfer learning-based framework that significantly reduces the amount of required training data while providing high accuracy and stability. This framework employs a size-independent Convolutional Neural Network (CNN) architecture for both the source and target models through the implementation of 1 x 1 convolutional transformation and global average pooling (GAP). The source model is well-trained to predict the Young's modulus of any 8 x 8 lattice metamaterial composed of four given unit cells. The target model can predict the Young's modulus even when new unit cells are introduced or when the size scale of the metamaterial changes. Moreover, the learned structure-Young's modulus knowledge can be successfully transferred to predict other mechanical or nonmechanical properties. Compared to conventional ML approaches, this framework requires only 1% of training data in the studied scenarios and no modifications to the model architecture. This work provides new avenues for building scalable and data-efficient metamaterial design spaces for various applications.

Thermoelectric generators (TEGs) are widely recognized as clean energy solutions to convert low-grade waste heat into electricity. However, low output power has limited their practical applications. In this paper, we present an innovative thermoelectric system that can improve the output power by up to 130 % compared to the existing design. This system incorporates an advanced metastructure heat sink and a turbulator within the cooling system. An experimentally validated Computational Fluid Dynamics (CFD) - Finite Element Method (FEM) model is developed to predict the system's output voltage and power for various metastructure heat sink and turbulator configurations. Unlike existing models that assume constant temperature distribution at the cold side, our model can make more realistic temperature predictions by accounting for the effects of water flow, geometric design of heat sink and turbulator on the convective heat transfer and thermoelectric conversion. This study reveals that optimizing the geometric design of the heat sink and turbulator is an effective strategy for enhancing output power. Our thermoelectric system performs more effectively in high-temperature environments as increasing the temperature by 40 degrees C can lead to an additional 2.1-fold enhancement in the output power, and a high power density of 33.13 mW/cm2 compared to the commercial TEGs.

Thermoelectric materials, which convert thermal gradients into electricity, offer an energy conversion technology with zero emissions, rapid response, and high reliability. Catalysts, on the other hand, can accelerate chemical reactions by reducing activation energy without consuming themselves. These two technologies exhibit strong synergy, particularly in systems where waste heat can be harvested to enhance catalytic processes. This work presents a detailed discussion of the concept of thermoelectrocatalysis (TECatal), including the current state of the art, existing challenges and potential future solutions. Its potential applications in hydrogen generation, CO2 conversion, environmental disinfection, and cancer therapy are discussed in detail. This work also offers a comprehensive perspective on interdisciplinary research on TECatal in materials design, system-level integration, and performance optimization.

This work explores the implications of mounting and fixturing of piezoelectric resonators (PRs) for applications in inductorless DC-DC power conversion. From a scalability and packaging perspective, various mounting strategies are evaluated, and impacts of solder bonding of PRs on performance and losses are quantified. COMSOL simulations are performed to verify shifts in PR resonance. Insights from fixturing are applied to assess the impact of mass augmentation strategies on mechanical losses and understand the feasibility of achieving higher power densities. Fixtured and augmented PRs are subsequently evaluated in a DC-DC converter to test the influence on converter efficiency.

Piezoelectric sensors are widely employed in industries for both acceleration and dynamic force sensing. However, achieving both functionalities typically requires multiple sensors, increasing complexity and complicating signal processing. To address this limitation, we present a multifunctional compressible piezocomposite sensor (CPS) that combines low-frequency acceleration sensing and wide-range dynamic force sensing in a single device. Dual functionality is enabled through a unique geometry that transits between uncompressed and compressed states. Two CPS prototypes are designed and fabricated using digital light processing (DLP) 3D printing with a customized slurry resin. In acceleration sensing mode, the CPS achieves ultra-high sensitivity of up to 455.9 mV/ g with a resonant frequency of 10-23 Hz. This makes it well-suited for monitoring human activities, infrastructure vibrations, and robotic movements. For dynamic force sensing, it exhibits 11.53 V/N (0.60 V/mJ) sensitivity and a force range exceeding 3.5 kN. Also, the CPS withstands extreme mechanical deformation and endures elongations up to 656 %, indicating its potential for large-strain sensing applications, such as extensive joint movements monitoring. Practical use of the CPS is demonstrated by integrating it into a robotic arm to monitor motion smoothness and detect transient clamping forces. This multifunctional sensor opens new avenues for compact and streamlined engineering systems.

Piezoelectric materials are extensively used across sensing and energy harvesting applications due to their ability to convert mechanical energy into electrical energy. Piezoelectric composites, typically consisting of a polymer matrix with ceramic reinforcements, can combine the advantages of each constituent. However, maintaining stable mechanical performance in high-temperature environments remains a significant challenge as conventional polymer matrices usually experience structural degradation. In this work, we developed a novel piezoelectric composite using a preceramic polymer (PCP) matrix with barium titanate (BTO) inclusions. PCPs are silicone-based polymers that undergo a unique polymer-to-ceramic phase transition upon heating. The strong and stable Si-O bond facilitates enhanced fracture resilience and improve structural integrity. Our analysis indicates that our piezo-composites can withstand compressive stress up to 30 MPa and strain up to 20 % without structural failure, even at 500 degrees C. Furthermore, they exhibit significantly improved d33 piezoelectric coefficients after repeated thermal cycles. It is found that the d33 piezoelectric coefficient can reach 6.98 pC/N after three thermal cycles, which is a 2.36-fold increase compared to samples subjected to a single thermal cycle. Digital Light Processing (DLP), a typical 3D printing technique, is utilized to fabricate samples with customized geometries and properties. This innovative piezo-composite opens new possibilities for sensing, energy harvesting, and actuation in high temperature environments.

Thermoelectric generators (TEGs) are widely recognized as clean energy solutions that can convert low-grade waste heat into electricity through a temperature gradient. Despite their significant potential, challenges such as low conversion efficiency and high costs have limited their practical applications. In this paper, we present an innovative metamaterial design concept for TEGs with significantly improved efficiency. A Finite Element Model is validated using Bi0.5Sb1.5Te3 bulk samples fabricated via the drop-cast method. This model can predict opencircuit voltage and output power as a function of an arbitrary metamaterial design using the commercial software ANSYS (R). Four different metastructure designs, including 2D Triangular Honeycomb, Re-entrant, body-centered cubic (BCC), and triply periodic minimal surface (TPMS) structures, are systematically investigated. Through experiments and numerical analysis, the effects of annealing temperature, porosity, and unit cell numbers (UCNs) on the performance of TE legs are explored. It is found that 2D Triangular Honeycomb and BCC structures outperform other configurations due to their capacity to maintain a higher thermal gradient. Optimizing their porosity and UCNs can further enhance the output power. Compared to the traditional designs with bulk TE legs, implementing a 2D metastructure design with 30 % porosity and UCNs of 4 x 4 x 4 can lead to approximately a 100 % increase in power output.

Piezoelectric energy harvesters (PEHs) have drawn considerable attention due to their unique ability to convert ambient mechanical energy to electrical energy. These devices are widely implemented in numerous applications such as wearable technology, structural health monitoring, and renewable energy systems. In this work, a novel approach that seamlessly integrates arbitrary metastructures into the substrate layer of cantilever beam-based PEHs is presented. The corresponding performance of each PEH design is evaluated via an experimentally validated finite element model. This is the first systematic study to explore 3D metastructures with various unit cell configurations, unit cell numbers, and porosity levels. Compared with the existing PEH designs, implementation of a 3D auxetic metastructure with 85% porosity single unit cell design can demonstrate a substantial enhancement in output power, reaching 48.16 mW, and a high normalized power density (NPD) of 2.1131 mu W mm-3 g-2 Hz-1. Results show that there are competing requirements for improving the performance of PEHs. On one hand, low metastructure stiffness is preferred to achieve high power output at low resonant frequency. On the other hand, metastructures designs with low stiffness may induce excessive distortion in the substrate layer, leading to mechanical energy loss. This deformation mechanism adversely affects the mechanical to electrical energy conversion efficiency. Detailed guidelines for designing and manufacturing high-performance PEHs are discussed in this work. Cantilever beam-based piezoelectric energy harvesters (PEHs) can convert vibrational mechanical energy to electrical energy. This study demonstrates a substantial enhancement in the PEH performance up to 13.26-fold by replacing the conventional solid substrate with a 3D auxetic unit-cell metastructure. Computational analysis suggests that enhancing the performance of an unimorph PEH requires selecting a metastructure substrate with low stiffness. However, it's crucial to prevent mechanical loss caused by internal metastructure deformation, which impedes the effective transfer of deformation to the piezoelectric patch. image

Metamaterials with a zero Poisson’s ratio offer significant advantages in robotic actuation and space exploration due to their precise control of deformation. However, existing machine learning techniques cannot be directly used to accelerate the design of such materials due to the scarcity of this property. We propose a few-shot learning-based framework to generate non-periodic metamaterials with zero Poisson’s ratio. Our framework incorporates an out-of-distribution (OOD) target-oriented sampler into a conditional variational autoencoder (cVAE). Unlike other metamaterial generative models that only deal with continuous pixel data, our approach can handle discrete unit cell patterns by computing their probability distributions. We found that controlling the learning focus during the training process can effectively mitigate the scarcity of acceptable data within the training set. This mitigation is achieved by repeatedly selecting target samples through the OOD target-oriented sampler. Incorporating active learning into the training process can further enhance model efficiency by adaptively adjusting the ratio between acceptable and unacceptable samples. The impacts of training data size, effective data composition, and the number of iterations in active learning on design efficiency are discussed in detail. Compared to random trial-and-error generation, our model demonstrates a substantial increase in the acceptable rate, from 0.3 % to 39 %.

Crack-based strain sensors (CBS), which are inspired by a spider's slit organ, can provide highly sensitive measurement with great flexibility. Fracture pattern design holds the key to meeting different sensing needs. In this article, a computational model is developed to understand the role of fracture patterns on sensitivity and sensing range of CBS that consist of a platinum (Pt) conductive layer and a polydimethylsiloxane (PDMS) substrate layer. Through the coupled mechanical-electrical finite element analysis, we find that a single mode I through crack can yield better sensing performance than a nonthrough crack in other orientations or a few discrete nonthrough cracks in the same orientation. Creating multiple mode I through cracks has a negligible effect on sensitivity. However, increasing the number of cracks can lead to a higher sensing range. When the same number of cracks is employed, even crack spacing can yield the highest sensing range. Sensitivity can be effectively improved by increasing the crack depth. Conclusions from the computational analysis can provide useful feedback for design and manufacturing of CBS in different applications.

The catalyst industry generates approximately $20 billion every year globally and plays a major role in the energy production sector. Traditional industrial catalysts (e.g., pellets) typically have mass and heat transfer limitations, and cause a pressure drop in continuous-flow reactors, lowering the efficiency of the catalytic processes. 3D printing technology has evolved rapidly over the past decade, and the 3D printing of catalysts with desired geometries can address many of the above-stated challenges with traditional catalysts. However, challenges remain to be addressed and opportunities remain to be explored before the full potential in the design and 3D printing of novel catalysts can be realized. This article reviews the recent development in the 3D printing of catalysts. It summarizes the 3D printing design, dimension, property, and performance of 3D printed catalysts. The applications of 3D printed catalysts, such as reforming, wastewater treatment, and CO2 capture and removal, are discussed, with a techno-economic analysis and life cycle analysis. Future research directions and opportunities for 3D printing of catalysts are also highlighted.

Progressive microforming is widely recognized as one of the most efficient and desirable methods of mass production in micromanufacturing arena. To predict the deformation behaviour and size effects of materials induced in progressive microforming, the finite element method (FEM) employed for modelling the microforming process needs to account for microstructure details and deformation/failure mechanisms of the materials. This led to the development of the novel crystal plasticity finite element method (CPFEM) and cohesive zone model (CZM). Previous research and application of CPFEM have been mainly limited to simple deformations such as uniaxial tension and compression, whereas the new method can provide physical insights into how the grain size affects the interplay between crystallographic slip and mechanical twinning in complex microforming, and further material deformation during sheet blanking. A case study was conducted to manufacture a hexagonal socket part using a three-step progressive microforming system, with the comparison between experiments and CPFEM simulations focusing on microstructure evolution, deformation load, and product quality. The CPFEM was found to be more reliable than the conventional FEM in predicting complex deformation, particularly in microstructure and texture evolution, dimensional accuracy and irregular geometries. Results showed that the total height of the part increases with the decreasing grain size, while the head diameter rises with grain size. Simulations successfully anticipated the distributions of dead metal zones and shear bands and identified hole and rollover geometries and corresponding grain size effects. In conclusion, this research facilitates the understanding of grain size effects on the deformation behaviour in progressive microforming and presents a novel approach and strategy for modelling, prediction, and product quality assurance in complex micro deformation and forming processes.

3D printing has the promising capability to fabricate engineered lattice structures with broadly tunable surface area and optimal geometries for maximizing structural and functional properties. This study characterizes the electrical conductivity of 3D lattices of varying size, structure, and porosity to guide additively manufactured electrode design in energy storage devices. Graph theory-based calculations and experiments comparing the conductivity of multiple strut lattice structures and illustrating the scaling laws governing architectures with either coated or solid conductive struts are presented. The lightweight lattices explored here show higher conductivity than random foams that lack a periodic mesostructure. It is experimentally demonstrated that the 3D lattice type influences the specific capacity when employed in supercapacitors, outperforming 2D supercapacitor counterparts, and other 3D printed electrodes while allowing for optimization of the design for different energy storage applications. Additionally, it is shown that tuning the physical structure of the lattices allows for precise control over the electrical response to mechanical loading, as confirmed through experimental measurements. The lattice structure programs the electrodes' mechanical stiffness, with higher relative density samples showing higher Young's modulus. These results can serve to guide the design of 3D printed electrodes in a variety of electrochemical and electromechanical device applications.