Bayesian Machine Learning in Metamaterial Design: Fragile Becomes Supercompressible

Inverse design of skull osteoinductive implants with multi-level pore structures through machine learning

How to accurately design a personalized matching implant that can induce skull regeneration is the focus of current research. However, the design space for the porous structure of implants is extensive, and the mapping relationships between these structures and their mechanical and osteogenic properties are complex. At present, the forward design of skull implants mainly relies on expert experience, leading to high cost and a lengthy process, while the existing inverse design approaches face challenges due to data dependence and manufacturing process errors. This study presents an efficient inverse design method for personalized multilevel structures of skull implants using a machine learning pipeline composed of a finite element method, topological optimization, and neural networks. Based on the mechanical response of the human body falls, this method can tailor multi-level structures for implants in various defect positions. The results show that the proposed method establishes a bidirectional relationship between topological parameters and mechanical properties, enabling the customization of mechanical behavior at low computational cost while accounting for manufacturing errors in the 3D printing process. Additionally, the design results are also mutually consistent with analytical relationships between lattice parameters and the elastic modulus obtained from experiments and finite element simulations. Thus, this study provides a general and practical approach to rapidly design skull osteoinductive implants.

Blueprints of Architected Materials: A Guide to Metamaterial Design for Tissue Engineering

Mechanical metamaterials are rationally designed structures engineered to exhibit extraordinary properties, often surpassing those of their constituent materials. The geometry of metamaterials' building blocks, referred to as unit cells, plays an essential role in determining their macroscopic mechanical behavior. Due to their hierarchical design and remarkable properties, metamaterials hold significant potential for tissue engineering; however their implementation in the field remains limited. The major challenge hindering the broader use of metamaterials lies in the complexity of unit cell design and fabrication. To address this gap, a comprehensive guide is presented detailing the design principles of well-established metamaterials. The essential unit cell geometric parameters and design constraints, as well as their influence on mechanical behavior, are summarized highlighting essential points for effective fabrication. Moreover, the potential integration of artificial intelligence techniques is explored in meta-biomaterial design for patient- and application-specific design. Furthermore, a comprehensive overview of current applications of mechanical metamaterials is provided in tissue engineering, categorized by tissue type, thereby showcasing the versatility of different designs in matching the mechanical properties of the target tissue. This review aims to provide a valuable resource for tissue engineering researchers and aid in the broader use of metamaterials in the field.

Automated discovery of reprogrammable nonlinear dynamic metamaterials

Harnessing the rich nonlinear dynamics of highly deformable materials has the potential to unlock the next generation of functional smart materials and devices. However, unlocking such potential requires effective strategies to spatially engineer material architectures within the nonlinear dynamic regime. Here we introduce an inverse-design framework to discover flexible mechanical metamaterials with a target nonlinear dynamic response. The desired dynamic task is encoded via optimal tuning of the full-scale metamaterial geometry through an inverse-design approach powered by a fully differentiable simulation environment. By deploying such a strategy, mechanical metamaterials are tailored for energy focusing, energy splitting, dynamic protection and nonlinear motion conversion. Furthermore, our design framework can be expanded to automatically discover reprogrammable architectures capable of switching between different dynamic tasks. For instance, we encode two strongly competing tasks-energy focusing and dynamic protection-within a single architecture, using static precompression to switch between these behaviours. The discovered designs are physically realized and experimentally tested, demonstrating the robustness of the engineered tasks. Our approach opens an untapped avenue towards designer materials with tailored robotic-like reprogrammable functionalities.

Biomimetic Mechanical Robust Cement-Resin Composites with Machine Learning-Assisted Gradient Hierarchical Structures

Biological materials relying on hierarchically ordered architectures inspire the emergence of advanced composites with mutually exclusive mechanical properties, but the efficient topology optimization and large-scale manufacturing remain challenging. Herein, this work proposes a scalable bottom-up approach to fabricate a novel nacre-like cement-resin composite with gradient brick-and-mortar (BM) structure, and demonstrates a machine learning-assisted method to optimize the gradient structure. The fabricated gradient composite exhibits an extraordinary combination of high flexural strength, toughness, and impact resistance. Particularly, the toughness and impact resistance of such composite attractively surpass the cement counterparts by factors of approximately 700 and 600 times, and even outperform natural rocks, fiber-reinforced cement-based materials and even some alloys. The strengthening and toughening mechanisms are clarified as the regional-matrix densifying and crack-tip shielding effects caused by the gradient BM structure. The developed gradient composite not only endows a promising structural material for protective applications in harsh scenarios, but also paves a new way for biomimetic metamaterials designing.

AI-Based Metamaterial Design

The use of metamaterials in various devices has revolutionized applications in optics, healthcare, acoustics, and power systems. Advancements in these fields demand novel or superior metamaterials that can demonstrate targeted control of electromagnetic, mechanical, and thermal properties of matter. Traditional design systems and methods often require manual manipulations which is time-consuming and resource intensive. The integration of artificial intelligence (AI) in optimizing metamaterial design can be employed to explore variant disciplines and address bottlenecks in design. AI-based metamaterial design can also enable the development of novel metamaterials by optimizing design parameters that cannot be achieved using traditional methods. The application of AI can be leveraged to accelerate the analysis of vast data sets as well as to better utilize limited data sets via generative models. This review covers the transformative impact of AI and AI-based metamaterial design for optics, acoustics, healthcare, and power systems. The current challenges, emerging fields, future directions, and bottlenecks within each domain are discussed.

Advancing materials science through next-generation machine learning

For over a decade, machine learning (ML) models have been making strides in computer vision and natural language processing (NLP), demonstrating high proficiency in specialized tasks. The emergence of large-scale language and generative image models, such as ChatGPT and Stable Diffusion, has significantly broadened the accessibility and application scope of these technologies. Traditional predictive models are typically constrained to mapping input data to numerical values or predefined categories, limiting their usefulness beyond their designated tasks. In contrast, contemporary models employ representation learning and generative modeling, enabling them to extract and encode key insights from a wide variety of data sources and decode them to create novel responses for desired goals. They can interpret queries phrased in natural language to deduce the intended output. In parallel, the application of ML techniques in materials science has advanced considerably, particularly in areas like inverse design, material prediction, and atomic modeling. Despite these advancements, the current models are overly specialized, hindering their potential to supplant established industrial processes. Materials science, therefore, necessitates the creation of a comprehensive, versatile model capable of interpreting human-readable inputs, intuiting a wide range of possible search directions, and delivering precise solutions. To realize such a model, the field must adopt cutting-edge representation, generative, and foundation model techniques tailored to materials science. A pivotal component in this endeavor is the establishment of an extensive, centralized dataset encompassing a broad spectrum of research topics. This dataset could be assembled by crowdsourcing global research contributions and developing models to extract data from existing literature and represent them in a homogenous format. A massive dataset can be used to train a central model that learns the underlying physics of the target areas, which can then be connected to a variety of specialized downstream tasks. Ultimately, the envisioned model would empower users to intuitively pose queries for a wide array of desired outcomes. It would facilitate the search for existing data that closely matches the sought-after solutions and leverage its understanding of physics and material-behavior relationships to innovate new solutions when pre-existing ones fall short.

Harnessing metamaterials for efficient wireless power transfer for implantable medical devices

Wireless power transfer (WPT) within the human body can enable long-lasting medical devices but poses notable challenges, including absorption by biological tissues and weak coupling between the transmitter (Tx) and receiver (Rx). In pursuit of more robust and efficient wireless power, various innovative strategies have emerged to optimize power transfer efficiency (PTE). One such groundbreaking approach stems from the incorporation of metamaterials, which have shown the potential to enhance the capabilities of conventional WPT systems. In this review, we delve into recent studies focusing on WPT systems that leverage metamaterials to achieve increased efficiency for implantable medical devices (IMDs) in the electromagnetic paradigm. Alongside a comparative analysis, we also outline current challenges and envision potential avenues for future advancements.

Deep Learning for Size-Agnostic Inverse Design of Random-Network 3D Printed Mechanical Metamaterials

Practical applications of mechanical metamaterials often involve solving inverse problems aimed at finding microarchitectures that give rise to certain properties. The limited resolution of additive manufacturing techniques often requires solving such inverse problems for specific specimen sizes. Moreover, the candidate microarchitectures should be resistant to fatigue and fracture. Such a multi-objective inverse design problem is formidably difficult to solve but its solution is the key to real-world applications of mechanical metamaterials. Here, a modular approach titled "Deep-DRAM" that combines four decoupled models is proposed, including two deep learning (DL) models, a deep generative model based on conditional variational autoencoders, and direct finite element (FE) simulations. Deep-DRAM integrates these models into a framework capable of finding many solutions to the posed multi-objective inverse design problem based on random-network unit cells. Using an extensive set of simulations as well as experiments performed on 3D printed specimens, it is demonstrate that: 1) the predictions of the DL models are in agreement with FE simulations and experimental observations, 2) an enlarged envelope of achievable elastic properties (e.g., rare combinations of double auxeticity and high stiffness) is realized using the proposed approach, and 3) Deep-DRAM can provide many solutions to the considered multi-objective inverse design problem.

Data-Driven Design for Metamaterials and Multiscale Systems: A Review

Metamaterials are artificial materials designed to exhibit effective material parameters that go beyond those found in nature. Composed of unit cells with rich designability that are assembled into multiscale systems, they hold great promise for realizing next-generation devices with exceptional, often exotic, functionalities. However, the vast design space and intricate structure-property relationships pose significant challenges in their design. A compelling paradigm that could bring the full potential of metamaterials to fruition is emerging: data-driven design. This review provides a holistic overview of this rapidly evolving field, emphasizing the general methodology instead of specific domains and deployment contexts. Existing research is organized into data-driven modules, encompassing data acquisition, machine learning-based unit cell design, and data-driven multiscale optimization. The approaches are further categorized within each module based on shared principles, analyze and compare strengths and applicability, explore connections between different modules, and identify open research questions and opportunities.

Rational Design of Flexible Mechanical Force Sensors for Healthcare and Diagnosis

Over the past decade, there has been a significant surge in interest in flexible mechanical force sensing devices and systems. Tremendous efforts have been devoted to the development of flexible mechanical force sensors for daily healthcare and medical diagnosis, driven by the increasing demand for wearable/portable devices in long-term healthcare and precision medicine. In this review, we summarize recent advances in diverse categories of flexible mechanical force sensors, covering piezoresistive, capacitive, piezoelectric, triboelectric, magnetoelastic, and other force sensors. This review focuses on their working principles, design strategies and applications in healthcare and diagnosis, with an emphasis on the interplay among the sensor architecture, performance, and application scenario. Finally, we provide perspectives on the remaining challenges and opportunities in this field, with particular discussions on problem-driven force sensor designs, as well as developments of novel sensor architectures and intelligent mechanical force sensing systems.

Rapid inverse design of metamaterials based on prescribed mechanical behavior through machine learning

Designing and printing metamaterials with customizable architectures enables the realization of unprecedented mechanical behaviors that transcend those of their constituent materials. These behaviors are recorded in the form of response curves, with stress-strain curves describing their quasi-static footprint. However, existing inverse design approaches are yet matured to capture the full desired behaviors due to challenges stemmed from multiple design objectives, nonlinear behavior, and process-dependent manufacturing errors. Here, we report a rapid inverse design methodology, leveraging generative machine learning and desktop additive manufacturing, which enables the creation of nearly all possible uniaxial compressive stress‒strain curve cases while accounting for process-dependent errors from printing. Results show that mechanical behavior with full tailorability can be achieved with nearly 90% fidelity between target and experimentally measured results. Our approach represents a starting point to inverse design materials that meet prescribed yet complex behaviors and potentially bypasses iterative design-manufacturing cycles.

Buckling Metamaterials for Extreme Vibration Damping

Damping mechanical resonances is a formidable challenge in an increasing number of applications. Many passive damping methods rely on using low stiffness, complex mechanical structures or electrical systems, which render them unfeasible in many of these applications. Herein, a new method for passive vibration damping, by allowing buckling of the primary load path in mechanical metamaterials and lattice structures, is introduced, which sets an upper limit for vibration transmission: the transmitted acceleration saturates at a maximum value in both tension and compression, no matter what the input acceleration is. This nonlinear mechanism leads to an extreme damping coefficient tanδ ≈ 0.23 in a metal metamaterial-orders of magnitude larger than the linear damping coefficient of traditional lightweight structural materials. This principle is demonstrated experimentally and numerically in free-standing rubber and metal mechanical metamaterials over a range of accelerations. It is also shown that damping nonlinearities even allow buckling-based vibration damping to work in tension, and that bidirectional buckling can further improve its performance. Buckling metamaterials pave the way toward extreme vibration damping without mass or stiffness penalty, and, as such, could be applicable in a multitude of high-tech applications, including aerospace, vehicles, and sensitive instruments.

Unleashing the Power of Artificial Intelligence in Materials Design

The integration of artificial intelligence (AI) algorithms in materials design is revolutionizing the field of materials engineering thanks to their power to predict material properties, design de novo materials with enhanced features, and discover new mechanisms beyond intuition. In addition, they can be used to infer complex design principles and identify high-quality candidates more rapidly than trial-and-error experimentation. From this perspective, herein we describe how these tools can enable the acceleration and enrichment of each stage of the discovery cycle of novel materials with optimized properties. We begin by outlining the state-of-the-art AI models in materials design, including machine learning (ML), deep learning, and materials informatics tools. These methodologies enable the extraction of meaningful information from vast amounts of data, enabling researchers to uncover complex correlations and patterns within material properties, structures, and compositions. Next, a comprehensive overview of AI-driven materials design is provided and its potential future prospects are highlighted. By leveraging such AI algorithms, researchers can efficiently search and analyze databases containing a wide range of material properties, enabling the identification of promising candidates for specific applications. This capability has profound implications across various industries, from drug development to energy storage, where materials performance is crucial. Ultimately, AI-based approaches are poised to revolutionize our understanding and design of materials, ushering in a new era of accelerated innovation and advancement.

Artificial Intelligence and Advanced Materials

Artificial intelligence (AI) is gaining strength, and materials science can both contribute to and profit from it. In a simultaneous progress race, new materials, systems, and processes can be devised and optimized thanks to machine learning (ML) techniques, and such progress can be turned into innovative computing platforms. Future materials scientists will profit from understanding how ML can boost the conception of advanced materials. This review covers aspects of computation from the fundamentals to directions taken and repercussions produced by computation to account for the origins, procedures, and applications of AI. ML and its methods are reviewed to provide basic knowledge of its implementation and its potential. The materials and systems used to implement AI with electric charges are finding serious competition from other information-carrying and processing agents. The impact these techniques have on the inception of new advanced materials is so deep that a new paradigm is developing where implicit knowledge is being mined to conceive materials and systems for functions instead of finding applications to found materials. How far this trend can be carried is hard to fathom, as exemplified by the power to discover unheard of materials or physical laws buried in data.

Nanostructured surface plasmon resonance sensors: Toward narrow linewidths

Surface plasmon resonance sensors have found wide applications in optical sensing field due to their excellent sensitivity to the slight refractive index change of surrounding medium. However, the intrinsically high optical losses in metals make it nontrivial to obtain narrow resonance spectra, which greatly limits the performance of surface plasmon resonance sensors. This review first introduces the influence factors of plasmon linewidths of metallic nanostructures. Then, various approaches to achieve narrow resonance linewidths are summarized, including the fabrication of nanostructured surface plasmon resonance sensors supporting surface lattice resonance/plasmonic Fano resonance or coupling with a photonic cavity, the preparation of surface plasmon resonance sensors with ultra-narrow resonators, as well as strategies such as platform-induced modification, alternating different dielectric layers, and the coupling with whispering-gallery-modes. Lastly, the applications and some existing challenges of surface plasmon resonance sensors are discussed. This review aims to provide guidance for the further development of nanostructured surface plasmon resonance sensors.

Utilizing ANN for Predicting the Cauchy Stress and Lateral Stretch of Random Elastomeric Foams under Uniaxial Loading

As a result of their cell structures, elastomeric foams exhibit high compressibility and are frequently used as buffer cushions in energy absorption. Foam pads between two surfaces typically withstand uniaxial loads. In this paper, we considered the effects of porosity and cell size on the mechanical behavior of random elastomeric foams, and proposed a constitutive model based on an artificial neural network (ANN). Uniform cell size distribution was used to represent monodisperse foam. The constitutive relationship between Cauchy stress and the four input variables of axial stretch λ_U, lateral stretch λ_L, porosity φ, and cell size θ was given by con-ANN. The mechanical responses of 500 different foam structures (20% < φ < 60%, 0.1 mm < θ < 0.5 mm) under compression and tension loads (0.4 < λ_U < 3) were simulated, and a dataset containing 100,000 samples was constructed. We also introduced a pre-ANN to predict lateral stretch to address the issue of missing lateral strain data in practical applications. By combining physical experience, we chose appropriate input forms and activation functions to improve ANN's extrapolation capability. The results showed that pre-ANN and con-ANN could provide reasonable predictions for λ_U outside the dataset. We can obtain accurate lateral stretch and axial stress predictions from two ANNs. The porosity affects the stress and λ_L, while the cell size only affects the stress during foam compression.

Bayesian-Optimization-Assisted Laser Reduction of Poly(acrylonitrile) for Electrochemical Applications

Laser reduction of polymers has recently been explored to rapidly and inexpensively synthesize high-quality graphitic and carbonaceous materials. However, in past work, laser-induced graphene has been restricted to semiaromatic polymers and graphene oxide; in particular, poly(acrylonitrile) (PAN) is claimed to be a polymer that cannot be laser-reduced successfully to form electrochemically active material. In this work, three strategies to surmount this barrier are employed: (1) thermal stabilization of PAN to increase its sp² content for improved laser processability, (2) prelaser treatment microstructuring to reduce the effects of thermal stresses, and (3) Bayesian optimization to search the parameter space of laser processing to improve performance and discover morphologies. Based on these approaches, we successfully synthesize laser-reduced PAN with a low sheet resistance (6.5 Ω sq^-1) in a single lasing step. The resulting materials are tested electrochemically, and their applicability as membrane electrodes for vanadium redox flow batteries is demonstrated. This work demonstrates electrodes that are processed in air, below 300 °C, which are cycled stably over 2 weeks at 40 mA cm^-2, motivating further development of laser reduction of porous polymers for membrane electrode applications such as RFBs.

Computational Design and Manufacturing of Sustainable Materials through First-Principles and Materiomics

Engineered materials are ubiquitous throughout society and are critical to the development of modern technology, yet many current material systems are inexorably tied to widespread deterioration of ecological processes. Next-generation material systems can address goals of environmental sustainability by providing alternatives to fossil fuel-based materials and by reducing destructive extraction processes, energy costs, and accumulation of solid waste. However, development of sustainable materials faces several key challenges including investigation, processing, and architecting of new feedstocks that are often relatively mechanically weak, complex, and difficult to characterize or standardize. In this review paper, we outline a framework for examining sustainability in material systems and discuss how recent developments in modeling, machine learning, and other computational tools can aid the discovery of novel sustainable materials. We consider these through the lens of materiomics, an approach that considers material systems holistically by incorporating perspectives of all relevant scales, beginning with first-principles approaches and extending through the macroscale to consider sustainable material design from the bottom-up. We follow with an examination of how computational methods are currently applied to select examples of sustainable material development, with particular emphasis on bioinspired and biobased materials, and conclude with perspectives on opportunities and open challenges.

A machine learning-based probabilistic computational framework for uncertainty quantification of actuation of clustered tensegrity structures

Clustered tensegrity structures integrated with continuous cables are lightweight, foldable, and deployable. Thus, they can be used as flexible manipulators or soft robots. The actuation process of such soft structure has high probabilistic sensitivity. It is essential to quantify the uncertainty of actuated responses of the tensegrity structures and to modulate their deformation accurately. In this work, we propose a comprehensive data-driven computational approach to study the uncertainty quantification (UQ) and probability propagation in clustered tensegrity structures, and we have developed a surrogate optimization model to control the flexible structure deformation. An example of clustered tensegrity beam subjected to a clustered actuation is presented to demonstrate the validity of the approach and its potential application. The three main novelties of the data-driven framework are: (1) The proposed model can avoid the difficulty of convergence in nonlinear Finite Element Analysis (FEA), by two machine learning methods, the Gauss Process Regression (GPR) and Neutral Network (NN). (2) A fast real-time prediction on uncertainty propagation can be achieved by the surrogate model, and (3) Optimization of the actuated deformation comes true by using both Sequence Quadratic Programming (SQP) and Bayesian optimization methods. The results have shown that the proposed data-driven computational approach is powerful and can be extended to other UQ models or alternative optimization objectives.

Synthesis of 2-DOF Decoupled Rotation Stage with FEA-Based Neural Network

Transfer printing technology has developed rapidly in the last decades, offering a potential demand for 2-DOF rotation stages. In order to remove decoupling modeling, improve motion accuracy, and simplify the control method, the 2-DOF decoupled rotation stages based on compliant mechanisms present notable merits. Therefore, a novel 2-DOF decoupled rotation stage is synthesized of which the critical components of decoupling are the topological arrangement and a novel decoupled compound joint. To fully consider the undesired deformation of rigid segments, an FEA-based neural network model is utilized to predict the rotation strokes and corresponding coupling ratios, and optimize the structural parameters. Then, FEA simulations are conducted to investigate the static and dynamic performances of the proposed 2-DOF decoupled rotation stage. The results show larger rotation strokes of 4.302 mrad in one-axis actuation with a 1.697% coupling ratio, and 4.184 and 4.151 mrad in two-axis actuation with undesired Rz rotation of 0.014 mrad with fewer actuators than other works. In addition, the first natural frequency of 2151 Hz is also higher, enabling a wider working frequency range.

Machine Learning of Implicit Combinatorial Rules in Mechanical Metamaterials

Combinatorial problems arising in puzzles, origami, and (meta)material design have rare sets of solutions, which define complex and sharply delineated boundaries in configuration space. These boundaries are difficult to capture with conventional statistical and numerical methods. Here we show that convolutional neural networks can learn to recognize these boundaries for combinatorial mechanical metamaterials, down to finest detail, despite using heavily undersampled training sets, and can successfully generalize. This suggests that the network infers the underlying combinatorial rules from the sparse training set, opening up new possibilities for complex design of (meta)materials.

Machine learning assisted metamaterial-based reconfigurable antenna for low-cost portable electronic devices

Antenna design has evolved from bulkier to small portable designs but there is a need for smarter antenna design using machine learning algorithms that can meet today's high growing demand for smart and fast devices. Here in this research, main focus is on developing smart antenna design using machine learning applicable in 5G mobile applications and portable Wi-Fi, Wi-MAX, and WLAN applications. Our design is based on the metamaterial concept where the patch is truncated and etched with a split ring resonator (SRR). The high gain requirement is met by adding metamaterial superstrates having thin wires (TW) and SRRs. The reconfigurability is achieved by adding three PIN diode switches. Multiple designs have been observed by adding superstrate layers ranging from one layer to four layers with interchanging TWs and SRRs. The TW metamaterial superstrate design with two layers is giving the best performance in gain, bandwidth, and the number of bands. The design is optimized by changing the path's physical parameters. To shrink simulation time, Extra Tree Regression based machine learning model is used to learn the behavior of the antenna and predict the reflectance value for a wide range of frequencies. Experimental results prove that the use of the Extra Tree Regression based model for simulation of antenna design can cut the simulation time, resource requirements by 80%.

Studying Complex Evolution of Hyperelastic Materials under External Field Stimuli using Artificial Neural Networks with Spatiotemporal Features in a Small-Scale Dataset

Deep-learning (DL) methods, in consideration of their excellence in dealing with highly complex structure-performance relationships for materials, are expected to become a new design paradigm for breakthroughs in material performance. However, in most cases, it is impractical to collect massive-scale experimental data or open-source theoretical databases to support training DL models with sufficient prediction accuracy. In a dataset consisting of 483 porous silicone rubber observations generated via ink-writing additive manufacturing, this work demonstrates that constructing low-dimensional, accurate descriptors is the prerequisite for obtaining high-precision DL models based on small experimental datasets. On this basis, a unique convolutional bidirectional long short-term memory model with spatiotemporal features extraction capability is designed, whose hierarchical learning mechanism further reduces the requirement for the amount of data by taking full advantage of data information. The proposed approach can be expected as a powerful tool for innovative material design on small experimental datasets, which can also be used to explore the evolutionary mechanisms of the structures and properties of materials under complex working conditions.

Compliant Cross-Axis Joints: A Tailoring Displacement Range Approach via Lattice Flexures and Machine Learning

Compliant joints are flexible elements that allow displacement due to the elastic deformations they experience under the action of external loading. The flexible parts responsible for these displacements are known as flexure hinges. Displacement, or motion range, in compliant joints depends on the stiffness of the flexure hinges and can be tailored through various parameters, including the overall dimensions, the base material, and the distribution within the hinge. Considering the distribution, we propose the stiffness modification of a compliant cross-axis joint via the use of lattice mechanical metamaterials. Due to the wide range of parameters that influence the stiffness of a lattice, different machine learning algorithms (artificial neural networks, support vector machine, and Gaussian progress regression) were proposed to forecast these parameters. Here, the machine learning algorithm with the best forecasting was the Gaussian progress regression; this algorithm has the advantage of well-tuning even with small regression databases, allowing these functions to more easily adjust to suit specific data, even if the dataset is small. Hexagonal, re-entrant, and square lattices were studied as flexure hinges. For each, the effect of the unit cell size and its orientation with respect to the principal axis on the effective stiffness were studied via computational and laboratory experiments on additively manufactured samples. Finite element predictions resulted in good agreement with the experimentally obtained data. As a result, using lattice-flexure hinges led to increments in displacement ranging from double to ten times those obtained with solid hinges. The most suitable machine learning algorithm was the Gaussian progress regression, with a maximum error of 0.12% when compared to the finite element analysis results.

Enhanced Cellular Materials through Multiscale, Variable-Section Inner Designs: Mechanical Attributes and Neural Network Modeling

In the current work, the mechanical response of multiscale cellular materials with hollow variable-section inner elements is analyzed, combining experimental, numerical and machine learning techniques. At first, the effect of multiscale designs on the macroscale material attributes is quantified as a function of their inner structure. To that scope, analytical, closed-form expressions for the axial and bending inner element-scale stiffness are elaborated. The multiscale metamaterial performance is numerically probed for variable-section, multiscale honeycomb, square and re-entrant star-shaped lattice architectures. It is observed that a substantial normal, bulk and shear specific stiffness increase can be achieved, which differs depending on the upper-scale lattice pattern. Subsequently, extended mechanical datasets are created for the training of machine learning models of the metamaterial performance. Thereupon, neural network (NN) architectures and modeling parameters that can robustly capture the multiscale material response are identified. It is demonstrated that rather low-numerical-cost NN models can assess the complete set of elastic properties with substantial accuracy, providing a direct link between the underlying design parameters and the macroscale metamaterial performance. Moreover, inverse, multi-objective engineering tasks become feasible. It is shown that unified machine-learning-based representation allows for the inverse identification of the inner multiscale structural topology and base material parameters that optimally meet multiple macroscale performance objectives, coupling the NN metamaterial models with genetic algorithm-based optimization schemes.

Spiderweb Nanomechanical Resonators via Bayesian Optimization: Inspired by Nature and Guided by Machine Learning

From ultrasensitive detectors of fundamental forces to quantum networks and sensors, mechanical resonators are enabling next-generation technologies to operate in room-temperature environments. Currently, silicon nitride nanoresonators stand as a leading microchip platform in these advances by allowing for mechanical resonators whose motion is remarkably isolated from ambient thermal noise. However, to date, human intuition has remained the driving force behind design processes. Here, inspired by nature and guided by machine learning, a spiderweb nanomechanical resonator is developed that exhibits vibration modes, which are isolated from ambient thermal environments via a novel "torsional soft-clamping" mechanism discovered by the data-driven optimization algorithm. This bioinspired resonator is then fabricated, experimentally confirming a new paradigm in mechanics with quality factors above 1 billion in room-temperature environments. In contrast to other state-of-the-art resonators, this milestone is achieved with a compact design that does not require sub-micrometer lithographic features or complex phononic bandgaps, making it significantly easier and cheaper to manufacture at large scales. These results demonstrate the ability of machine learning to work in tandem with human intuition to augment creative possibilities and uncover new strategies in computing and nanotechnology.

Bayesian Machine Learning for Efficient Minimization of Defects in ALD Passivation Layers

Atomic layer deposition (ALD) is an enabling technology for encapsulating sensitive materials owing to its high-quality, conformal coating capability. Finding the optimum deposition parameters is vital to achieving defect-free layers; however, the high dimensionality of the parameter space makes a systematic study on the improvement of the protective properties of ALD films challenging. Machine-learning (ML) methods are gaining credibility in materials science applications by efficiently addressing these challenges and outperforming conventional techniques. Accordingly, this study reports the ML-based minimization of defects in an ALD-Al₂O₃ passivation layer for the corrosion protection of metallic copper using Bayesian optimization (BO). In all experiments, BO consistently minimizes the layer defect density by finding the optimum deposition parameters in less than three trials. Electrochemical tests show that the optimized layers have virtually zero film porosity and achieve five orders of magnitude reduction in corrosion current as compared to control samples. Optimized parameters of surface pretreatment using Ar/H₂ plasma, the deposition temperature above 200 °C, and 60 ms pulse time quadruple the corrosion resistance. The significant optimization of ALD layers presented in this study demonstrates the effectiveness of BO and its potential outreach to a broader audience, focusing on different materials and processes in materials science applications.

Metamaterial Reverse Multiple Prediction Method Based on Deep Learning

Metamaterials and their related research have had a profound impact on many fields, including optics, but designing metamaterial structures on demand is still a challenging task. In recent years, deep learning has been widely used to guide the design of metamaterials, and has achieved outstanding performance. In this work, a metamaterial structure reverse multiple prediction method based on semisupervised learning was proposed, named the partially Conditional Generative Adversarial Network (pCGAN). It could reversely predict multiple sets of metamaterial structures that can meet the needs by inputting the required target spectrum. This model could reach a mean average error (MAE) of 0.03 and showed good generality. Compared with the previous metamaterial design methods, this method could realize reverse design and multiple design at the same time, which opens up a new method for the design of new metamaterials.

Strength through defects: A novel Bayesian approach for the optimization of architected materials

We use a previously unexplored Bayesian optimization framework, “evolutionary Monte Carlo sampling,” to systematically design the arrangement of defects in an architected microlattice to maximize its strain energy density before undergoing catastrophic failure. Our algorithm searches a design space with billions of 4 × 4 × 5 3D lattices, yet it finds the global optimum with only 250 cost function evaluations. Our optimum has a normalized strain energy density 12,464 times greater than its commonly studied defect-free counterpart. Traditional optimization is inefficient for this microlattice because (i) the design space has discrete, qualitative parameter states as input variables, (ii) the cost function is computationally expensive, and (iii) the design space is large. Our proposed framework is useful for architected materials and for many optimization problems in science and elucidates how defects can enhance the mechanical performance of architected materials.

Inverted and Programmable Poynting Effects in Metamaterials

The Poynting effect generically manifests itself as the extension of the material in the direction perpendicular to an applied shear deformation (torsion) and is a material parameter hard to design. Unlike isotropic solids, in designed structures, peculiar couplings between shear and normal deformations can be achieved and exploited for practical applications. Here, a metamaterial is engineered that can be programmed to contract or extend under torsion and undergo nonlinear twist under compression. First, it is shown that the system exhibits a novel type of inverted Poynting effect, where axial compression induces a nonlinear torsion. Then the Poynting modulus of the structure is programmed from initial negative values to zero and positive values via a pre-compression applied prior to torsion. The work opens avenues for programming nonlinear elastic moduli of materials and tuning the couplings between shear and normal responses by rational design. Obtaining inverted and programmable Poynting effects in metamaterials inspires diverse applications from designing machine materials, soft robots, and actuators to engineering biological tissues, implants, and prosthetic devices functioning under compression and torsion.

Deciphering controversial results of cell proliferation on TiO2 nanotubes using machine learning

With the rapid development of biomedical sciences, contradictory results on the relationships between biological responses and material properties emerge continuously, adding to the challenge of interpreting the incomprehensible interfacial process. In the present paper, we use cell proliferation on titanium dioxide nanotubes (TNTs) as a case study and apply machine learning methodologies to decipher contradictory results in the literature. The gradient boosting decision tree model demonstrates that cell density has a higher impact on cell proliferation than other obtainable experimental features in most publications. Together with the variation of other essential features, the controversy of cell proliferation trends on various TNTs is understandable. By traversing all combinational experimental features and the corresponding forecast using an exhausted grid search strategy, we find that adjusting cell density and sterilization methods can simultaneously induce opposite cell proliferation trends on various TNTs diameter, which is further validated by experiments. This case study reveals that machine learning is a burgeoning tool in deciphering controversial results in biomedical researches, opening up an avenue to explore the structure-property relationships of biomaterials.

Reduced Order Multiscale Simulation of Diffuse Damage in Concrete

Damage in concrete structures initiates as the growth of diffuse microcracks that is followed by damage localisation and eventually leads to structural failure. Weak changes such as diffuse microcracking processes are failure precursors. Identification and characterisation of these failure precursors at an early stage of concrete degradation and application of suitable precautionary measures will considerably reduce the costs of repair and maintenance. To this end, a reduced order multiscale model for simulating microcracking-induced damage in concrete at the mesoscale level is proposed. The model simulates the propagation of microcracks in concrete using a two-scale computational methodology. First, a realistic concrete specimen that explicitly resolves the coarse aggregates in a mortar matrix was generated at the mesoscale. Microcrack growth in the mortar matrix is modelled using a synthesis of continuum micromechanics and fracture mechanics. Model order reduction of the two-scale model is achieved using a clustering technique. Model predictions are calibrated and validated using uniaxial compression tests performed in the laboratory.

Design of Pseudo-Mechanisms and Multistable Units for Mechanical Metamaterials

Mechanisms-collections of rigid elements coupled by perfect hinges which exhibit a zero-energy motion-motivate the design of a variety of mechanical metamaterials. We enlarge this design space by considering pseudo-mechanisms, collections of elastically coupled elements that exhibit motions with very low energy costs. We show that their geometric design generally is distinct from those of true mechanisms, thus opening up a large and virtually unexplored design space. We further extend this space by designing building blocks with bistable and tristable energy landscapes, realize these by 3D printing, and show how these form unit cells for multistable metamaterials.

Knowledge extraction and transfer in data-driven fracture mechanics

Data-driven approaches promise to usher in a new phase of development in fracture mechanics, but very little is currently known about how data-driven knowledge extraction and transfer can be accomplished in this field. As in many other fields, data scarcity presents a major challenge for knowledge extraction, and knowledge transfer among different fracture problems remains largely unexplored. Here, a data-driven framework for knowledge extraction with rigorous metrics for accuracy assessments is proposed and demonstrated through a nontrivial linear elastic fracture mechanics problem encountered in small-scale toughness measurements. It is shown that a tailored active learning method enables accurate knowledge extraction even in a data-limited regime. The viability of knowledge transfer is demonstrated through mining the hidden connection between the selected three-dimensional benchmark problem and a well-established auxiliary two-dimensional problem. The combination of data-driven knowledge extraction and transfer is expected to have transformative impact in this field over the coming decades.

Oligomodal metamaterials with multifunctional mechanics

Mechanical metamaterials are man-made materials with extraordinary properties that come from their geometrical structure rather than their chemical composition. For instance, they can be engineered to be extremely light and stiff; to shrink sideways when compressed, instead of expanding; or to exhibit programmable shape changes. Such properties often rely on zero-energy modes. In this work, we created a class of mechanical metamaterials with zero-energy modes that can exhibit multiple properties at the same time within a single structure. In particular, we created a metamaterial that can either shrink or expand on the side when compressed, depending on how fast it is compressed. These metamaterials could lead to novel adaptable devices for, for example, robotics and energy absorption applications.

Mechanical metamaterials are artificial composites that exhibit a wide range of advanced functionalities such as negative Poisson’s ratio, shape shifting, topological protection, multistability, extreme strength-to-density ratio, and enhanced energy dissipation. In particular, flexible metamaterials often harness zero-energy deformation modes. To date, such flexible metamaterials have a single property, for example, a single shape change, or are pluripotent, that is, they can have many different responses, but typically require complex actuation protocols. Here, we introduce a class of oligomodal metamaterials that encode a few distinct properties that can be selectively controlled under uniaxial compression. To demonstrate this concept, we introduce a combinatorial design space containing various families of metamaterials. These families include monomodal (i.e., with a single zero-energy deformation mode); oligomodal (i.e., with a constant number of zero-energy deformation modes); and plurimodal (i.e., with many zero-energy deformation modes), whose number increases with system size. We then confirm the multifunctional nature of oligomodal metamaterials using both boundary textures and viscoelasticity. In particular, we realize a metamaterial that has a negative (positive) Poisson’s ratio for low (high) compression rate over a finite range of strains. The ability of our oligomodal metamaterials to host multiple mechanical responses within a single structure paves the way toward multifunctional materials and devices.

Artificial intelligence and machine learning in design of mechanical materials

Artificial intelligence, especially machine learning (ML) and deep learning (DL) algorithms, is becoming an important tool in the fields of materials and mechanical engineering, attributed to its power to predict materials properties, design de novo materials and discover new mechanisms beyond intuitions. As the structural complexity of novel materials soars, the material design problem to optimize mechanical behaviors can involve massive design spaces that are intractable for conventional methods. Addressing this challenge, ML models trained from large material datasets that relate structure, properties and function at multiple hierarchical levels have offered new avenues for fast exploration of the design spaces. The performance of a ML-based materials design approach relies on the collection or generation of a large dataset that is properly preprocessed using the domain knowledge of materials science underlying chemical and physical concepts, and a suitable selection of the applied ML model. Recent breakthroughs in ML techniques have created vast opportunities for not only overcoming long-standing mechanics problems but also for developing unprecedented materials design strategies. In this review, we first present a brief introduction of state-of-the-art ML models, algorithms and structures. Then, we discuss the importance of data collection, generation and preprocessing. The applications in mechanical property prediction, materials design and computational methods using ML-based approaches are summarized, followed by perspectives on opportunities and open challenges in this emerging and exciting field.

Exploring the potential of transfer learning for metamodels of heterogeneous material deformation

From the nano-scale to the macro-scale, biological tissue is spatially heterogeneous. Even when tissue behavior is well understood, the exact subject specific spatial distribution of material properties is often unknown. And, when developing computational models of biological tissue, it is usually prohibitively computationally expensive to simulate every plausible spatial distribution of material properties for each problem of interest. Therefore, one of the major challenges in developing accurate computational models of biological tissue is capturing the potential effects of this spatial heterogeneity. Recently, machine learning based metamodels have gained popularity as a computationally tractable way to overcome this problem because they can make predictions based on a limited number of direct simulation runs. These metamodels are promising, but they often still require a high number of direct simulations to achieve an acceptable performance. Here we show that transfer learning, a strategy where knowledge gained while solving one problem is transferred to solving a different but related problem, can help overcome this limitation. Critically, transfer learning can be used to leverage both low-fidelity simulation data and simulation data that is the outcome of solving a different but related mechanical problem. In this paper, we extend Mechanical MNIST, our open source benchmark dataset of heterogeneous material undergoing large deformation, to include a selection of low-fidelity simulation results that require ≈ 2 - 4 orders of magnitude less CPU time to run. Then, we show that transferring the knowledge stored in metamodels trained on these low-fidelity simulation results can vastly improve the performance of metamodels used to predict the results of high-fidelity simulations. In the most dramatic examples, metamodels trained on 100 high fidelity simulations but pre-trained on 60,000 low-fidelity simulations achieves nearly the same test error as metamodels trained on 60,000 high-fidelity simulations (1 - 1.5% mean absolute percent error). In addition, we show that transfer learning is an effective method for leveraging data from different load cases, and for leveraging low-fidelity two-dimensional simulations to predict the outcomes of high-fidelity three-dimensional simulations. Looking forward, we anticipate that transfer learning will enable us to better capture the influence of tissue spatial heterogeneity on the mechanical behavior of biological materials across multiple different domains.

Foundations for Soft, Smart Matter by Active Mechanical Metamaterials

Emerging interest to synthesize active, engineered matter suggests a future where smart material systems and structures operate autonomously around people, serving diverse roles in engineering, medical, and scientific applications. Similar to biological organisms, a realization of active, engineered matter necessitates functionality culminating from a combination of sensory and control mechanisms in a versatile material frame. Recently, metamaterial platforms with integrated sensing and control have been exploited, so that outstanding non-natural material behaviors are empowered by synergistic microstructures and controlled by smart materials and systems. This emerging body of science around active mechanical metamaterials offers a first glimpse at future foundations for autonomous engineered systems referred to here as soft, smart matter. Using natural inspirations, synergy across disciplines, and exploiting multiple length scales as well as multiple physics, researchers are devising compelling exemplars of actively controlled metamaterials, inspiring concepts for autonomous engineered matter. While scientific breakthroughs multiply in these fields, future technical challenges remain to be overcome to fulfill the vision of soft, smart matter. This Review surveys the intrinsically multidisciplinary body of science targeted to realize soft, smart matter via innovations in active mechanical metamaterials and proposes ongoing research targets that may deliver the promise of autonomous, engineered matter to full fruition.

Devising Materials Manufacturing Toward Lab-to-Fab Translation of Flexible Electronics

Flexible electronics have witnessed exciting progress in academia over the past decade, but most of the research outcomes have yet to be translated into products or gain much market share. For mass production and commercialization, industrial adoption of newly developed functional materials and fabrication techniques is a prerequisite. However, due to the disparate features of academic laboratories and industrial plants, translating materials and manufacturing technologies from labs to fabs is notoriously difficult. Therefore, herein, key challenges in the materials manufacturing of flexible electronics are identified and discussed for its lab-to-fab translation, along the four stages in product manufacturing: design, materials supply, processing, and integration. Perspectives on industry-oriented strategies to overcome some of these obstacles are also proposed. Priorities for action are outlined, including standardization, iteration between basic and applied research, and adoption of smart manufacturing. With concerted efforts from academia and industry, flexible electronics will bring a bigger impact to society as promised.

Enhancing City Sustainability through Smart Technologies: A Framework for Automatic Pre-Emptive Action to Promote Safety and Security Using Lighting and ICT-Based Surveillance

The scope of the present paper is to promote social, cultural and environmental sustainability in cities by establishing a conceptual framework and the relationship amongst safety in urban public space (UPS), lighting and Information and Communication Technology (ICT)-based surveillance. This framework uses available technologies and tools, as these can be found in urban equipment such as lighting posts, to enhance security and safety in UPS, ensuring protection against attempted criminal activity. Through detailed literary research, publications on security and safety concerning crime and lighting can be divided into two periods, the first one pre-1994, and the second one from 2004–2008. Since then, a significant reduction in the number of publications dealing with lighting and crime is observed, while at the same time, the urban nightscape has been reshaped with the immersion of light-emitting diode (LED) technologies. Especially in the last decade, where most municipalities in the EU28 (European Union of all the member states from the accession of Croatia in 2013 to the withdrawal of the United Kingdom in 2020) are refurbishing their road lighting with LED technology and the consideration of smart networks and surveillance is under development, the use of lighting to deter possible attempted felonies in UPS is not addressed. To capitalize on the potential of lighting as a deterrent, this paper proposes a framework that uses existing technology, namely, dimmable LED light sources, presence sensors, security cameras, as well as emerging techniques such as artificial intelligence (AI)-enabled image recognition algorithms and big data analytics and presents a possible system that could be developed as a stand-alone product to alert possible dangerous situations, deter criminal activity and promote the perception of safety thus linking lighting and ICT-based surveillance towards safety and security in UPS.

Molecular generation targeting desired electronic properties via deep generative models

As we seek to discover new functional materials, we need ways to explore the vast chemical space of precursor building blocks, not only generating large numbers of possible building blocks to investigate, but trying to find non-obvious options, that we might not suggest by chemical experience alone. Artificial intelligence techniques provide a possible avenue to generate large numbers of organic building blocks for functional materials, and can even do so from very small initial libraries of known building blocks. Specifically, we demonstrate the application of deep recurrent neural networks for the exploration of the chemical space of building blocks for a test case of donor-acceptor oligomers with specific electronic properties. The recurrent neural network learned how to produce novel donor-acceptor oligomers by trading off between selected atomic substitutions, such as halogenation or methylation, and molecular features such as the oligomer's size. The electronic and structural properties of the generated oligomers can be tuned by sampling from different subsets of the training database, which enabled us to enrich the library of donor-acceptors towards desired properties. We generated approximately 1700 new donor-acceptor oligomers with a recurrent neural network tuned to target oligomers with a HOMO-LUMO gap <2 eV and a dipole moment <2 Debye, which could have potential application in organic photovoltaics.

The Critical Raw Materials in Cutting Tools for Machining Applications: A Review

A variety of cutting tool materials are used for the contact mode mechanical machining of components under extreme conditions of stress, temperature and/or corrosion, including operations such as drilling, milling turning and so on. These demanding conditions impose a seriously high strain rate (an order of magnitude higher than forming), and this limits the useful life of cutting tools, especially single-point cutting tools. Tungsten carbide is the most popularly used cutting tool material, and unfortunately its main ingredients of W and Co are at high risk in terms of material supply and are listed among critical raw materials (CRMs) for EU, for which sustainable use should be addressed. This paper highlights the evolution and the trend of use of CRMs) in cutting tools for mechanical machining through a timely review. The focus of this review and its motivation was driven by the four following themes: (i) the discussion of newly emerging hybrid machining processes offering performance enhancements and longevity in terms of tool life (laser and cryogenic incorporation); (ii) the development and synthesis of new CRM substitutes to minimise the use of tungsten; (iii) the improvement of the recycling of worn tools; and (iv) the accelerated use of modelling and simulation to design long-lasting tools in the Industry-4.0 framework, circular economy and cyber secure manufacturing. It may be noted that the scope of this paper is not to represent a completely exhaustive document concerning cutting tools for mechanical processing, but to raise awareness and pave the way for innovative thinking on the use of critical materials in mechanical processing tools with the aim of developing smart, timely control strategies and mitigation measures to suppress the use of CRMs.