Search

International audience; Advancements in machine learning have sparked significant interest in designing mechanical metamaterials, i.e., materials that derive their properties from their inherent microstructure rather than just their constituent material. We propose a data-driven exploration of the design space of growth-based cellular metamaterials based on star-shaped distances. These two-dimensional metamaterials are based on periodically-repeating unit cells consisting of material and void patterns with non-trivial geometries. Machine learning models exploiting large datasets are then employed to inverse design growth-based metamaterials for tailored anisotropic stiffness. Firstly, a forward model is created to bypass the growth and homogenization process and accurately predict the mechanical properties given a finite set of design parameters. Secondly, an inverse model is used to invert the structure–property maps and enable the accurate prediction of designs for a given anisotropic stiffness query. We successfully demonstrate the frameworks’ generalization capabilities by inverse designing for stiffness properties chosen from outside the domain of the design space.

We propose a new approach for unsupervised learning of hyperelastic constitutive laws with physics-consistent deep neural networks. In contrast to supervised learning, which assumes the availability of stress–strain pairs, the approach only uses realistically measurable full-field displacement and global reaction force data, thus it lies within the scope of our recent framework for Efficient Unsupervised Constitutive Law Identification and Discovery (EUCLID) and we denote it as NN-EUCLID. The absence of stress labels is compensated for by leveraging a physics-motivated loss function based on the conservation of linear momentum to guide the learning process. The constitutive model is based on input-convex neural networks, which are capable of learning a function that is convex with respect to its inputs. By employing a specially designed neural network architecture, multiple physical and thermodynamic constraints for hyperelastic constitutive laws, such as material frame indifference, material stability, and stress-free reference configuration are automatically satisfied. We demonstrate the ability of the approach to accurately learn several hidden isotropic and anisotropic hyperelastic constitutive laws – including e.g., Mooney–Rivlin, Arruda–Boyce, Ogden, and Holzapfel models – without using stress data. For anisotropic hyperelasticity, the unknown anisotropic fiber directions are automatically discovered jointly with the constitutive model. The neural network-based constitutive models show good generalization capability beyond the strain states observed during training and are readily deployable in a general finite element framework for simulating complex mechanical boundary value problems with good accuracy., Journal of the Mechanics and Physics of Solids, 169, ISSN:0022-5096, ISSN:1873-4782

To tackle the challenge in artificial spinal disc (ASD) design of restoring the mechanics of a natural disc, this study proposes an innovative lattice-based ASD for reproducing a patient-specific anisotropic rotational response, inspired by the design freedom provided by lattice structures. Motivated by the great potential of machine learning to improve computational design processes, a method is proposed for computationally efficient topology optimization using artificial neural networks (ANNs) and a subsequent member sizing for automating the design of patient-specific ASDs. The results reported in this study show a good match between the optimized ASDs’ six rotational stiffnesses with those of both L2-L3 and L4-L5 human lumbar discs. Additionally, the fast convergence rate of the optimization verifies the application of ANNs and the proposed strategy to reduce the design space by formulating the design problem as optimizing the unit-cell distribution in a predefined grid. Therefore, the study demonstrates that a lattice-based ASD is able to reproduce patient-specific anisotropic rotational response and that a machine-learning-based method improves the computational efficiency of an automated design process to produce personalized ASD designs., Computers in Biology and Medicine, 152, ISSN:0010-4825

Biocompatibility and high responsiveness to magnetic fields are fundamental requisites to translate magnetic small-scale robots into clinical applications. The magnetic element iron exhibits the highest saturation magnetization and magnetic susceptibility while exhibiting excellent biocompatibility characteristics. Here, a process to reliably fabricate iron microrobots by means of template-assisted electrodeposition in 3D-printed micromolds is presented. The 3D molds are fabricated using a modified two-photon absorption configuration, which overcomes previous limitations such as the use of transparent substrates, low writing speeds, and limited depth of field. By optimizing the geometrical parameters of the 3D molds, metallic structures with complex features can be fabricated. Fe microrollers and microswimmers are realized that demonstrate motion at ≈20 body lengths per second, perform 3D motion in viscous environments, and overcome higher flow velocities than those of "conventional 3D printed helical microswimmers." The cytotoxicity of these microrobots is assessed by culturing them with human colorectal cancer (HCT116) cells for four days, demonstrating their good biocompatibility characteristics. Finally, preliminary results regarding the degradation of iron structures in simulated gastric acid liquid are provided.