The dislocation behaviors in body-centered cubic (BCC) multi-principal element alloys (MPEAs) containing ordered phases remain incompletely understood. This study elucidates the dislocation-dominated deformation mechanism in the Al-12(HfNbTiZr)(88) with ordered phases on the scale of tens to one hundred nanometers. Advanced techniques, including HAADF-STEM, 3D APT, and DFT calculations, were employed to reveal the presence of substantial B2-type ordered phases coherently embedded in the BCC matrix. In-situ TEM observations combined with molecular dynamics simulations demonstrate that individual dislocations pair into superdislocations within the ordered phases. Interestingly, these superdislocations can readily enter and exit the ordered phases, without forming the pronounced pile-ups typically observed at ordered phase-matrix interfaces. Notably, these superdislocations frequently undergo cross-slip from the primary {110}< 111 > to the secondary {112}< 111 > slip system upon encountering pronounced local lattice friction. Such intriguing superdislocation behaviors enable the Al-12(HfNbTiZr)(88) alloy to achieve a good balance between strength and ductility, compared with the chemically disordered HfNbTiZr, short-range-ordered Al-8(HfNbTiZr)(92) (the size of the ordered phase < 1 nm), and intermetallic compound Al-16(HfNbTiZr)(84) (the size of the ordered phase > 500 nm). These findings deepen our understanding of dynamic dislocation slip behavior and elucidate the critical influence of nanoscale ordering on plastic deformation in BCC MPEAs.

Room-temperature compressive creep deformation threatens the long-term reliability of deep-sea pressure hulls. Its distinct stress states and thermal activation energies make conventional high-temperature or tensile creep mechanisms inadequate for this scenario. This study investigates the compressive creep mechanism of Ti80 alloy, focusing on microstructural evolution and dislocation behavior under prolonged near-yield stress conditions. Using multiscale characterization techniques from the micron to atomic scale, including quasi in-situ and in-situ methods, we demonstrate that premature dislocation nucleation at interfaces-the primary mechanism for roomtemperature compressive creep-is linked to nanoscale gamma phases at alpha/(3 interfaces, which have lower critical resolved shear stress (CRSS) for dislocation activation and serve as bridges for dislocations transmission. Our findings provide mechanistic insight into room-temperature compressive creep and offer guidance for alloy design and service life prediction of deep-sea pressure hulls.

The unfitted finite element (FE) methods offer significant advantages in engineering analysis by embedding the structure within a simple background mesh and eliminating the need for complex and labor-intensive mesh generation. A high solution accuracy can still be achieved via constructing, on each coarse element, standard polynomial shape functions, or numerical (piecewise) shape functions. However, the strategy incurs significant technical challenges due to the unavoidable occurrence of cut elements of arbitrarily small size that may greatly deteriorate the condition number of the stiffness matrix. To address the issue, we propose the concept of reduced coarse elements. By formulating high-order numerical shape functions as the product of a boundary interpolator and a boundary-interior mapping, a detailed condition number analysis reveals the dependence of the numerical shape function conditioning on the boundary interpolator. Based on these findings, we develop a new type of reduced coarse elements and their associated numerical shape functions to address the conditioning challenges. The stability, accuracy, convergence rate, and efficiency of the approach are tested through various numerical examples in comparison with other cutting-edge approaches. Its performance on a multi-material printed circuit board (PCB) example of 183 million fine mesh nodes is also tested.

Existing industrial 3D garment meshes already cover most real-world clothing geometries, yet their texture diversity remains limited. To acquire more realistic textures, generative methods are often used to extract Physically-based Rendering (PBR) textures and materials from large collections of wild images and project them back onto garment meshes. However, most image-conditioned texture generation approaches require strict topological consistency between the input image and the input 3D mesh, or rely on accurate mesh deformation to match to the image poses, which significantly constrains the texture generation quality and flexibility. To address the challenging problem of non-isometric image-based garment texture generation, we construct 3D Garment Videos, a physically simulated, garment-centric dataset that provides consistent geometry and material supervision across diverse deformations, enabling robust cross-pose texture learning. We further employ Nano Banana for high-quality non-isometric image editing, achieving reliable cross-topology texture generation between non-isometric image-geometry pairs. Finally, we propose an iterative baking method via uncertainty-guided view selection and reweighting that fuses multi-view predictions into seamless, production-ready PBR textures. Through extensive experiments, we demonstrate that our feedforward dual-branch architecture generates versatile and spatially aligned PBR materials suitable for industry-level 3D garment design.

Recent breakthroughs in 3D generation have enabled the synthesis of high-fidelity individual assets. However, generating 3D compositional objects from single images–particularly under occlusions–remains challenging. Existing methods often degrade geometric details in hidden regions and fail to preserve the underlying object-object spatial relationships (OOR). We present a novel framework Interact3D designed to generate physically plausible interacting 3D compositional objects. Our approach first leverages advanced generative priors to curate high-quality individual assets with a unified 3D guidance scene. To physically compose these assets, we then introduce a robust two-stage composition pipeline. Based on the 3D guidance scene, the primary object is anchored through precise global-to-local geometric alignment (registration), while subsequent geometries are integrated using a differentiable Signed Distance Field (SDF)-based optimization that explicitly penalizes geometry intersections. To reduce challenging collisions, we further deploy a closed-loop, agentic refinement strategy. A Vision-Language Model (VLM) autonomously analyzes multi-view renderings of the composed scene, formulates targeted corrective prompts, and guides an image editing module to iteratively self-correct the generation pipeline. Extensive experiments demonstrate that Interact3D successfully produces promising collsion-aware compositions with improved geometric fidelity and consistent spatial relationships.

CAD/CAE interoperability remains a fundamental bottleneck in simulation-driven product development because design and analysis models represent different but interdependent views of the same product. CAD provides the authoritative geometric and semantic model for editing, manufacturing definition, assembly management, and product-data control, whereas CAE requires a physically meaningful computational domain with appropriate abstraction, materials, boundary conditions, discretization, and solver assumptions. Interoperability is therefore not merely file translation, but a bidirectional association between geometry and physics across the CAD-to-CAE-to-CAD loop.From a CAD perspective, this review focuses on the model transformations that determine whether simulation evidence can be generated, interpreted, and returned to design. It examines defect diagnosis and repair for model validity, simplification and idealization for analysis readiness, embedded and unfitted analysis-model generation, associative traceability for attribute and result transfer, and CAE-to-CAD feedback for design update. Industrial platforms, AI-assisted workflows, and benchmarks are discussed where they affect these transformations. Across these topics, geometric, topological, semantic, numerical, and provenance contracts are used to clarify what information is preserved, approximated, discarded, or recorded. The review identifies a research agenda centered on evidence-aware transformation, semantic correspondence, analysis-aware abstraction, traceable embedded analysis and verification-driven AI.

Reconstructing editable feature-based computer-aided design (CAD) models from boundary representation (B-Rep) data is valuable for accelerating design iteration and improving model reuse, but it is hindered by (1) feature identification, especially under feature interactions, and (2) non-unique feature sequence determination. This paper proposes a two-level learning pipeline that combines a multi-task classification network and a link prediction network. The multi-task network jointly predicts per-face feature types and per-edge precedence relations, while the link prediction network infers missing connections to merge fragmented regions into complete feature instances. On the Fusion 360 Segmentation dataset (35,680 user-created models), 85,511 feature instances are manually annotated for training and evaluation. The proposed classifier achieves face/edge accuracies of 94.51%/87.21% and intersection-over-union (IoU) scores of 78.13%/56.41% (all reported as means), improving face IoU by 4.52 and 1.03 percentage points over UV-Net (a typical B-Rep learning baseline) and BRepNet (a typical topological message-passing baseline for B-Rep models), respectively. The link predictor reaches an accuracy of 83.87% with an IoU of 72.22% (means), enabling feature instance identification with a mean instance identification accuracy (mInsIdenAcc) of 74.11% and a mean instance identification IoU (mInsIdenIoU) of 87.83%. Based on the resulting feature precedence graph, valid feature sequences are automatically generated and can be replayed to reconstruct interpretable and editable feature histories for downstream CAD applications.

We propose a novel approach to structure-aware topology optimization (SATO) to generate physically plausible multi-component structures with diverse stylistic variations. Traditional TO methods often operate within a discrete voxel-defined design space, overlooking the underlying structure-aware, which limits their ability to accommodate stylistic design preferences. Our approach leverages variational autoencoders (VAEs) to encode both geometries and corresponding structures into a unified latent space, capturing part arrangement features. The design target is carefully formulated as a topology optimization problem taking the VAE code as design variables under physical constraints, and solved numerically via analyzing the associated sensitivity with respect to the VAE variables. Our numerical examples demonstrate the ability to generate lightweight structures that balance geometric plausibility and structural performance with much enhanced stiffness that outperforms existing generative techniques. The method also enables the generation of diverse and reliable designs, maintaining structural integrity throughout, via a direct smooth interpolation between the optimized designs. The findings highlight the potential of our approach to bridge the gap between generative design and physics-based optimization by incorporating deep learning techniques.

We propose an augmented sphere tracing (AST) pipeline that seamlessly integrates editing, rendering, and slicing of mega-scale periodic shell-lattice structures. Traditional STL-based pipelines face challenges such as time-consuming format conversions, high storage requirements, and complex blending issues between discrete lattice and shell components, often resulting in a loss of geometric accuracy. Alternatively, implicit-based pipelines excel at smooth modeling and robust Boolean operations but require inefficient and error-prone conversions of STL shells into implicit forms, complicating the rendering process. To address these issues, AST combines hybrid implicit lattice and mesh shell representations, eliminating the need for explicit 3D model construction and unnecessary geometric format conversions. It overcomes the major challenges of hybrid forms and mega-scale rendering by using an augmented tracing distance query that avoids costly signed distance field (SDF) calculations while preserving geometric details. Additionally, it employs a local tracing distance query within a single cell, leveraging lattice periodicity for efficiency. The pipeline also supports various types of shell-lattices in industrial applications, including blending, warping, field-directed distributions, region-specific cell types, and produces arbitrary directional slicing for manufacturing. As demonstrated by various examples implemented in WebGPU, AST archives high efficiency and accuracy in real-time rendering of shell-lattices with billions of beams on an RTX 3090, outperforming traditional pipelines in storage, frame time, and detail preservation.

For deep-sea submersibles, the service life of titanium alloys in lightweight pressure hulls is dictated by roomtemperature creep deformation. Under high stress in deep-sea, the mechanism for room-temperature creep is primarily dominated by dislocation slip in the soft grains. Guided by the principles of dislocation pile-up and back stress hardening, a pre-compression treatment was applied to Ti80 alloy. Via pre-compression, the dislocation density increased significantly by 1.9-2.5 times in some softer grains whereas the average dislocation density only experienced a 22 % increase. This treatment effectively elevated the critical resolved shear stress (CRSS) and enhanced the resistance to dislocation motion. Accordingly, the creep stress threshold is raised by at least 14 % and the total creep strain is reduced by 80 % after 1000 h of creep at 90 % the yield strength. A creep constitutive model based on back stress evolution was developed to accurately describe the creep behavior of Ti80 alloy. That model incorporates an estimation of the initial back stress induced by pre-compression treatment and its effect on dislocation slip. The results demonstrate crucial insights into the optimization of materials for deep-sea pressure hulls and their long-term performance prediction.

Converting topology-optimization results into parametric models is crucial for manufacturing lightweight, highstiffness products. However, currently available technologies cannot effectively automate and perform this conversion, especially for three-axis CNC machining. To bridge this gap, this study proposes an automatic approach for generating parametric models from topology-optimization results. First, integrating the machining characteristics of three-axis CNC machining, surface voxel accessibility analysis and different voxel clustering are carried out to determine the optimal machining directions and removable geometry of the (common) raw material model. Then, a parametric sketch contours generation method is presented for the removable geometry. This also provides the essential preparation for generating a parametric model by adding subtractive features to the raw material model. Particularly, a classified layered projection method is developed to ensure the final parametric model preserves the shape of the topology-optimization result as much as possible. This method can project and fit the removable geometry into quadratic curve sketch contours in a layer-by-layer scheme. Based on the sketch contour of each layer, the corresponding subtractive feature can be generated and added to the raw material model to remove the corresponding removable geometry. Performing this subtractive process layer by layer can generate the final parametric model of the topology-optimization result. Herein, certain constraints are also implemented during the subtractive process to ensure that the final model can preserve the stiffness of the topology-optimization result. Finally, the automatic conversion experiments on two complex and representative topology-optimization results show an average reduction of 19.3 % in maximum displacement (i.e., compliance) and an average increase of 45.2 % in mass as well as 28.4 % increase in volume when averaging the changes across both generated parametric models compared with their original topology-optimization results. The methodological comparisons also show that the presented approach has special benefits, including the ability to convert topology-optimization results automatically and effectively into parametric models, while maintaining stiffness and being processable by three-axis CNC machining.

Topology optimization can maximally leverage the high DOFs and mechanical potentiality of porous foams but faces challenges in adapting to free-form outer shapes, maintaining full connectivity between adjacent foam cells, and achieving high simulation accuracy. Utilizing the concept of Voronoi tessellation may help overcome the challenges owing to its distinguished properties on highly flexible topology, natural edge connectivity, and easy shape conforming. However, a variational optimization of the so-called Voronoi foams has not yet been fully explored. In addressing the issue, a concept of explicit topology optimization of open-cell Voronoi foams is proposed that can efficiently and reliably guide the foam's topology and geometry variations under critical physical and geometric requirements. Taking the site (or seed) positions and beam radii as the DOFs, we explore the differentiability of the open-cell Voronoi foams w.r.t. its seed locations, and propose a highly efficient local finite difference method to estimate the derivatives. During the gradient-based optimization, the foam topology can change freely, and some seeds may even be pushed out of shape, which greatly alleviates the challenges of prescribing a fixed underlying grid. The foam's mechanical property is also computed with a much-improved efficiency by an order of magnitude, in comparison with benchmark FEM, via a new material-aware numerical coarsening method on its highly heterogeneous density field counterpart. We show the improved performance of our Voronoi foam in comparison with classical topology optimization approaches and demonstrate its advantages in various settings.

The simulation of elastic slender objects like cables is essential for industrial applications in predicting elastic behaviors and life cycles. The Cosserat model and its variants are the dominant approaches due to their high efficiency and accuracy. However, these assume cables with homogeneous interiors and thus cannot simulate hybrid cables containing different materials. We address this by developing a novel coarsened-shell-based Cosserat (CSC) model. The CSC model constructs a material-aware elastic energy function along the cable’s cross-section to describe the global elastic behavior. The CSC model is specifically developed by carefully leveraging the strengths of three approaches: the Cosserat theory to model slender cables, the Kirchhoff–Love shell theory to model the cable’s cross-sectional energy, and numerical coarsening to reduce the degrees of freedom in the shell simulation via constructing a set of new types of material-aware shape/base functions. This allows the more accurate computation of the local and global deformations of hybrid cables, surpassing the classical Cosserat models in accuracy.

Numerical coarsening is an approach that constructs coarsened shape (or basis) functions for simulating heterogeneous structures on a coarse mesh, and seeks the fine mesh solution on a finite dimensional space spanned by these shape functions. It avoids the challenging issue of conforming meshing, reduces computational costs while maintaining high simulation accuracy, and has huge potential in the simulation of composites or even complex CAD models in fictitious domain methods. The C1-continuity of the coarsened shape functions is of vital importance in determining the simulation quality, such as the continuity of stress solution, but till now has not been developed. In addressing the issue, the study proposes to construct coarsened shape functions using B-spline patches (piecewise surfaces in 2D or volumes in 3D). By carefully devising an intermediate interpolation patch that covers the coarse element boundaries, important properties of C1-continuity and partition of unity are achieved, being able to produce high simulation stability and accuracy. By being explicitly formulated as two sequent displacement mappings from coarse nodes to the boundary fine nodes and then to the interior fine nodes, the shape functions are cheaply derived with small efforts in solving a small-size linear system for each coarse element. Performance of the coarsened shape functions, implemented within the material point method (MPM), is demonstrated through various numerical examples in comparisons with benchmark results in terms of stability, simulation accuracy, and computational efficiency.

Ti alloys with dual alpha and beta phases are commonly used in various applications due to their superior mechanical properties. In the present work, to unravel the effect of orientation angle theta between the alpha and beta phases, molecular dynamics simulations were performed on Ti alloy bicrystal samples with the orientation angle varied from 0 to 90. Under compressive stress, the phase interface was observed to nucleate defects, leading to four distinct deformation mechanisms depending on the orientation angle. Prismatic < a > dislocations were identified at orientation angles around 0, while shear bands were observed in the range of 5(degrees) < theta < 35. Basal < a > dis-locations were present at 35 < theta < 70(degrees), and pyramidal II < c + a > dislocations were observed at 70(degrees) < theta < 90. The orientation-dependent variation of interface-mediated deformation mechanisms is explained by virtue of the critical resolved shear stress of each deformation mechanism and the Schmid factor on the corresponding slip system. Our findings provide insight into the role played by phase interfaces in determining the plastic deformation mechanisms of two-phase Ti alloys.

The planning of multi-branch cable harness layouts holds significant practical importance in aircraft industrial contexts, yet it has received limited attention in prior research. This study aims to address the matter concerning the significance of managing multiple constraints and preventing loops. It formulates the problem as an optimization problem in 3D free-form space and resolves it using an extended A* path planning approach in combination with the ant colony optimization algorithm. Initially, a feasible search space for wiring is established through the repair and simplification of the input CAD model. Subsequently, the topology of a multi-branched wiring harness is identified, taking into account industrial requirements related to cable physics, turning, support, bundling, and electromagnetic compatibility constraints. Specifically, the disassembly or merging of branches and loops is employed to avoid wire loops. Ultimately, we propose an A*–ant colony optimization algorithm (A*-ACO) with an enhanced heuristic function for neighboring points, incorporating a concentration increment model. Experimental tests illustrate the effectiveness of this approach in minimizing wire loops and reducing the total cable layout cost, considering factors such as length, bundling, and turning costs. It results in a reduction of 67.0%, 68.5%, and 51.1% compared to A*, ACO, and manual wiring methods, respectively.

Overhangs are usually inevitable in Additive Manufacturing (AM). Moreover, making their downward surfaces reach higher accuracy after printing has an imperative impact on improving the accuracy of them and their corresponding as-printed parts. Although extensive studies have been devoted to improving the accuracy of as-printed parts, the effective, general, reliable, and less-consumption (in time and material) approach to make an overhang (especially on its downward surfaces) reach its ideal accuracy is still rare. Hence, a new machine-leaning-based geometric compensation approach is proposed in this study. First, based on the Taguchi method, a series of (overhang) benchmarks are designed and printed to collect the (geometric) deviations of the downward surfaces. With these data, a new deviation predictor is established based on the gaussian process regression model. The predictor can effectively predict the deviation(s) (as well as its corresponding quantified uncertainty) of each downward surface (after printing) by using a pointwise manner. Meanwhile, to make the downward surface of an overhang meet its ideal accuracy (after printing) with high confidence, a specific stochastic chance-constrained programming problem is first formulated to evaluate the new and suitable overhang height of each point on the downward surface. Based on the above-mentioned predictor, the programming problem gets solved by developing a new overhang-height optimization method that integrates the Monte Carlo simulation with the particle swarm optimization. After that, a support structure-associated compensation method is presented to make a pointwise compensation (according to the above-evaluated overhang height) on the downward surface and determine the support structures’ optimized number and anchor positions. Finally, experiments on several representative overhangs are also implemented based on a typical material extrusion machine to validate the effectiveness of the proposed approach. The results show that the proposed approach reduces the deviations of the downward surface by up to 63.92% on average, saving up to 40% of material (less support structures). In addition, methodological comparisons with state-of-the-art approaches are also implemented. The comparisons show that the proposed approach has the great potential to reliably improve the accuracy of the overhang’s downward surface(s) after printing in a less-consumption manner.