Computer-Aided Design最新文献

We present an approach for systematic design of generalized Plesiohedra, a new type of 3D space-filling shapes that can even include unchained handlebodies. We call these handlebody plesiohedra unchained, since they are topologically interlocked, i.e., they can be assembled and disassembled without breaking any of the solids apart and they can keep in place with a set of boundary constraints. These space-filling shapes (i.e. congruent prototiles) are obtained from the Voronoi decomposition of symmetric Delone (Delaunay) point sets. To create this new class of shapes, we generalize the design space of classical Plesiohedra by introducing two novel geometric steps: (a) extension of point sites to piecewise linear approximations of higher-dimensional geometries and (b) extension of symmetries to 3D crystallographic symmetries. We show how these specific collections of higher-dimensional geometries can admit the symmetric Delone property. A Voronoi partitioning of 3D space using these specific collections of higher-dimensional shapes as Voronoi sites naturally results in congruent prototiles. This generalizes the idea of classical Plesiohedra by allowing for piecewise linear approximation of curved edges and faces, non-convex boundaries, and even handlebodies with positive genus boundaries to provide truly volumetric material systems in contrast to traditional planar or shell-like systems. To demonstrate existence of these solid shapes, we produced a large set of unchained congruent space-filling handlebodies as proofs of concept. For this, we focused our investigation using isometries of some space-filling polyhedra, such as a cube and a truncated octahedron with circles, and curve complexes as Voronoi sites. These results point to a rich and vast parametric design space of unchained handlebody plesiohedra making them an excellent representations for engineering applications such as topologically interlocked architectured materials.

This article introduces a novel computer-aided procedure to design optimised coated structures with precise shell thickness control using the Smallest Univalue Segment Assimilating Nucleus operator and a novel augmentation-projection technique. Structures with heterogeneous sections, or coated structures, combine two different materials for the nucleus and the shell, which are generally chosen so that the material in the infill is lighter and the material in the coating is stiffer, which in this work are supposed homogeneous. Solving the interface problem requires material properties interpolation equations that consider three material phases, accurate placement of the coating over the base material, and precise control over the coating's thickness. The formation of the coating is controlled by the Smallest Univalue Segment Assimilating Nucleus, an edge detection operator developed in Digital Image Processing. The coating's thickness is controlled by an innovative methodology consisting of the projection of an augmented contour field, which is shown to create a constant thickness coating around the material domain. The optimisation problem is solved with the Sequential Element Rejection and Admission method. The validity of the procedure has been verified by solving various numerical application examples.

Recovering geometric regularities from scanned mesh models with various types of surface features has always been a challenging task in reverse engineering. To address this problem, this paper presents a regularity detection and enhancement framework for surface mesh reconstruction. Initially, surface patches are identified by decomposing the original model into planar, quadric and freeform surface patches. Similar surface patches are aligned with each other by pairwise registration, and symmetry patterns are detected from the accumulated affine transformations using an improved grid fitting method. Regular relations between symmetry patterns and individual surface patches are enumerated and progressively strengthened by orientation, dimension and placement optimizations. Finally, the resultant model with enhanced regularities is obtained by projecting surface patches onto the optimized parametric surfaces iteratively. Comparative experiments on test models demonstrate that the proposed method outperforms existing methods in recovering both lower- and higher-level regularities of engineering models, especially those with freeform surfaces.

This paper presents a novel mesh denoising approach designed specifically for developable models with curved folds, going beyond traditional denoising metrics to focus on restoring the model’s developability. We introduce a metric based on normal variation to assess mesh developability and integrate it into an optimization problem that aims to increase the sparsity of the normal vector field, leading to a dedicated mesh denoising algorithm. The performance of our method is evaluated across a wide range of criteria, including standard metrics and surface developability determined through Gaussian curvature. Through testing on a variety of noisy models and comparison with several state-of-the-art mesh denoising and developability optimization techniques, our approach demonstrates superior performance in both traditional metrics and the enhancement of mesh developability.

Functionally Graded Cellular Materials (FGCM) with variable volume fractions have demonstrated significant advantages, including weight reduction, improved stiffness, and enhanced load distribution, when compared to uniform density counterparts. Their design is often characterized by the application of a density distribution to locally modify Representative Volume Elements (RVEs). Current studies have explored the application of Triply Periodic Minimal Surfaces (TPMS) topologies, given their capability to create seamless and interconnected structures, thus avoiding stress concentration issues commonly encountered in traditional lattice configurations. Consequently, this paper introduces a design methodology tailored to TPMS-based FGCM allowing for independent or simultaneous adjustments of RVE thickness and size. Models for predicting relative density as a function of the RVE design parameters of Primitive and Gyroid topologies are presented and discussed. These models are employed to adapt the topologies to three-dimensional density distributions. The proposed method is implemented as a set of design tools and is illustrated for the studied TPMS topologies.

We present a method to resolve visual artifacts of a state-of-the-art iso-surface extraction algorithm by generating feature-preserving surface patches for isolated arbitrarily complex, single voxels without the need for further adaptive subdivision. In the literature, iso-surface extraction from a 3D voxel grid is limited to a single sharp feature per minimal unit, even for algorithms such as Cubical Marching Squares that produce feature-preserving surface reconstructions. In practice though, multiple sharp features can meet in a single voxel. This is reflected in the triple dexel model, which is used in simulation of CNC manufacturing processes. Our approach generalizes the use of normal information to perfectly preserve multiple sharp features for a single voxel, thus avoiding visual artifacts caused by state-of-the-art procedures.

Shell structures with high stiffness-to-weight ratios are desirable in various engineering applications. Topology optimization serves as a popular and effective tool for generating optimal shell structures. The solid isotropic material with penalization (SIMP) method is often chosen because of its simplicity and convenience. However, SIMP method is typically integrated with conventional Finite Element Analysis (FEA) which has limitations in computational accuracy. Achieving high accuracy with FEA necessitates a substantial number of elements, leading to computational burdens. In addition, the discrete representation of the material distribution function may result in rough boundaries. Owing to these limitations, this paper proposes an Isogeometric Analysis (IGA) based SIMP method for optimizing the topology of shell structures based on Reissner–Mindlin theory. This method uses Non-Uniform Rational B-Splines (NURBS) to represent both the shell structure and the material distribution function with the same basis functions, allowing for higher accuracy and smoother boundaries. The optimization model takes compliance as the objective function with a volume fraction constraint and the coefficients of the density function as design variables, resulting in an optimized shell structure defined by the material distribution function. To obtain fairing boundaries of the holes in the optimized shell structure, further process is conducted by fitting the boundaries with fair B-spline curves automatically. Furthermore, the proposed IGA-SIMP framework is applied to generate porous shell structures by imposing different local volume fraction constraints. Numerical examples are provided to demonstrate the feasibility and efficiency of the IGA-SIMP method, showing that it outperforms the FEA-SIMP method and produces smoother boundaries.

Automatic control of a workpiece being manufactured is a requirement to ensure in-line correction and thus move towards a more intelligent manufacturing system. There is therefore a need to develop control strategies which are capable of taking precise account of real working conditions and enabling first-time-right control. As part of such a smart-control strategy, this paper introduces a machine learning-based approach capable of accurately predicting a priori the 3D coverage of a part according to a scan configuration given as input, i.e. predicting before scanning it which areas of the part will be acquired for real. This corresponds to a paradigm shift, where coverage estimation no longer relies on theoretical visibility criteria, but on rules learned from a large amount of data acquired in real-life conditions. The proposed 3D Scan Coverage Prediction Network (3DSCP-Net) is based on a 3D feature encoding and decoding module, which is capable of taking into account the specifics of the scan configuration whose impact on the 3D coverage is to be predicted. To take account of real working conditions, features are extracted at various levels, including geometric ones, but also features characterising the way structured-light projection behaves. The method is thus able to incorporate inter-reflection and overexposure issues into the prediction process. The database used for the training was built using an ad-hoc platform specially designed to enable the automatic acquisition and labelling of numerous point clouds from a wide variety of scan configurations. Experiments on several parts show that the method can efficiently predict the scan coverage, and that it outperforms conventional approaches based on purely theoretical visibility criteria.

The generation of support for 3D models toward 3D printing is a highly challenging task that is of great need in many additive manufacturing processes. In this work, we explore the use of multiresolution geometric lattices to generate support with controlled contact locations. That is, with bounds on the maximal distance between adjacent local support points. A variety of end-user controls over the synthesized support are provided, such as the angular slopes in the model that are provided with support and/or controls on the dimensions and sizes of the support lattice tiles. These controls are augmented with the option of an automated optimization via a direct link to analysis. We demonstrate this proposed lattice approach for support synthesis on several 3D models of different types.