Abstract:Despite international efforts to support specification and exchange of CAD data through acknowledged standards and data interfaces, in practice several problems still remain. One type of problem is related to the reconstruction and verification of CAD data containing defects. Major problems causing such defects are related to differences in the internal modeling accuracy of systems. Generated models that sometimes contain degenerated edges and faces having either almost zero length or zero area exacerbate the situation. Defects in the topology and / or the geometry of such geometric models have a severe impact on computer-aided post-design processes such as finite element analysis, STL file generation, tool path generation, etc. In the meantime, until improvements can be made in the processes that generated those corrupted CAD data, correction of possible errors in CAD data can be achieved by introducing an intermediate post-processing step aimed at reconstructing an error-free, valid solid model. Within the given scope of this paper, we introduce and investigate the application of a face-sewing operator, which can be applied to CAD data based either on surface models or on solid models. Operation and quality of the model reconstruction are determined by the tolerances of spatial model space points located on the face boundaries and their distances apart. Due to the use of reference kernel architecture featuring a Brep scheme within an integrated multi-dimensional topology supporting both surface models and solid models, a CAD data interface based on a standard data exchange format and an internal translator that classifies, converts and correlates pure geometric entities, sub-setting geometric entities and topological entities, open and highly flexible system architecture in respect to considered CAD data and geometric models can be maintained.

SolidGen: An Autoregressive Model for Direct B-rep Synthesis

BrepGen: A B-rep Generative Diffusion Model with Structured Latent Geometry

3D-Aware Image Synthesis Via Learning Structural and Textural Representations

3D Solid Model Generation Method Based on a Generative Adversarial Network

Generative Creativity: Adversarial Learning For Bionic Design

Brep2Seq: A dataset and hierarchical deep learning network for reconstruction and generation of computer-aided design models

B-rep Matching for Collaborating Across CAD Systems

eCAD-Net: Editable Parametric CAD Models Reconstruction from Dumb B-Rep Models Using Deep Neural Networks

Self-Supervised Representation Learning for CAD

Split-and-Fit: Learning B-Reps via Structure-Aware Voronoi Partitioning

Solid Model Reconstruction from trimmed Surfaces Based on B-ReP Face Sewing

Don't Mesh with Me: Generating Constructive Solid Geometry Instead of Meshes by Fine-Tuning a Code-Generation LLM

PolyGen: An Autoregressive Generative Model of 3D Meshes

3D Image Synthesis Forb-Reps Objects

Vitruvion: A Generative Model of Parametric CAD Sketches

Research on Shape Feature Recognition of Boundary Representation CAD Model

Automatic Generation of B-Rep Model Based on Iges

GenCAD: Image-Conditioned Computer-Aided Design Generation with Transformer-Based Contrastive Representation and Diffusion Priors

Converting Hybrid Wire-Frames to B-rep Models