Editable Watertight B-Rep Conversion for Generative CAD Geometry
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Generative design solvers convert input solids to volumetric representations, leading to surface precision loss, especially near sharp edges and holes, causing part interferences and manufacturing defects due to inaccurate surface geometry representation.
Innovation Solution
A fully automatic method converts generative designs into editable, watertight B-Rep models by embedding exact input solid boundary surfaces and approximating other boundaries with globally smooth, editable surfaces, using T-Spline or T-NURCC surfaces to ensure accurate representation and prevent interferences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If generative design solvers convert input solids to volumetric representations for efficient computation, then productivity and ease of operation are improved, but manufacturing precision and surface accuracy deteriorate
Solution Approach 1:
The mesh surface is segmented into multiple patches, with each patch converted to a separate NURBS surface. This segmentation allows precise representation of different surface regions while maintaining computational efficiency during the generative design process.
Solution Approach 2:
A mesh-to-NURBS conversion process serves as an intermediary step between the volumetric representation (level set) and the final B-Rep model. This intermediary converts the approximate mesh surface into precise NURBS surfaces that maintain manufacturing accuracy.
2Ease of operation
If B-Reps are sampled and replaced with volumetric representations for generative design computation, then ease of operation is improved, but loss of information regarding surface precision occurs
Solution Approach 1:
The input solid B-Rep is copied and converted into a level set volumetric representation for generative design computation. After the generative process completes, the original B-Rep information is recovered and combined with the generated geometry to preserve surface precision.
Solution Approach 2:
The representation parameters are changed dynamically: the model uses volumetric level set parameters during generative computation for ease of operation, then transitions to precise B-Rep surface parameters for the final output, thereby preventing information loss.
3Stability of the object's composition
If general surface reconstruction is used to create watertight B-Rep from generative mesh, then completeness of the model is improved, but manufacturing precision deteriorates due to surface inaccuracies
Solution Approach 1:
The patent uses NURBS (Non-Uniform Rational B-Splines) surfaces, which are mathematically precise curved surfaces, to represent the generative design output. This replaces approximate mesh surfaces with exact mathematical representations that maintain manufacturing precision while achieving watertight completeness.
Solution Approach 2:
The surface representation parameters are changed from approximate mesh geometry to precise NURBS mathematical surfaces. This parameter transformation maintains the watertight completeness of the model while significantly improving surface accuracy for manufacturing.
Data Source
Figure 1
Figure 2~3
Figure 4~6
AI summary
Methods, systems, and apparatus, including medium-encoded computer program products, for computer aided design of physical structures include, in at least one aspect, a fully automatic method of converting a generative design into an editable, watertight B-Rep by leveraging the generative solver input and representation to: (1) embed the exact input solid boundary surfaces where the design coincides with the input, (2) approximate everywhere else the design boundary with globally smooth, editable "organic" surfaces, and (3) join all surfaces to form a generative design output B-Rep.