Automatic CAD Parametrization via Minimal Spanning Tree
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current CAD technologies lack efficient methods for automatically generating parameters and parametric constraints, particularly for 3D models, leading to issues with over- or under-constraining, which can result in unresolvable design requirements and inefficient editing processes.
Innovation Solution
A computer-implemented method for automatic parametrization of CAD models that determines a minimal spanning subset of pairwise numerical constraints, allowing for the selection of a convenient parameter set and corresponding parametric constraints without prior assumptions, ensuring the CAD model maintains its original geometric properties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a distance parameter and parametric constraint are introduced for each numerical distance value between pairs of faces, then the CAD model becomes fully constrained, but this leads to over-constraining the model
Solution Approach 1:
The patent extracts only the necessary constraints from the complete set of pairwise distance constraints. By using graph theory to identify a spanning tree of the constraint graph, the system selects a minimal subset of constraints that maintains connectivity between all faces without including redundant constraints. This extraction process reduces the number of parameters and constraints from N(N-1)/2 possible constraints to exactly N-1 constraints, eliminating over-constraining while preserving model reliability.
Solution Approach 2:
The patent segments the complete constraint system into a minimal spanning subset. The constraint graph is divided into essential constraints (those forming the spanning tree) and redundant constraints (those creating cycles). By segmenting the constraint set this way, the system maintains only the essential constraints needed for full connectivity, reducing complexity while preserving the ability to maintain all relative distances between faces.
2Device complexity
If random selection of numerical distance values is used to create parameters and constraints, then the number of constraints is reduced, but this may split faces into subsets without proper constraints between subsets
Solution Approach 1:
The patent performs preliminary graph analysis before selecting constraints. It first constructs the complete constraint graph with all faces as nodes and all pairwise distance constraints as edges. Then, using graph theory algorithms, it identifies a spanning tree that guarantees connectivity between all faces. This preliminary structural analysis ensures that the selected constraint subset maintains proper connectivity without requiring random selection, thus avoiding division into disconnected subsets.
3Ease of operation
If automatic generation of parameters and constraints is implemented, then user effort is reduced, but ensuring geometric properties are preserved becomes more complex
Solution Approach 1:
The patent implements self-service through automatic constraint generation based on the existing geometric data. The system analyzes the CAD model's face geometry and automatically generates appropriate distance constraints without requiring user input. By using the actual measured distances between faces as the basis for constraints, the system self-adjusts to preserve the original geometric properties while reducing manual user effort in the constraint creation process.
Data Source
AI summary
A CAD model comprises a set of at least three subentities. A candidate set of pairwise numerical constraints is obtained for the set of subentities, such that a first graph, representing the subentities of the set of subentities as nodes and the pairwise numerical constraints of the candidate set as edges, is connected. A minimal spanning subset of pairwise numerical constraints is obtained from the candidate set, such that a second graph, representing the subentities of the set of subentities as nodes and the pairwise numerical constraints of the minimal spanning subset as edges, is a spanning tree. A parameter set to parameterize the pairwise numerical constraints of the minimal spanning subset is determined. The parameter set and parametric constraints, based on the parameter set and the numerical values of the numerical constraints of the minimal spanning subset, are added to the CAD model.