BUILD SIMULATION GRAPH (ONTOLOGY)¶
Builds a comprehensive knowledge graph from a completed LS-DYNA simulation, capturing parts, materials, contacts, loads, sensors, results, and responses as typed nodes and edges. The graph can be serialized to JSON and loaded into an ARC2 ontology store for AI agent traversal and semantic querying.
When to use¶
Tagged: arc2, binout, contacts, d3hsp, graph_builder, knowledge_graph, lsdyna, materials.
Inputs¶
| Label | ID | Type | Default | Required | Description |
|---|---|---|---|---|---|
| Simulation | simulation_id | remote_lookup | — | ✓ | Platform simulation record to graph-build from; must reference a completed LS-DYNA run whose d3hsp and binout files are accessible. |
| Response Template (Optional) | response_template_id | remote_lookup | — | Optional response template whose applied KPI responses are added as result nodes in the graph; omit when no pre-defined template is available. |
Outputs¶
| Label | ID | Type | Description |
|---|---|---|---|
| Status | status | string | Human-readable summary string reporting the simulation name, total node count, edge count, and any non-critical warnings encountered during graph construction. |
| Simulation Graph | graph | graph | In-memory graph object containing all typed nodes (parts, materials, contacts, sensors, results, etc.) and directed edges representing their relationships; consumed by downstream ontology or agent workers. |
| Simulation Graph File | graph_file | file | Downloadable JSON file of the full simulation graph written to the platform temp directory; path pattern is simulation_{id}_graph.json. |
| Node Count | node_count | scalar | Integer count of all graph nodes generated; useful for verifying graph completeness or comparing across simulation variants. |
| Edge Count | edge_count | scalar | Integer count of all directed edges generated; indicates the richness of cross-entity relationships captured in the graph. |
| Contact Interaction Matrix | contact_interaction_matrix | dataset | Tabular dataset (part × part) quantifying contact forces between interacting parts, derived from binout contact-force extraction; empty when no binout data is available. |
Disciplines¶
- ai_ml.agents
- ai_ml.domain_reasoning
- cae.postprocessing.extraction
- cae.postprocessing.response
- engineering.crash.dynamics
- platform.ontology
Auto-generated from platform schema. Worker id: lsdyna_build_simulation_graph. Schema hash: 2c120d0c88ac. Hand-curated docs in workerexamples/ override this page when present.