Integrating code-based and CAD-based parameterized knowledge rules for multi-fidelity aircraft model generation

This contribution explores improvements of the practical applicability and user accessibility of knowledge-based, multi-fidelity aircraft geometry model generation approaches. A recent implementation of this approach is the Fuselage Geometry Assembler (FUGA), which automates the generation of 3D geometry models by applying knowledge-based rules to aircraft design. Currently, existing FUGA rules for the dynamic generation of geometries are purely code-based, which makes the generation of more complex, high-fidelity geometries challenging. To increase the efficiency of the creation of knowledge rules for the generation of complex geometries and to improve the accessibility for a larger user base, an architecture for the integration of parameterizable CAD models into the knowledge-based engineering process was developed as part of this contribution. Parameterizable CAD models can now flexibly be integrated into FUGA in the form of special knowledge rules and the underlying CAD models can dynamically be parameterized during rule execution. The developed integration is evaluated in a case study on automated generation of structural geometry models. Compared to the existing code-based approach, using automatically parameterizable CAD models allows for the creation of detailed geometries without requiring extensive programming knowledge.