A method for designing filament-wound composite frame structures using a data-driven evolutionary optimisation algorithm EvoDN2

ABSTRACT

A methodology of optimising composite frame structures has been applied to three selected geometries. A cyclic process driven by predefined objectives achieved the most desirable parameters through adjustments in winding angles and tube ply thicknesses. A geometry resembling a crane structure underwent initial analyses, allowing for the determination of appropriate settings for the surrogate model’s training phase, considering accuracy and computational time. Its final design was influenced by prevalent bending and tension loads, resulting in near-zero winding angles and a range of thicknesses that met displacement, strength, and weight requirements. A second geometry with further restrictions was also considered. Finally, for a third geometry, winding angles were tailored to accommodate torsion forces. The presented optimisation process resulted in volume reduction while maintaining displacement and strength parameters. These findings highlight the effectiveness and transferability of the optimisation approach across different geometries.

KEYWORDS:

• Laminate
• tube
• optimisation
• evolutionary
• deep learning

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study received support from the Grant Agency of the Czech Technical University in Prague [grant number SGS21/151/OHK2/3T/12].