In aerospace structural design, recent manufacturing advancements have enabled increasingly complex stiffening layouts beyond conventional ones, expanding the design space and requiring dedicated design methodologies. In parallel, bio-inspired approaches have gained relevance, as natural systems exhibit efficient architectures shaped by evolutionary processes. Among these processes, multi-scale architectures allow interacting reinforcement levels, promoting effective stiffness redistribution and improved structural response. This thesis proposes a framework for designing multi-scale stiffened panels for instability-driven applications through a two-step procedure. A problem-oriented genetic algorithm optimises the primary stiffening layout, while a structural-to-geometric mapping generates a sub-reinforcement pattern based on the mechanical response of the primary structure. The mapping relies on anisotropic centroidal Voronoi tessellations, whose mathematical formulation allows a consistent translation of structural information into geometric features; similar patterns are common in nature and associated with mechanically efficient configurations. To ensure computational efficiency and simple integration within the procedure, structural evaluations are performed through a semi-analytical Ritz-based formulation. Although the two scales are generated sequentially, rather than through a fully coupled procedure, the framework enables the systematic development of structurally effective multi-scale configurations and can be extended to other design problems.
Nel campo della progettazione aerospaziale i recenti progressi produttivi hanno consentito lo sviluppo di irrigidimenti più complessi rispetto a quelli tradizionali, ampliando lo spazio di progetto e richiedendo strategie mirate. Parallelamente, i concetti bio-mimetici hanno acquisito importanza poiché la natura spesso presenta soluzioni efficienti esito di processi evolutivi. Tra queste, le architetture multi-scala sono caratterizzate dall’interazione tra diversi livelli di rinforzo favorendo una redistribuzione efficiente della rigidezza strutturale. Questa tesi propone un metodo, articolato in due fasi, per progettare pannelli irrigiditi multi-scala in applicazioni governate da fenomeni di instabilità. Un algoritmo genetico ad hoc ottimizza la posa dei rinforzi primari, mentre una mappatura strutturale-geometrica genera i rinforzi secondari a partire dalla risposta meccanica della struttura primaria. La mappatura impiega le tassellazioni di Voronoi centroidali anisotrope, la cui definizione matematica consente una traduzione coerente delle informazioni strutturali in caratteristiche geometriche; configurazioni simili sono osservabili in natura e associate a soluzioni meccanicamente efficienti. Per ridurre l'onere computazionale e garantire una agevole integrazione nel processo, la risposta strutturale è valutata mediante formulazione semi-analitica basata sul metodo di Ritz. Sebbene i due livelli siano generati in sequenza, e non in modo accoppiato, il metodo consente la generazione sistematica di pannelli multi-scala efficaci ed è estendibile ad altri problemi strutturali.
Bio-inspired design of a multi-scale stiffened panel
Agari, Ermal
2024/2025
Abstract
In aerospace structural design, recent manufacturing advancements have enabled increasingly complex stiffening layouts beyond conventional ones, expanding the design space and requiring dedicated design methodologies. In parallel, bio-inspired approaches have gained relevance, as natural systems exhibit efficient architectures shaped by evolutionary processes. Among these processes, multi-scale architectures allow interacting reinforcement levels, promoting effective stiffness redistribution and improved structural response. This thesis proposes a framework for designing multi-scale stiffened panels for instability-driven applications through a two-step procedure. A problem-oriented genetic algorithm optimises the primary stiffening layout, while a structural-to-geometric mapping generates a sub-reinforcement pattern based on the mechanical response of the primary structure. The mapping relies on anisotropic centroidal Voronoi tessellations, whose mathematical formulation allows a consistent translation of structural information into geometric features; similar patterns are common in nature and associated with mechanically efficient configurations. To ensure computational efficiency and simple integration within the procedure, structural evaluations are performed through a semi-analytical Ritz-based formulation. Although the two scales are generated sequentially, rather than through a fully coupled procedure, the framework enables the systematic development of structurally effective multi-scale configurations and can be extended to other design problems.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_03_Agari_Tesi.pdf
accessibile in internet solo dagli utenti autorizzati
Dimensione
21.31 MB
Formato
Adobe PDF
|
21.31 MB | Adobe PDF | Visualizza/Apri |
|
2026_03_Agari_Executive_Summary.pdf
accessibile in internet solo dagli utenti autorizzati
Dimensione
996.27 kB
Formato
Adobe PDF
|
996.27 kB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/251037