In the last decade, also due to the onset of Additive Manufacturing technology, the research regarding Computational Geometry has witnessed a substantial development in all areas. The possibility of physically realizing the elaborate design con gurations outputted through CG optimization processes through 3D printing boosted the research in the sector, allowing an exponential growth of the study output. Computational Geometry algorithms are able to process and generate complex topological features through the use of pure computational procedures. When such algorithms are integrated into Design software suites, they guarantee the complete de nition of virtually any spatial structures. Before the advent 3D printing, many of the complex geometries obtained through CG algorithms were not attainable, because of the impossibility to process them through traditional manufacturing techniques. Additive Manufacturing procedures, on the other side, allow the generation of the convoluted shapes obtained by Computational Geometry. The principal aim of the present research is the implementation of a general Computational Geometry-based procedure for weight minimization. The algorithm relies on the iterative generation of a Centroidal Voronoi Tessellation and on a nal hollowing procedure, which delivers the de nitive lightweight topology design. The complete structure generated through such algorithm is a closed-cell foam-like arrangement, which is guaranteed to thoroughly reduce the weight of the nal structure while ensuring the respect of the structural constraints on the object. The solid structure is initially fractured through a Voronoi Tessellation-generating recursive algorithm and divided into cells, which are subsequently hollowed creating the de nitive lightweight structure. The resulting topology provides an optimal solution for the heuristic optimization problem and, considering some minor approximations, for the real system. The general procedure is nally applied to a case study, aimed to improve the design of a Fan Outlet Guide Vane blade stage for a high-bypass turbofan engine by means of a signi cant weight reduction, while at the same time ful lling the structural constraints required for the correct operation of the system. The principal constraint for the analysis is to guarantee the stability of the structure in operating conditions, and therefore the process relies heavily on Finite Element Analysis. The initial simulation results, obtained from the solid system, provide an input for the computational procedure that yields a de nitive geometry, which is then again processed through the FEM to supply a validation for the result.
Negli ultimi anni, l'ascesa dell'Additive Manufacturing e la commercializzazione dei processi di stampa 3D ha provocato un considerevole aumento nell'output di ricerca legato alla Geometria Computazionale. La possibilit a concreta di realizzare le complesse topologie tramite algoritmi di GC ha permesso un aumento sostanziale dell'output di ricerca nel settore. Gli algoritmi di Geometria Computazionale permettono di generare e modi- care una grande variet a di topologie complesse per mezzo di procedure informatiche. Queste procedure, integrate in appropriati software di Design, permettono la de nizione di una molteplicit a di strutture quasi illimitata. Prima dell'avvento delle tecnologie di Additive Manufacturing, le complesse geometrie risultanti da procedure di GC non erano facilmente ottenibili, data la di colt a estrema di costruzione tramite le tecniche manufatturiere tradizionali. Le tecnologie di 3D printing invece permettono la creazione delle pi u disparate forme derivate da tali algoritmi. Lo scopo del seguente studio e l'implementazione di un algoritmo generale ed opensource di GC per la minimizzazione della massa. La procedura e basata sulla computazione iterativa di una Tassellazione Centroidale di Voronoi e su un simultaneo processo di svuotamento. La topologia nale e assimilabile ad un oggetto Lightweight". La struttura nale generata e compatibile con quella di una schiuma a cellule chiuse, che riduce sensibilmente la massa dell'oggetto considerato senza impedire le condizioni di utilizzo dello stesso. La struttura solida viene disgiunta dall'algoritmo ricorsivo in cellule per generare la Tassellazione di Voronoi. Le cellule sono successivamente svuotate creando la struttura Lightweight" nale. Il risultato del processo corrisponde ad una soluzione del problema di ottimizzazione euristico formulato per la computazione e, considerando le sempli cazioni introdotte, per il sistema. L'algoritmo e in ne applicato ad un caso concreto, mirato all'ottimizzazione del design di un Fan Outlet Guide Vane per un motore turbofan ad alto tasso di bypass tramite una signi cativa riduzione di massa, senza in ciare i vincoli strutturali richiesti per la corretta operativit a del pezzo. Il vincolo principale considerato dall'algoritmo e la distribuzione degli stress in condizioni di crociera del velivolo, motivo per cui il processo si basa sull'Analisi ad Elementi Finiti. Il risultato dell'analisi iniziale della struttura solida fornisce l'input per il processo che restituisce la topologia ottimizzata, che a sua volta e sottoposta all'analisi FEM per essere validata.
Structural optimization algorithm for weight reduction of fan outlet guide vane
RASETTI, FRANCESCO
2020/2021
Abstract
In the last decade, also due to the onset of Additive Manufacturing technology, the research regarding Computational Geometry has witnessed a substantial development in all areas. The possibility of physically realizing the elaborate design con gurations outputted through CG optimization processes through 3D printing boosted the research in the sector, allowing an exponential growth of the study output. Computational Geometry algorithms are able to process and generate complex topological features through the use of pure computational procedures. When such algorithms are integrated into Design software suites, they guarantee the complete de nition of virtually any spatial structures. Before the advent 3D printing, many of the complex geometries obtained through CG algorithms were not attainable, because of the impossibility to process them through traditional manufacturing techniques. Additive Manufacturing procedures, on the other side, allow the generation of the convoluted shapes obtained by Computational Geometry. The principal aim of the present research is the implementation of a general Computational Geometry-based procedure for weight minimization. The algorithm relies on the iterative generation of a Centroidal Voronoi Tessellation and on a nal hollowing procedure, which delivers the de nitive lightweight topology design. The complete structure generated through such algorithm is a closed-cell foam-like arrangement, which is guaranteed to thoroughly reduce the weight of the nal structure while ensuring the respect of the structural constraints on the object. The solid structure is initially fractured through a Voronoi Tessellation-generating recursive algorithm and divided into cells, which are subsequently hollowed creating the de nitive lightweight structure. The resulting topology provides an optimal solution for the heuristic optimization problem and, considering some minor approximations, for the real system. The general procedure is nally applied to a case study, aimed to improve the design of a Fan Outlet Guide Vane blade stage for a high-bypass turbofan engine by means of a signi cant weight reduction, while at the same time ful lling the structural constraints required for the correct operation of the system. The principal constraint for the analysis is to guarantee the stability of the structure in operating conditions, and therefore the process relies heavily on Finite Element Analysis. The initial simulation results, obtained from the solid system, provide an input for the computational procedure that yields a de nitive geometry, which is then again processed through the FEM to supply a validation for the result.File | Dimensione | Formato | |
---|---|---|---|
Francesco_Rasetti.pdf
non accessibile
Dimensione
13.27 MB
Formato
Adobe PDF
|
13.27 MB | 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/10589/175980