A high-order Virtual Element Method (p-VEM) framework is developed for the geometrically nonlinear analysis of curvilinearly stiffened variable stiffness panels. The proposed formulation allows the use of arbitrary polygonal elements with curved edges, simplifying mesh generation for complex curved geometries. However, the standard VEM requires a stabilization term, which is arbitrary and can reduce accuracy if poorly chosen. To address this issue, self-stabilized formulations are adopted in this work, eliminating the need for heuristic stabilization. Furthermore, the presence of variable coefficients in the bilinear form, unavoidable in variable stiffness panels, requires special care in standard VEM settings. Thus, a novel Variable Coefficients-VEM (VC-VEM) approach is proposed in this thesis. The VEM formulation is applied to the linear case and then is extended to geometrically nonlinear analyses by replacing nonlinear terms with computable approximations. The thesis is organized into two main contributions. The first is devoted to the benchmarking of several stabilized and self-stabilized formulations, as well as VC-VEM, across a range of PDEs. In particular, the proposed VC-VEM demonstrates improved robustness and efficiency compared to standard VEM. The second part applies the developed formulation to the nonlinear analysis of variable stiffness structures. The method proves to be very efficient in the presence of high-gradient solutions and in dealing with complex geometries with curvilinear stiffeners. Moreover, it maintains its accuracy in the nonlinear regime despite the introduced approximations. Overall, the proposed approach ensures both flexibility and efficiency in the analysis of innovative variable stiffness structures in the nonlinear regime, demonstrating that the VEM is a valuable alternative to classical methods.
In questa tesi si propone una formulazione del Metodo agli Elementi Virtuali (VEM) di ordine elevato per l’analisi non lineare di pannelli a rigidezza variabile irrigiditi da correnti curvilinei. La formulazione consente l’uso di elementi poligonali arbitrari con bordi curvi, semplificando la generazione della discretizzazione per geometrie complesse e curve. Tuttavia, il VEM richiede l’uso di un termine di stabilizzazione, la cui scelta e definizione non è univoca e, se inadeguata, può ridurre l’accuratezza del metodo. Per superare questo problema, si propone l’uso di formulazioni auto-stabilizzanti, eliminando così la necessità di stabilizzazioni euristiche. Inoltre, la presenza di coefficienti variabili nella forma bilineare, inevitabile nei pannelli a rigidezza variabile, richiede particolare attenzione nelle formulazioni VEM standard. Per affrontare questa criticità, viene proposto un nuovo approccio denominato Coefficienti Variabili-VEM (VC-VEM). La formulazione VEM è applicata ad analisi lineari e, successivamente, viene estesa anche ad analisi non lineari, sostituendo i termini non lineari con approssimazioni calcolabili. La tesi si articola in due linee principali. La prima riguarda lo studio dettagliato delle diverse formulazioni stabilizzate e auto-stabilizzanti, nonché del VC-VEM, su una gamma di Equazioni alle Derivate Parziali. In particolare, il VC-VEM dimostra una robustezza ed efficienza superiore rispetto al VEM standard. La seconda linea riguarda l’applicazione della formulazione sviluppata all’analisi non lineare di strutture a rigidezza variabile. Il metodo si dimostra molto efficiente nella gestione di soluzioni con forti gradienti e nella gestione di geometrie complesse con correnti curvilinei. Si dimostra inoltre la possibilità di applicare con successo il metodo proposto in ambito non lineare, nonostante le approssimazioni introdotte. Complessivamente, l’approccio proposto garantisce sia flessibilità sia efficienza nell’analisi di strutture innovative a rigidezza variabile nel regime non lineare, dimostrando che il VEM è una valida alternativa ai metodi classici.
Novel p-virtual element method for the nonlinear analysis of curvilinearly stiffened panels
Foligno, Paola Pia
2025/2026
Abstract
A high-order Virtual Element Method (p-VEM) framework is developed for the geometrically nonlinear analysis of curvilinearly stiffened variable stiffness panels. The proposed formulation allows the use of arbitrary polygonal elements with curved edges, simplifying mesh generation for complex curved geometries. However, the standard VEM requires a stabilization term, which is arbitrary and can reduce accuracy if poorly chosen. To address this issue, self-stabilized formulations are adopted in this work, eliminating the need for heuristic stabilization. Furthermore, the presence of variable coefficients in the bilinear form, unavoidable in variable stiffness panels, requires special care in standard VEM settings. Thus, a novel Variable Coefficients-VEM (VC-VEM) approach is proposed in this thesis. The VEM formulation is applied to the linear case and then is extended to geometrically nonlinear analyses by replacing nonlinear terms with computable approximations. The thesis is organized into two main contributions. The first is devoted to the benchmarking of several stabilized and self-stabilized formulations, as well as VC-VEM, across a range of PDEs. In particular, the proposed VC-VEM demonstrates improved robustness and efficiency compared to standard VEM. The second part applies the developed formulation to the nonlinear analysis of variable stiffness structures. The method proves to be very efficient in the presence of high-gradient solutions and in dealing with complex geometries with curvilinear stiffeners. Moreover, it maintains its accuracy in the nonlinear regime despite the introduced approximations. Overall, the proposed approach ensures both flexibility and efficiency in the analysis of innovative variable stiffness structures in the nonlinear regime, demonstrating that the VEM is a valuable alternative to classical methods.| File | Dimensione | Formato | |
|---|---|---|---|
|
PhD_Thesis_Foligno.pdf
accessibile in internet per tutti
Descrizione: PhD Thesis Foligno
Dimensione
84.79 MB
Formato
Adobe PDF
|
84.79 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/10589/249758