Micro-Electro-Mechanical Systems (MEMS) are widely used in sensing applications where miniaturization and reliability are essential. In case of very severe dynamic loads such as shocks or drops, MEMS devices can experience internal impacts between the proof mass and mechanical stoppers. Accurate modeling of these events is important for predicting the performance and structural integrity of the devices. However, full-order finite element simulations of contact dynamics are computationally expensive, especially for systems with many degrees of freedom. This thesis proposes and validates a reduced-order modeling framework capable of re producing the transient dynamics of MEMS devices undergoing impact and that is also scalable to systems with many degrees of freedom. Several reduction strategies are ex amined, including modal truncation, switching between pre and post impact dynamics, static enrichment, and Component mode synthesis techniques such as Craig-Bampton and Rubin methods. The different approaches are first tested on simplified one- and two-dimensional spring mass systems, in order to be compared in accuracy and scalability. The most promising method is then applied to a three-dimensional FE model of a MEMS accelerometer. In addition, the influence of key physical parameters such as contact stiffness and damp ing is investigated to assess how these factors influence the impact dynamics and the performance of the reduced-order model. The results demonstrate that the static enrichment and Rubin methods provide accurate and scalable tools for simulating impact-driven MEMS dynamics while reducing the computational time by some orders of magnitude.
I sistemi microelettromeccanici (MEMS) sono ampiamente utilizzati nelle applicazioni di rilevamento in cui la miniaturizzazione e l’affidabilita’ sono essenziali. In caso di carichi dinamici molto severi, come urti o cadute, i dispositivi MEMS possono subire impatti in terni tra la massa vibrante e gli elementi di arresto meccanici. Per prevedere le prestazioni e l’integrita’ strutturale dei dispositivi, e’ importante modellare accuratamente questi eventi. Tuttavia le simulazioni agli elementi finiti di ordine completo della dinamica di contatto sono computazionalmente costose, soprattutto per sistemi con molti gradi di liberta’. Questa tesi propone e convalida una tecnica di modellazione di ordine ridotto in grado di riprodurre la dinamica transitoria dei dispositivi MEMS sottoposti ad impatto e che e’ anche scalabile a sistemi con molti gradi di liberta’. Vengono esaminate diverse strategie di riduzione, tra cui il troncamento modale, il passaggio tra dinamica pre e post impatto, l’arricchimento statico e tecniche di sintesi modale per componenti, come i metodi di Craig-Bampton e Rubin. I diversi approcci vengono prima testati su sistemi semplificati mono o bidimensionali formati da masse e molle, per essere confrontati in termini di accuratezza e scalabilita’. Il metodo piu’ promettente viene poi applicato ad un modello agli elementi finiti tridi mensionale di un accelerometro MEMS. Inoltre, viene analizzata l’influenza di alcuni parametri fisici, come la rigidezza di contatto e lo smorzamento, per valutare come questi influenzino la dinamica di impatto e le prestazioni del modello ridotto. I risultati dimostrano che i metodi di arricchimento statico e Rubin forniscono strumenti accurati e scalabili per simulare la dinamica dei MEMS guidata dall’impatto, riducendo anche il tempo computazionale di alcuni ordini di grandezza.
Reduced-order modeling of impact phenomena in MEMS
CLEMENTINI, TOMMASO
2024/2025
Abstract
Micro-Electro-Mechanical Systems (MEMS) are widely used in sensing applications where miniaturization and reliability are essential. In case of very severe dynamic loads such as shocks or drops, MEMS devices can experience internal impacts between the proof mass and mechanical stoppers. Accurate modeling of these events is important for predicting the performance and structural integrity of the devices. However, full-order finite element simulations of contact dynamics are computationally expensive, especially for systems with many degrees of freedom. This thesis proposes and validates a reduced-order modeling framework capable of re producing the transient dynamics of MEMS devices undergoing impact and that is also scalable to systems with many degrees of freedom. Several reduction strategies are ex amined, including modal truncation, switching between pre and post impact dynamics, static enrichment, and Component mode synthesis techniques such as Craig-Bampton and Rubin methods. The different approaches are first tested on simplified one- and two-dimensional spring mass systems, in order to be compared in accuracy and scalability. The most promising method is then applied to a three-dimensional FE model of a MEMS accelerometer. In addition, the influence of key physical parameters such as contact stiffness and damp ing is investigated to assess how these factors influence the impact dynamics and the performance of the reduced-order model. The results demonstrate that the static enrichment and Rubin methods provide accurate and scalable tools for simulating impact-driven MEMS dynamics while reducing the computational time by some orders of magnitude.| File | Dimensione | Formato | |
|---|---|---|---|
|
2025_12_Clementini_Tesi.pdf
non accessibile
Descrizione: testo tesi
Dimensione
22.26 MB
Formato
Adobe PDF
|
22.26 MB | Adobe PDF | Visualizza/Apri |
|
2025_12_Clementini_Executive Summary.pdf
non accessibile
Descrizione: executive summary
Dimensione
1.46 MB
Formato
Adobe PDF
|
1.46 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/10589/246558