In recent years, the trend toward electric mobility has increased the focus on the NVH (Noise Vibrations and Harshness) characteristics of vehicles in the automotive industry. Virtual Prototype Assembly (VPA) techniques are being increasingly adopted. These models consist of active and passive components coupled in a virtual environment to reconstruct complex assemblies and enable the prediction of their NVH performance at any stage of the design phase. Proper characterization of the components is required. In particular, active components (source) are described by their blocked forces, which are estimated using component-based TPA techniques and dependent solely on the dynamics of the source. The thesis aims to improve the reliability of in situ blocked forces estimation by optimizing the receiver of the test setup and the sensor arrangement on it. The accuracy depends on the sensors’ ability to observe the source-receiver interface, which directly translates into the conditioning of the admittance matrix of the coupled system. The target component of the specific application is a steering rack with Electric Power Steering (EPS) system, whose actuator is and internal gearing are responsible for generating noise and vibrations. As is common in modern-day NVH problems, the frequency range of interest is very broad, in the order of kHz. To date, all approaches regarding the optimal sensor placement are limited to applications characterized by much narrower frequency ranges and focused on very few structural modes. Therefore, this thesis proposes a workflow suitable for the target application, featuring an integrated tool for identifying the optimal layout of the sensors. Although the geometry of the receiver has an impact on the conditioning of the matrix, the outcome of this study highlights that sensor layout optimization has much greater influence. Therefore, other design constraints can be taken into account, such as the manufacturing cost or the versatility of the component for research purposes. Promising results are achieved by implementing a Tuned Mass Damper (TMD) on the receiver, aiming to reduce isolated peaks in the condition number curve at specific critical frequencies, thus laying the foundations for further research.
Negli ultimi anni, la crescente diffusione della mobilità elettrica ha aumentato l’interesse verso le caratteristiche NVH (Noise, Vibrations and Harshness) nell’industria automo tive. Le tecniche di Virtual Prototype Assembly (VPA) vengono sempre più adottate: questi modelli sono costituiti da componenti attivi e passivi accoppiati in un ambiente virtuale, con l’obiettivo di ricostruire complessi assiemi e consentire la previsione delle loro prestazioni NVH in qualsiasi fase della progettazione. È necessaria una caratterizzazione accurata dei componenti. In particolare, i componenti attivi (sorgenti) sono descritti dalle loro blocked forces, stimate tramite tecniche di Component-Based TPA e dipendenti esclusivamente dalla dinamica del componente sorgente. La tesi mira a migliorare la stima delle in situ blocked forces ottimizzando sia il ricevitore del test setup che la disposizione dei sensori. L’accuratezza dipende dalla capacità dei sen sori di osservare l’interfaccia sorgente-ricevitore, che si traduce nel condizionamento della matrice di ammettenza del sistema accoppiato. Il componente target dell’applicazione specifica è uno steering rack equipaggiato con un sistema di Electric Power Steering (EPS), responsabile della generazione di rumore e vibrazioni. Come tipico dei problemi NVH, l’intervallo di frequenze di interesse è molto ampio, nell’ordine dei kHz. Ad oggi, gli approcci relativi all’ottimizzazione del posizionamento dei sensori sono limitati ad appli cazioni caratterizzate da intervalli di frequenza molto più ristretti e focalizzati su pochi modi strutturali. Perciò, questa tesi propone un workflow adatto all’applicazione tar get, corredato da un’algoritmo per l’identificazione della disposizione ottimale dei sensori. Sebbene la geometria del componente ricevitore abbia un impatto sul condizionamento della matrice, i risultati di questo studio evidenziano che l’ottimizzazione del layout dei sensori ha un’influenza molto maggiore. Pertanto, è possibile considerare ulteriori vincoli progettuali, come il costo di produzione o la versatilità del componente ricevitore. Risultati promettenti sono stati ottenuti inserendo un Tuned Mass Damper (TMD) sul ricevitore. Tale indagine è stata effettuata con l’obiettivo di ridurre eventuali picchi isolati nella curva che descrive il numero di condizionamento in frequenza. Un eventuale impiego su più larga scala necessita, tuttavia, ulteriori approfondimenti.
A methodology for designing the receiving structure and sensor layout for component-based TPA in virtual prototype assembly
PALMUCCI, STEFANO
2024/2025
Abstract
In recent years, the trend toward electric mobility has increased the focus on the NVH (Noise Vibrations and Harshness) characteristics of vehicles in the automotive industry. Virtual Prototype Assembly (VPA) techniques are being increasingly adopted. These models consist of active and passive components coupled in a virtual environment to reconstruct complex assemblies and enable the prediction of their NVH performance at any stage of the design phase. Proper characterization of the components is required. In particular, active components (source) are described by their blocked forces, which are estimated using component-based TPA techniques and dependent solely on the dynamics of the source. The thesis aims to improve the reliability of in situ blocked forces estimation by optimizing the receiver of the test setup and the sensor arrangement on it. The accuracy depends on the sensors’ ability to observe the source-receiver interface, which directly translates into the conditioning of the admittance matrix of the coupled system. The target component of the specific application is a steering rack with Electric Power Steering (EPS) system, whose actuator is and internal gearing are responsible for generating noise and vibrations. As is common in modern-day NVH problems, the frequency range of interest is very broad, in the order of kHz. To date, all approaches regarding the optimal sensor placement are limited to applications characterized by much narrower frequency ranges and focused on very few structural modes. Therefore, this thesis proposes a workflow suitable for the target application, featuring an integrated tool for identifying the optimal layout of the sensors. Although the geometry of the receiver has an impact on the conditioning of the matrix, the outcome of this study highlights that sensor layout optimization has much greater influence. Therefore, other design constraints can be taken into account, such as the manufacturing cost or the versatility of the component for research purposes. Promising results are achieved by implementing a Tuned Mass Damper (TMD) on the receiver, aiming to reduce isolated peaks in the condition number curve at specific critical frequencies, thus laying the foundations for further research.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_03_Palmucci_Executive Summary.pdf
accessibile in internet per tutti a partire dal 28/02/2029
Descrizione: executive summary
Dimensione
6.06 MB
Formato
Adobe PDF
|
6.06 MB | Adobe PDF | Visualizza/Apri |
|
2026_03_Palmucci_Tesi.pdf
accessibile in internet per tutti a partire dal 28/02/2029
Descrizione: Testo della tesi
Dimensione
59.42 MB
Formato
Adobe PDF
|
59.42 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/10589/253292