1,720,993 research outputs found
Investigation of computational and accuracy issues in POD-based reduced order modeling of dynamic structural systems
No full textIn this paper, we investigate the performance of reduced order modeling of dynamic structural systems
based on the proper orthogonal decomposition (POD) technique. Singular value decomposition of the socalled
snapshot matrix is adopted to generate the reduced space, onto which the system equations of
motion are projected to speedup the computations.
To get insights into the achievable speedup and the capability of POD to provide an input-independent
reduced model, we consider the 39-story Pirelli tower in Milan-Italy. First, we assume that a shear model
of the building is excited by the May 18-1940, Mw 7.1, El Centro earthquake, and generate the data
ensemble necessary to build the reduced model. Second, we assess the local and global accuracies of
the same reduced model in tracking the dynamics of the building, if excited by the May 6-1976, Mw
6.4, Friuli earthquake and by the January 17-1995, Mw 6.8, Kobe earthquake, which differ from the El Centro
one in terms of excited vibration frequencies. We show that POD allows to attain a speedup approaching
250, when the reduced order model is asked to feature a high accuracy; moreover, POD tends to
outperform a standard modal analysis at increasing number of modes retained in the model
Dual estimation of partially observed nonlinear structural systems: a particle filter approach
No full textDual estimation consists of tracking the whole state of partially observed systems, and simultaneously
estimating unknown model parameters. In case of nonlinearly evolving systems, standard filtering procedures
may provide unreliable model calibrations, either because of estimates affected by bias or due
to diverging filter response. In this paper, we propose a particle filter (PF) wherein particles, i.e. system
realizations evolving in a stochastic frame, are first sampled from the current probability density function
of the system and then moved towards the region of high probability by an extended Kalman filter. We
show that the proposed filter works much better than a standard PF, in terms of accuracy of the estimates
and of computing time
Towards real-time health monitoring of structural systems via recursive Bayesian filtering and reduced order modelling
Full text linkA method for the structural health monitoring (SHM) of compliant,
thin plates is discussed. With a specific focus on lightweight composite
structures, a proposal for the optimal deployment of a network of
surface-mounted inertial micro-sensors (MEMS) is reviewed. Allowing for the
measurements gathered through the sensor network as (partial) observations
of the structural state, a hybrid Kalman-particle filtering scheme is adopted to
track the response of the plate to the external excitations and simultaneously
identify unknown model parameters, among which damage or integrity
indices. To move towards a real-time SHM procedure, the mentioned tracking
and identification tasks are performed on a reduced-order model of the
structure, continuously tuned after damage inception by a further Kalman
filter. Results are reported for the exemplary case of a square plate, simply
supported along its boundary, loaded by a concentrated force at its centre and
developing a uniform damage in regions of its mid-plane area
POD-based reduced order modeling of dynamic systems
No full textIn this paper, we study reduced order modeling of dynamic structural systems through the proper orthogonal decomposition (POD) technique. Principal component analysis is adopted to generate the reduced space, where the equations of motion of the system are projected onto. Numerical results concerning the Pirelli Tower in Milan, excited by an earthquake, are shown to get insights into the capability and efficiency of the proposed methodology
Particle and sigma-point Kalman filters: a comparison of performance
No full textAiming to track the full state of a nonlinearly evolving structural system and simultaneously calibrate its constitutive model, in this paper we compare the performances of the sigma-point Kalman filter (S-PKF), of a standard particle filter (PF), and of an extended Kalman-particle filter (EK-PF). To ensure that structural responses do not affect the filters’ performances, we focus on a single degree-of-freedom (DOF) system and show that the three filters are all capable to track possible failure mechanisms. As far as model calibration is concerned, the newly proposed EK-PF performs better than the other two filters, providing unbiased parameter estimates even when the system is diverging because of a failure event
Online damage detection in structural systems via dynamic inverse analysis: A recursive Bayesian approach
Full text linkIn this paper, a framework is presented for the joint state tracking and parameter estimation of partially observed structural systems characterized by a relatively large number of degrees of freedom. To pursue this aim in real-time, the order of the system is reduced via an optimal set of bases, or proper orthogonal modes (POMs) obtained through proper orthogonal decomposition. Since the aforementioned POMs are sensitive to damage, which is defined as a change in the stiffness of the structural model, the variation in the characteristics of the POMs themselves is also tracked online. Taking advantage of the linear relationship between the observation process and the components of the POMs, a solution to the whole problem is obtained with an extended Kalman filter or a hybrid extended Kalman particle filter for the joint tracking-estimation purposes, and with a further Kalman filter for the model update purposes. The efficiency of the proposed method is assessed through simulated experiments on a 8-story shear type building
Stochastic system identification using Kalman filtering
No full textSimultaneous state tracking and calibration of constitutive laws for stochastic structural systems is usually pursued via the extended Kalman filter. However, in the presence of severe nonlinearities due to damage inception and growth, filtering may become unstable. To improve outcomes, a statistical linearization of the system equations has been recently adopted within the sigma-point Kalman filtering approach.
In this study we compare the performances of the extended and sigma-point Kalman
filters, and show that the latter one is superior in calibrating softening materials laws
On-chip testing: A miniaturized lab to assess sub-micron uncertainties in polysilicon MEMS
No full textAn increasing impact of micromechanically governed uncertainties is nowadays foreseen due to the trend of progressively reducing the footprint of MEMS (Microelectromechanical Systems) devices. For polysilicon MEMS, the two major sources of uncertainties, as resulting from the microfabrication process, are linked to the polycrystalline morphology and to the etching. In this review, we summarize some of our recent results related to the statistical assessment of the aforementioned sources, on the basis of experimental data acquired via an on-chip testing device specifically designed to enhance such effects. Through standard electrostatic actuation and readout, the scattering in the response of a series of nominally identical cantilever structures is analyzed to determine characteristic features of etching defects, and of the overall stiffness of the polysilicon film constituting the movable parts of the tested devices
SHM and Efficient Strategies for Reduced-Order Modeling
Full text linkWithin model-based approaches to structural health monitoring (SHM), numerical simulations must be tailored to continuously adapt to the degradation processes and to the possibly changing environment. This model update stage of the analysis brings two competing requirements: the accuracy of the model, with a more detailed description of the phenomena required where damage is supposed to take place; the efficiency of the model, to reduce the overall computational burden and allow for real-time (or close to real-time) computing. Without resorting to AI-based strategies, approaches solely based on proper orthogonal decomposition (POD) and domain decomposition (DD) techniques proved rather efficient in handling the aforementioned trade-off between the diverging requirements of accuracy and efficiency. In this work, we discuss a further improvement over our recently proposed methodology that consists of: a DD of the entire structure into sub-regions, which can be designed to decouple regions more prone to get damaged from regions that are instead less affected by the degradation processes; a POD-based selective model order reduction for all the domains, with adjustable and heterogeneous accuracy requirements. The approach is assessed through an illustrative example related to beam dynamics, with results provided in terms of both accuracy and computational efficiency, or speedup with respect to the full-order model
Online damage detection in plates via vibration measurements
No full textIn this work, we propose a new framework for the online detection of damage in plates via vibration measurements. To this end, a finite element model of the plate is handled by a recursive Bayesian filter for simultaneous state and parameter estimation. To drastically reduce the computational costs and enhance the robustness of the filter, such model is projected onto a (sub-)space spanned by a few vibration modes only, which are provided by a snapshot-based proper orthogonal decomposition (POD) method. A challenge in using such approach for damaging structures stems from the fact that vibration modes can be adjusted only during the training stage of the analysis; if damage occurs or grows when the reduced-order model is at work, the training stage has to be re-started. Here, an alternate method is proposed to concurrently update the sub-space spanned by the modes and to provide estimates of damage location and amplitude. The robustness and accuracy of the proposed approach are ascertained through an ad-hoc pseudo-experimental campaign
- …
