Journal: Mechanical Systems and Signal Processing

Abbreviation

Mech. syst. signal process.

Publisher

Elsevier

Journal Volumes

ISSN

0888-3270
1096-1216

Description

Filters

Reset filters

Search Results

Publications 1 - 10 of 61
  • Novelty detection by multivariate kernel density estimation and growing neural gas algorithm
    Item type: Journal Article
    Fink, Olga; Zio, Enrico; Weidmann, Ulrich (2015)
    Mechanical Systems and Signal Processing
  • A non-intrusive model-order reduction of geometrically nonlinear structural dynamics using modal derivatives
    Item type: Journal Article
    Karamooz Mahdiabadi, Morteza; Tiso, Paolo; Brandt, Antoine; et al. (2021)
    Mechanical Systems and Signal Processing
    Non-intrusive model-order reduction methods are beneficial for reducing the computational costs of dynamic analysis of nonlinear finite element models, developed in programs that do not release nonlinear element forces and Jacobians (e.g., commercial software). One of the key aspects for developing a displacement-based non-intrusive reduced order model is a proper construction of the reduction basis, which has to be small in size, easy to compute, and must span the subspace in which the full solution lives. In this paper, we propose a non-intrusive model order reduction method based on modal derivatives stemming from a selected set of vibration modes of the linearized system. By definition, modal derivatives do not require the knowledge of the applied load. We name this load-independent basis. The method we propose is also simulation-free, meaning that no nonlinear dynamic simulations of the full model are required to construct the reduction basis. The method is tested with three examples of increasing complexity.
  • On noise covariance estimation for Kalman filter-based damage localization
    Item type: Journal Article
    Wernitz, Stefan; Chatzi, Eleni; Hofmeister, Benedikt; et al. (2022)
    Mechanical Systems and Signal Processing
    In Structural Health Monitoring, Kalman filters can be used as prognosis models, and for damage detection and localization. For a proper functioning, it is necessary to tune these filters with noise covariance matrices for process and measurement noise, which are unknown in practice. Therefore, in the presented work, we apply an autocovariance least-squares method with semidefinite constraints solely based on model parameters. We facilitate this novel approach by formulating the considered innovations covariance function in infinite horizon, which follows inherently from the assumption of linear time-invariant systems. For damage analysis, we adapt a framework based on state-projection estimation errors that was recently established, and yet only applied using H∞ filters. These estimators represent an alternative to Kalman filters, and are considered robust. Because of this property, the necessity of filter tuning is relaxed, and a naive design is often considered. Based on the damage analysis framework, we derive a new damage indicator that features a high sensitivity towards localized damage. We demonstrate the efficacy of the proposed schemes for noise covariance estimation and damage analysis in a series of simulations inspired by a preceding laboratory test. We finally offer experimental validation, based on vibration test data of a cantilever beam featuring damages at multiple positions, where high sensitivity towards small local stiffness changes is achieved. In our investigations, we compare the damage detection and localization performance of Kalman and H∞ filters as well as differences in mode shape curvatures (MSC). In the simulation studies, the proposed Kalman filter-based approach outperforms the alternative strategy using H∞ estimators. The experimental investigations demonstrate a significantly higher sensitivity of the filters towards localized damage compared to differences in MSCs. Considering the totality of investigations, the combined application of both estimators can lead to an increased robustness and sensitivity regarding damage detection and localization.
  • Time-sharing orbit jump and energy harvesting in nonlinear piezoelectric energy harvesters using a synchronous switch circuit
    Item type: Journal Article
    Zhao, Bao; Wang, Jiahua; Hu, Guobiao; et al. (2023)
    Mechanical Systems and Signal Processing
    Nonlinearity has enabled energy harvesting to advance toward higher power output and broader bandwidth in monostable, bistable, and multistable systems. However, operating in the preferable high-energy orbit (HEO) rather than the low-energy orbit (LEO) for making such advancement has restricted their applications. Based on a monostable nonlinear system, this paper proposes a self-contained solution for time-sharing orbit jump and energy harvesting. The joint dynamics of an electromechanical assembly consisting of a nonlinear energy harvester and a switched-mode piezoelectric interface circuit for high-capability energy harvesting is studied. The proposed solution is carried out by utilizing a cutting-edge switched-mode bidirectional energy conversion circuit (BECC), which enables time-sharing dual functions of energy harvesting and vibration exciting. A theoretical model is established based on impedance analysis and multiple time scales method to analyze the stability, frequency response, and phase evolution of the autonomous and nonautonomous nonlinear energy harvesting systems. In particular, the detailed dynamics for the orbit jumps with the vibration exciting mode of BECC are studied. Experiments are performed to validate the full-hysteresis-range orbit jumps with the monostable nonlinear energy harvester. The harvested power after orbit jumps yields a nine-fold increase, compensating for the energy consumption under vibration exciting mode quickly. The proposed solution also refrains the system from extra mechanical or electrical energy sources for orbit jumps, which leads to the first self-contained solution for simultaneous energy harvesting and orbit jump in nonlinear piezoelectric energy harvesting. This work enhances the practical utility of nonlinear energy harvesting technologies toward engineering applications.
  • Full-field structural monitoring using event cameras and physics-informed sparse identification
    Item type: Journal Article
    Lai, Zhilu; Alzugaray, Ignacio; Chli, Margarita; et al. (2020)
    Mechanical Systems and Signal Processing
  • Nonlinear model calibration of a shear wall building using time and frequency data features
    Item type: Journal Article
    Asgarieh, Eliyar; Moaveni, Babak; Barbosa, Andre R.; et al. (2017)
    Mechanical Systems and Signal Processing
  • A composite experimental dynamic substructuring method based on partitioned algorithms and localized Lagrange multipliers
    Item type: Journal Article
    Abbiati, Giuseppe; La Salandra, Vincenzo; Bursi, Oreste S.; et al. (2018)
    Mechanical Systems and Signal Processing
  • Nonlinear model reduction for a cantilevered pipe conveying fluid: A system with asymmetric damping and stiffness matrices
    Item type: Journal Article
    Li, Mingwu; Yan, Hao; Wang, Lin (2023)
    Mechanical Systems and Signal Processing
    We construct reduced-order models (ROMs) for a geometrically nonlinear viscoelastic cantilevered pipe conveying fluid, a dynamical system that includes asymmetric damping and stiffness matrices and nonlinear internal forces. The asymmetries of the damping and stiffness matrices are resulted from flow-induced gyroscopic and follower forces respectively. The ROMs are constructed using spectral submanifolds (SSMs). Specifically, we use a simulation-free approach implemented in SSMTool, an open-source package that automates the computation of SSMs, to construct the SSM-based ROMs. This approach takes equations of motion of the pipe as inputs and no simulation of the full system is involved. The ROMs enable efficient and accurate predictions of nonlinear dynamics of the system, including free vibration, isolas in forced response curves, bifurcations of periodic and quasi-periodic orbits, and heteroclinic and homoclinic orbits.
  • Physics-guided Deep Markov Models for learning nonlinear dynamical systems with uncertainty
    Item type: Journal Article
    Liu, Wei; Lai, Zhilu; Bacsa, Kiran; et al. (2022)
    Mechanical Systems and Signal Processing
    In this paper, we propose a probabilistic physics-guided framework, termed Physics-guided Deep Markov Model (PgDMM). The framework targets the inference of the characteristics and latent structure of nonlinear dynamical systems from measurement data, where exact inference of latent variables is typically intractable. A recently surfaced option pertains to leveraging variational inference to perform approximate inference. In such a scheme, transition and emission functions of the system are parameterized via feed-forward neural networks (deep generative models). However, due to the generalized and highly versatile formulation of neural network functions, the learned latent space often lacks physical interpretation and structured representation. To address this, we bridge physics-based state space models with Deep Markov Models, thus delivering a hybrid modeling framework for unsupervised learning and identification of nonlinear dynamical systems. The proposed framework takes advantage of the expressive power of deep learning, while retaining the driving physics of the dynamical system by imposing physics-driven restrictions on the side of the latent space. We demonstrate the benefits of such a fusion in terms of achieving improved performance on illustrative simulation examples and experimental case studies of nonlinear systems. Our results indicate that the physics-based models involved in the employed transition and emission functions essentially enforce a more structured and physically interpretable latent space, which is essential for enhancing and generalizing the predictive capabilities of deep learning-based models.
  • A structure-preserving machine learning framework for accurate prediction of structural dynamics for systems with isolated nonlinearities
    Item type: Journal Article
    Najera-Flores, David A.; Quinn, D. Dane; Garland, Anthony; et al. (2024)
    Mechanical Systems and Signal Processing
    The nonlinearities present in structural systems are often found in isolated regions within the structure, such as those containing joints or interfaces. However, despite the localized nature of these nonlinearities their presence serves to couple together the modes of the underlying linear system and significantly complicate the development of appropriate reduced-order models; the localized nonlinearities have a global effect on the dynamics of the system. Further, in the presence of evolving structural health the nonlinearities can arise from accumulating damage, with dynamics distinct from those observed in the healthy state. The present work develops a data-driven formulation to identify and include the contributions of the isolated nonlinearities on the dynamics of the underlying linear structure. A novel coordinate separation is developed that decomposes those nonlinearities restricted to the isolated subdomain from the known linear system defined over the entire domain, and the influence of the isolated nonlinearities is reintroduced as an appropriately identified traction at the boundary of the isolated subdomain, referred to as the deviatoric force. In the region exterior to the nonlinear subdomain the response of the ideal linear system recovers that of the original nonlinear system. In this work, the deviatoric force component is predicted using a structure-preserving multilayer perceptron, based only on measured responses at the boundary of the isolated subdomain. Therefore introduction of the perceptron is able to bypass the direct numerical simulation of the nonlinearities within the isolated subdomain. This approach is illustrated through a simple structural system in which an interior region contains cubic nonlinearities and hysteretic damping. Once trained, the machine learning system is able to accurately predict the deviatoric force so that the ideal system recovers the response of the original system in the region outside the isolated nonlinear subdomain. Moreover, the data-driven approach is able to accurately predict the response when the system is subject to differing initial conditions and external excitation without the need for retraining, so that the proposed approach provides a robust description of the structural dynamics of the overall system.
Publications 1 - 10 of 61