Current dynamic substructuring methods as approximations to condensation model reduction

Model order reduction (MOR) techniques have been used in structural dynamics for over sixty years as a quick and effective way to examine the dynamic behavior of complex finite element (FE) models. Various macroscopic phenomena arise from the physics and mechanics of the underlying microstructure, and MOR approaches can be employed in conjunction with the multiscale computational homogenization approach, which is deriving local macroscopic constitutive responses from underlying microstructure; to predict the dynamic behavior of heterogeneous macrostructures. The accuracy and efficiency of the most representative approaches from dynamic condensation, component mode synthesis (CMS), and general model reduction (GMR) methods are compared in this paper. The techniques considered are Guyan and the improved reduced system (IRS) from classical dynamic condensation, Craig-Bampton (CB) and enhanced Craig-Bampton (ECB) from CMS, IRS-based substructuring, GMR, and GMRPlus. All the substructuring methods employ primal assembly at the boundary interface of the substructures. The accuracy and computing efficiency of the techniques are compared in-depth using real-world engineering problems, and the effect of the number of substructures on the accuracy and computational cost is also examined.

Component mode synthesis (CMS) is a widely employed model reduction technique used to reduce the computational cost associated with the dynamic analysis of complex engineering structures. To generate CMS models, specifically the formulation of Craig and Bampton, both normal fixed-interface modes and constraint modes of the component’s structure are calculated. These modes are used in conjunction with the component level mass and stiffness matrices to generate reduced mass and stiffness matrices used in the final analyses. For some component models, the most computationally expensive part of this procedure is calculating the component normal modes information. Several different approaches are utilized to investigate the sensitivity of system level responses to variations in several aspects of the CMS models. One approach evaluates changes due to modifications of the reduced mass and stiffness matrices assuming that the mode shapes do not change. The second approach assumes that the mode shapes change but the reduced mass and stiffness matrices do not change. An example is presented to show the influence of these two approaches.

Uncertainty propagation in dynamic sub-structuring by model reduction integrated domain decomposition

<div class="section abstract"><div class="htmlview paragraph">Increasing demand for simulation accuracy often leads to increased finite element model complexity, which in turn, results in higher computational costs. As a provision, component mode synthesis approaches are employed to approximate the system response by using dynamic substructuring and model reduction techniques in linear systems. However, the use of available model reduction techniques in nonlinear problems containing the contact type of nonlinearities remains an interesting topic. In this paper, the application of a component mode synthesis method in squeak and rattle nonlinear simulation has been investigated. Critical regions for squeak and rattle of the side door model of a passenger car were modelled by nonlinear contact definition in finite element simulation. Craig-Bampton model reduction method was employed to substructure the finite element model while keeping the nonlinear contacts in the model. The model response was evaluated using the modal assurance criterion, frequency response analysis and contact force magnitude in comparison with the baseline model. Results showed that a great reduction in computational time (about 98%) can be achieved while the accuracy of the system response was maintained at an acceptable range for the intended application for squeak and rattle simulation. Although the prediction of impact events in time was done accurately, the contact force magnitude was estimated with average error of 2.5% to 22%, compared with the baseline results. The outcomes of the study show that to empower squeak and rattle prediction by including contact interfaces in finite element simulations, implementation of the model reduction approach can compensate the simulation cost.</div></div>

Application of Wilson-Ritz vectors in dynamic substructuring

Reliable and efficient vibroacoustic loads prediction is often critical in structural design, yet it remains a challenging task for many applications. Spacecraft structures are characterised by extensive use of composite materials, complex connections between components and various non-trivial geometrical features. Accurate treatment necessitates the construction of highly detailed numerical models, traditionally employing deterministic representations. Simultaneously, the broadband acoustic excitation due to the diffuse sound field experienced during launch requires modelling the fluid domain and solving the resulting elasto-acoustic interaction at = multiple frequencies. To alleviate the computational demand implications for large problem sizes, substructuring and reduction techniques for the structural domain are commonplace, component mode synthesis (CMS) being a framework widely adopted in the aerospace industry. Nevertheless, despite ongoing research, the topic still presents a range of difficulties when a universal, robust method of accounting for model uncertainties is sought. In this study, two CMS based approaches are proposed and evaluated. Firstly, the Craig-Bampton stochastic method (CBSM) is improved via a set of modifications enhancing its efficiency, and subsequently adapted for use in a vibroacoustic setting. Optimal perturbation levels and scope of validity of the technique are established against a probabilistic structural analysis (PSA) simulation for a spacecraft structure. Secondly, a novel stochastic finite element method (FEM) is presented. The underlying mathematical foundation is derived so that uncertainty can naturally be controlled at the subsystem level, in partitions of the corresponding condensed mass and stiffness matrices. This decomposition based approach ensures that realisations of the random matrices have key properties such as positive (semi)definiteness strictly preserved, guaranteeing complete robustness. The method is validated with a spacecraft test case, comparing its predictions against PSA, the improved CBSM and experimental data. A coupling scheme with a hierarchical matrix accelerated boundary element method is formulated, resulting in the construction of a complete stochastic vibroacoustic solver.

Correction of stiffness and mass matrices utilizing simulated measured modal data

A dynamic condensation method with free interface substructuring

<div class="htmlview paragraph">With highly complex structures, it is sometimes desirable to derive a dynamic model of the system from experimental vibration data. This paper presents algorithms for synthesizing the mass and stiffness matrices from experimentally derived modal data in a way which preserves the physical significance of the individual mass and stiffness elements. The synthesizing procedures allow for the incorporation of other mass and stiffness data, whether empirical or based on the analyst's insight. The mass and stiffness matrices are derived for a cantilever beam example and are compared with those obtained using earlier techniques.</div>

Many real systems such as moving vehicles on roads, missiles following trajectories and ships in sea water have asymmetric mass, damping and stiffness matrices. Eigen‐sensitivity analysis methods for the symmetric damped system cannot be used in the asymmetric damped case. Therefore, a method for calculating eigenpair sensitivity of the asymmetric damped system is needed. To do this, a modal method employing a modal superposition idea was recently developed. Since the accuracy of the modal method is dependent on the number of modes used in calculation, the modal method needs higher eigenvectors to guarantee the accuracy. In large‐scale systems, however, only a few lower modes are generally considered for the dynamic analysis. Hence, if the modal method is used to obtain the eigen‐sensitivity of the large‐scale system, the significant errors could not be avoided due to the lack of the information of higher modes. In this paper, the modified modal methods for computing the sensitivities of the eigenpairs of asymmetric damped system using a few lowest sets of modes are proposed. Numerical example shows that the proposed methods achieve better calculating efficiency than the previous modal method. Copyright © 2004 John Wiley Sons, Ltd.

Model updating of large-scale structures is difficult to carry out when using a frequency response function (FRF) for damage identification, as the solutions for the global system matrices with too many degrees of freedom are required in each iteration. In this paper, a substructure damage identification method is proposed based on the model updating of the acceleration FRF. The original finite element model is divided into several substructures using the improved reduced system (IRS) by the dynamic condensation method, resulting in the simplified substructure model. The final simplified model is composed of the simplified mass matrix and stiffness matrix of the substructure considered. The damage acceleration FRF to be identified is used to iteratively update the simplified model. The locations and extents of the damage elements are obtained by updating the results, which reduces the number of uncertain parameters to be updated and leads to the rapid convergence of the optimization process. In the iteration, L1 norm regularization is introduced to solve the ill-posed problem, which improves the stability of the identification results. A numerical simulation of a six-story steel frame structure under various working conditions was carried out to verify the effectiveness of the proposed method, which was also validated by the experiments. The robustness and performance of the proposed damage identification method based on substructures have been demonstrated.

BASEBAND METHODS OF COMPONENT MODE SYNTHESIS FOR NON-PROPORTIONALLY DAMPED SYSTEMS

This paper reports a detailed study of modal reduction based on either linear normal mode(LNM) analysis or proper orthogonal decomposition(POD) for modeling a single α-D glucopyranose monomer as well as a chain of monomers. Also a modal reduction method combining POD and component modal synthesis(CMS) is developed. The accuracy and efficiency of these methods are reported. The focus of this study is to determine to what extent these methods can reduce the time and cost of molecular modeling and simultaneously provide the required accuracy. It has been demonstrated that a linear reduced order model(ROM) is valid for small amplitude excitation and low frequency excitation. It is found that a nonlinear ROM based on POD modes provides a good approximation even for large excitation while the nonlinear ROM using linear eigenmodes as the basis vectors is less effective for modeling molecules with a strong nonlinearity. The ROM based on CMS using POD modes for each component also gives a good approximation. With the reduction in the dimension of the system using these methods the computational time and cost can be reduced significantly.

This paper introduces a new reduced order model (ROM) for nonlinear forced response analysis of mistuned bladed disks with localized contact nonlinearities. The developed technique is based on a Component Mode Synthesis (CMS) approach and introduces small frequency mistuning into the final ROM as deviations in the mass matrix. Critical to the development of CMS-based ROMs for mistuned systems, is the construction of a reduction basis that remains invariant to mistuning parameters, ensuring the ROM’s suitability for statistical analyses. Traditionally, CMS-based reduction techniques for small mistuned systems fall into two major groups: 1. System-level approaches: in which mistuned system modeshapes are represented as a linear combination of tuned modes, typically overlooking mistuning impact on static constraint modes. 2. Component-level approaches: in which mistuning is confined to individual components (or superelements), preserving the independence of the reduction basis from mistuning patterns, typically overlooking mistuning impact on interface degrees-of-freedom (DOFs). The proposed technique combines the strengths of both methods. It adopts a system-level approach, utilizing modeshapes of the fully stuck system within the reduction basis. Furthermore, it eliminates common assumptions on static constraint modes by incorporating mistuning effects directly into the mass matrix. As a result, the constraint modes employed in the reduction basis are precise and remain unaffected by mistuning, ensuring their invariance. A comprehensive comparison between ROMs constructed using the new approach and the full model demonstrates the method strengths through modal and forced response analyses. The proposed approach presents a promising solution for accurate and efficient nonlinear dynamic analysis of mistuned bladed disk systems with shroud friction contacts.

This paper proposes a method to identify mass and stiffness matrices of structures. Element mass and stiffness matrices have mass and stiffness distributions. Therefore, a ratio between two elements in these matrices has a proper value. Element mass and stiffness matrices are divided into several groups considering this relation. A group is represented by a sparse matrix (called group matrix) with only elements in the group matrix. Mass and stiffness matrices are modified group by group. The objective function to be minimized is the sum of each norm of difference between analytical and identified group matrices. A numerical example is given to show the effectiveness of the method.