| 1. |
- Abrahamsson, Thomas, 1956, et al.
(författare)
-
Calibration and Validation of a Car Subframe Finite Element Model Using Frequency Responses
- 2015
-
Ingår i: Conference Proceedings of the Society for Experimental Mechanics Series. - Cham : Springer International Publishing. - 2191-5644 .- 2191-5652. - 9783319152516 ; 10, s. 9-22
-
Konferensbidrag (refereegranskat)abstract
- A finite element model of a car front subframe has been calibrated against test data. Stepped-sine testing has been used to give frequency response function estimates on an ensemble of seemingly identical subframes. Therefore, the deviation between test data and simulation results can be compared in a meaningful way by the outcome of model calibration and cross-validation. Emphasis has been put on the preparation of the test pieces for high fidelity testing and on bettering the chances of getting a calibration outcome that provides insight into the physical processes that govern the subframe dynamics. The front subframe model has more than 200,000 degrees-of-freedom and 17 model calibration parameters. The efficiency of the calibration procedure under these conditions is reported. To achieve efficiency, a calibration with a smooth deviation metric is used together with a damping equalization method that eliminates the need for matching of experimental and analytical eigenmodes. The method is combined with surrogate model frequency response evaluation based on model reduction for increased speed. The Matlab based open-domain software tool FEMcali that employs the Levenberg-Marquardt minimizer with randomized starts has been used for calibration and an unregularized Gauss-Newton minimizer has been used in the cross-validation.
|
|
| 2. |
- Abrahamsson, Thomas, 1956, et al.
(författare)
-
Finite element model calibration using frequency responses with damping equalization
- 2015
-
Ingår i: Mechanical Systems and Signal Processing. - : Elsevier BV. - 0888-3270 .- 1096-1216. ; 62-63, s. 218-234
-
Tidskriftsartikel (refereegranskat)abstract
- Model calibration is a cornerstone of the finite element verification and validation procedure, in which the credibility of the model is substantiated by positive comparison with test data. The calibration problem, in which the minimum deviation between finite element model data and experimental data is searched for, is normally characterized as being a large scale optimization problem with many model parameters to solve for and with deviation metrics that are nonlinear in these parameters. The calibrated parameters need to be found by iterative procedures, starting from initial estimates. Sometimes these procedures get trapped in local deviation function minima and do not converge to the globally optimal calibration solution that is searched for. The reason for such traps is often the multi-modality of the problem which causes eigenmode crossover problems in the iterative variation of parameter settings. This work presents a calibration formulation which gives a smooth deviation metric with a large radius of convergence to the global minimum. A damping equalization method is suggested to avoid the mode correlation and mode pairing problems that need to be solved in many other model updating procedures. By this method, the modal damping of a test data model and the finite element model is set to be the same fraction of critical modal damping. Mode pairing for mapping of experimentally found damping to the finite element model is thus not needed. The method is combined with model reduction for efficiency and employs the Levenberg-Marquardt minimizer with randomized starts to achieve the calibration solution. The performance of the calibration procedure, including a study of parameter bias and variance under noisy data conditions, is demonstrated by two numerical examples.
|
|
| 3. |
- Abrahamsson, Thomas, 1956, et al.
(författare)
-
What Is Normal About Normal Modes?
- 2016
-
Ingår i: Conference Proceedings of the Society for Experimental Mechanics Series. - Cham : Springer International Publishing. - 2191-5644 .- 2191-5652. ; 10, s. 97-110
-
Konferensbidrag (refereegranskat)abstract
- A normal mode of a vibrating system is a mode that is orthogonal to all other normal modes of that system. The orthogonality is in a weighted sense. For an undamped discretized linear mechanical system, the orthogonality is defined with respect to stiffness and mass such that normal modes are mutually stiffness and mass orthogonal. Another commonly used definition of an oscillating normal mode is that it is a pattern of motion in which all parts of the system vibrate harmonically with the same frequency and therefore with fixed relative phase relations between parts. The normality of a mode is thus not in a statistical sense. For lightly damped linear systems, a normal observation, i.e. one very common observation in the statistical sense, is that the phase relation between the motion of different parts of the system deviates very little from zero or pi. However, this normally occurring behavior should not lead us to think that that always has to be the case. Here it is shown by example that the normal modes of an undamped system may have arbitrary phase relations. One such mode of vibrationmay then possess the property of moving nodal lines, which is often attributed to non-proportionally damped or nonself-adjoint systems. The proper normalization of such modes is discussed and their relation to the well-known modal mass and MAC concepts and also to state-space based normalizations that are usually being used for complex-valued eigenmodes.
|
|
| 4. |
- Andersson, N., et al.
(författare)
-
Feature-based design analysis for automatic classification of simulated nonlinear responses using machine learning
- 2016
-
Ingår i: Proceedings of Isma2016 International Conference on Noise and Vibration Engineering and Usd2016 International Conference on Uncertainty in Structural Dynamics, ed. by Sas, P; Moens, D; VanDeWalle, A.. - 9789073802940 ; , s. 3259-3273
-
Konferensbidrag (refereegranskat)abstract
- A practical concept design analysis approach for automatic processing and classification of simulated responses is presented. Deterministic and nonlinear dynamics are studied under ideal loading and low noise conditions to determine fundamental system properties, how they vary and possibly interact. Responses should be classified into characteristic types, such as periodic or nonperiodic, resonant or nonresonant, linear or nonlinear, and further into subcategories such as single or dual frequency responses, hardening or softening. For this, time-signals are evaluated using methods and metrics commonly used within structural dynamics and then possibly associated with qualitative features according to measures based on a set of rules-of-thumb criteria. A support vector machine is trained to determine whether a single feature, or combinations of features, applies or not. This paper describes elements of the analysis, report practical considerations and discuss the effectiveness of evaluated features using known few-degree-of-freedom examples.
|
|
| 5. |
- Andersson, Niclas S, 1971, et al.
(författare)
-
Efficient Component Reductions in a Large-Scale Flexible Multibody Model
- 2018
-
Ingår i: SAE International Journal of Vehicle Dynamics, Stability, and NVH. - : SAE International. - 2380-2162 .- 2380-2170. ; 2:1, s. 5-26
-
Tidskriftsartikel (refereegranskat)abstract
- To make better use of simulations in the automotive driveline design process there is a need for both improved predictive capabilities of typical system models and increased number of variant evaluations carried out during system concept design phase. A previously developed large-scale multibody rotor dynamical powertrain model that combines detailed linear-elastic finite element components and nonlinear joints is used to more accurately simulate system response modes and their variations across the operating-range. However, the total simulation time is too long to include extensive parameter evaluations during the rapid design iterations, which will have a negative influence on the total understanding of the designed system's behaviour. Therefore this paper is about reducing such a large-scale model to one that runs faster, but without losing the ability to predict the most fundamental system characteristics. Reduction methods considering defined stimuli-response relations are well established and used within the field of control systems, to balance prediction accuracy and evaluation effort, but are not yet commonly applied to large-scaled structural models and analysis of vibrations in continuous and lightly damped structures. Here, an implementation of two such state-space reduction methods into a common computational software workflow is described and their overall efficiency is compared to standard methods. Reductions are applied to two major structural components of the powertrain model. Steady-state simulations are performed for multiple engine speeds and responses related to vehicle noise and vibrations are compared using a quantitative error metric. The prediction accuracy, reduction and response simulation times of different model orders are evaluated, as well as the corresponding mode frequency spectra.
|
|
| 6. |
- Andersson, Niclas S, 1971, et al.
(författare)
-
Feature-Based Response Classification in Nonlinear Structural Design Simulations
- 2018
-
Ingår i: SAE International Journal of Vehicle Dynamics, Stability, and NVH. - : SAE International. - 2380-2162 .- 2380-2170. ; 2:3, s. 185-202
-
Tidskriftsartikel (refereegranskat)abstract
- An applied system design analysis approach for automated processing and classification of simulated structural responses is presented. Deterministic and nonlinear dynamics are studied under ideal loading and low noise conditions to determine fundamental system properties, how they vary and possibly interact. Using powerful computer resources, large amounts of simulated raw data can be produced in a short period of time. Efficient tools for data processing and interpretation are then needed, but existing ones often require much manual preparation and direct human judgement. Thus, there is a need to develop techniques that help to treat more virtual prototype variants and efficiently extract useful information from them. For this, time signals are evaluated by methods commonly used within structural dynamics and statistical learning. A multi-level multi-frequency stimulus function is constructed and simulated response signals are combined into frequency domain functions. These are associated with qualitative system features, such as being periodic or aperiodic, linear or nonlinear and further into subcategories of nonlinear systems, such as fundamental, sub or super harmonic and even or odd order types. Appropriate classes are then determined from selected feature metrics and rules-of-thumb criteria. To automate the classification of large data sets, a support vector machine is trained on categorised responses to determine whether a single feature, or combinations of features, applies or not. The trained classifier can then efficiently process new sets of data and pick out cases that are associated with possible vibrational problems, which subsequently can be further analysed and understood. This article describes elements of the analysis, discuss the effectiveness of evaluated feature metrics, reports practical considerations and results from two separate training study examples.
|
|
| 7. |
- Brewick, P., et al.
(författare)
-
Hybrid Time/Frequency Domain Identification of Real Base-Isolated Structure
- 2016
-
Ingår i: Conference Proceedings of the Society for Experimental Mechanics Series. - Cham : Springer International Publishing. - 2191-5644 .- 2191-5652. - 9783319297514 ; 2, s. 303-311
-
Konferensbidrag (refereegranskat)abstract
- This paper presents a case study using hybrid time-and frequency-domain identifications in a synergistic manner to develop models of a full-scale experimental base-isolated structure. This four-story reinforced-concrete building on an isolation layer (of rubber bearings, elastic sliding bearings, passive metallic yielding dampers, and controllable oil dampers) was designed and constructed at the large-scale Japanese NIED E-Defense earthquake engineering laboratory. A variety of sensors, including accelerometers, were mounted within the structure to measure building response to shake table excitations. While the building was ultimately subjected to historical and synthetic ground motions, the recorded table and building accelerations during a number of random excitation tests are used to identify the structure's natural frequencies, damping ratios and mode shapes. The substantial damping provided by the isolation layer necessitates adopting a hybrid time-and frequency-domain approach for identification. The modes of the structure are separated by frequency content wherein lower frequency modes are identified using time domain approaches from the subspace identification family of methods and higher frequency modes are identified using frequency response functions. Individually, neither approach is able to successfully identify all of the desired modes but, through their combination, the modal properties of the structure are successfully characterized.
|
|
| 8. |
- Bylin, Axel, et al.
(författare)
-
Experimental-analytical state-space synthesis of passenger car components
- 2018
-
Ingår i: Proceedings of ISMA2018 International Conference on Noise and Vibration Engineering and USD2018 International Conference on Uncertainty in Structural Dynamics. - 9789073802995 ; 2018, s. 4021-4035
-
Konferensbidrag (refereegranskat)abstract
- In this paper an experimental model of a Volvo XC90 body-in-white and a finite element model of a rear subframe are coupled for vibration assessment using a state-space synthesis method. This problem is challenging due to the complex structures considered. They are exhibiting high modal densities, high damping and the coupling is through four rubber bushings. The models are connected through four virtual points created at each coupling position. The state-space synthesis method relies on a system identification procedure for the experimental model, which in itself pose a significant challenge for the modally dense body-in-white. In this paper a newly proposed system identification technique focusing not to the least on direct accelerances, based on a subspace state-space method, is used. It also aims to fulfil physically motivated constraints. Results from the coupled system are shown to agree well with experimental data of the system assembly, and the method outperforms a finite element based coupling.
|
|
| 9. |
- Chen, Yousheng, 1985-, et al.
(författare)
-
A Pretest Planning Method for Model Calibration for Nonlinear Systems
- 2016
-
Ingår i: Model Validation and Uncertainty Quantification, Volume 3. - Cham : Springer. - 2191-5644 .- 2191-5652. - 9783319297538 - 9783319297545 ; , s. 371-379
-
Konferensbidrag (övrigt vetenskapligt/konstnärligt)abstract
- With increasing demands on more flexible and lighter engineering structures, it has been more common to take nonlinearity into account. Model calibration is an important procedure for nonlinear analysis in structural dynamics with many industrial applications. Pretest planning plays a key role in the previously proposed calibration method for nonlinear systems, which is based on multi-harmonic excitation and an effective optimization routine. This paper aims to improve the pretest planning strategy for the proposed calibration method. In this study, the Fisher information matrix (FIM), which is calculated from the gradients with respect to the chosen parameters with unknown values, is used for determining the locations, frequency range, and the amplitudes of the excitations as well as the sensor placements. This pretest planning based model calibration method is validated by a structure with clearance nonlinearity. Synthetic test data is used to simulate the test procedure. Model calibration and K-fold cross validation are conducted for the optimum configurations selected from the pretest planning as well as three other configurations. The calibration and cross validation results show that a more accurate estimation of parameters can be obtained by using test data from the optimum configuration.
|
|
| 10. |
- Chen, Yousheng, 1985-, et al.
(författare)
-
An Efficient Simulation Method for Large-Scale Systems with Local Nonlinearities
- 2016
-
Ingår i: Special topics in structural dynamics, 34th IMAC. - Cham : Springer. - 2191-5644 .- 2191-5652. - 9783319299105 - 9783319299099
-
Konferensbidrag (övrigt vetenskapligt/konstnärligt)abstract
- In practice, most mechanical systems show nonlinear characteristics within the operational envelope. However, the nonlinearities are often caused by local phenomena and many mechanical systems can be well represented by a linear model enriched with local nonlinearities. Conventional nonlinear response simulations are often computationally intensive; the problem which becomes more severe when large-scale nonlinear systems are concerned. Thus, there is a need to further develop efficient simulation techniques. In this work, an efficient simulation method for large-scale systems with local nonlinearities is proposed. The method is formulated in a state-space form and the simulations are done in the Matlab environment. The nonlinear system is divided into a linearized system and a nonlinear part represented as external nonlinear forces acting on the linear system; thus taking advantage in the computationally superiority in the locally nonlinear system description compared to a generally nonlinear counterpart. The triangular-order hold exponential integrator is used to obtain a discrete state-space form. To shorten the simulation time additionally, auxiliary matrices, similarity transformation and compiled C-codes (mex) to be used for the time integration are studied. Comparisons of the efficiency and accuracy of the proposed method in relation to simulations using the ODE45 solver in Matlab and MSC Nastran are demonstrated on numerical examples of different model sizes.
|
|
| 11. |
- Chen, Yousheng, 1985-, et al.
(författare)
-
Experimental Validation of a Nonlinear Model Calibration Method Based on Multiharmonic Frequency Responses
- 2017
-
Ingår i: Journal of Computational and Nonlinear Dynamics. - : ASME Press. - 1555-1415 .- 1555-1423. ; 12:4
-
Tidskriftsartikel (refereegranskat)abstract
- Correlation and calibration using test data are natural ingredients in the process of validating computational models. Model calibration for the important subclass of nonlinear systems which consists of structures dominated by linear behavior with the presence of local nonlinear effects is studied in this work. The experimental validation of a nonlinear model calibration method is conducted using a replica of the École Centrale de Lyon (ECL) nonlinear benchmark test setup. The calibration method is based on the selection of uncertain model parameters and the data that form the calibration metric together with an efficient optimization routine. The parameterization is chosen so that the expected covariances of the parameter estimates are made small. To obtain informative data, the excitation force is designed to be multisinusoidal and the resulting steady-state multiharmonic frequency response data are measured. To shorten the optimization time, plausible starting seed candidates are selected using the Latin hypercube sampling method. The candidate parameter set giving the smallest deviation to the test data is used as a starting point for an iterative search for a calibration solution. The model calibration is conducted by minimizing the deviations between the measured steady-state multiharmonic frequency response data and the analytical counterparts that are calculated using the multiharmonic balance method. The resulting calibrated model's output corresponds well with the measured responses.
|
|
| 12. |
- Chen, Yousheng, 1985-, et al.
(författare)
-
Informative Data for Model Calibration of Locally Nonlinear Structures Based on Multi-Harmonic Frequency Responses
- 2016
-
Ingår i: Journal of Computational and Nonlinear Dynamics. - : ASME Press. - 1555-1415 .- 1555-1423. ; 11:5
-
Tidskriftsartikel (refereegranskat)abstract
- In industry, linear FE-models commonly serve as baseline models to represent the global structural dynamics behavior. However, available test data may show evidence of significant nonlinear dynamic characteristics. In such a case, the baseline linear model may be insufficient to represent the dynamics of the structure. The causes of the nonlinear characteristics may be local in nature and the remaining parts of the structure may be satisfactorily represented by linear descriptions. Although the baseline model can then serve as a good foundation, the physical phenomena needed to substantially increase the model's capability of representing the real structure are most likely not modelled in it. Therefore, a set of candidate nonlinear property parameters to control the nonlinear effects have to be added and subjected to calibration to form a credible model. The selection of the calibration parameters and the choice of data for a calibration metric form a coupled problem. An over-parameterized model for calibration may result in parameter value estimates that do not survive a validation test. The parameterization is coupled to the test data and should be chosen so that the expected co-variances of the chosen parameter's estimates are made small. Accurate test data, suitable for calibration, is often obtained from sinusoidal testing. Because a pure mono-sinusoidal excitation is difficult to achieve during a test of a nonlinear structure, the excitation is here designed to contain sub and super harmonics besides the fundamental harmonic. The steady-state responses at the side frequencies are shown to contain valuable information for the calibration process that can improve the accuracy of the parameter estimates. The nonlinear steady-state solutions can be found efficiently using the multi-harmonic balance method. In this paper, synthetic test data from a model of a nonlinear benchmark structure are used for illustration. The model calibration and an associated K-fold cross-validation are based on the Levenberg-Marquardt and the undamped Gauss-Newton algorithm, respectively. Starting seed candidates for calibration are found by the Latin hypercube sampling method. The realization that gives the smallest deviation to test data is selected as a starting point for the iterative search for a calibration solution. The calibration result shows good agreement with the true parameter setting, and the K-fold cross validation result shows that the variance of the estimated parameters shrinks when adding sub and super harmonics to the nonlinear frequency response functions.
|
|
| 13. |
- Chen, Yousheng, et al.
(författare)
-
Validation of a model calibration method through vibrational testing of a mechanical system with local clearance
- 2016
-
Ingår i: Proceedings of ISMA2016 International conference on noise and vibration engineering and USD2016 International conference on uncertainty in structural dynamics. - Leuven, Belgium : Katholieke University Leuven. - 9789073802940 ; , s. 2581-2595
-
Konferensbidrag (refereegranskat)abstract
- Nonlinear finite element models are often validated using experimental data. A previously proposed calibration method, which concerns pre-test planning, multi-sinusoidal excitation and an effective optimization routine, is improved with an extended version of the pre-test planning. The improved method is validated using a test structure with a clearance type nonlinearity. From the pretest planning, an optimal configuration for the data acquisition is determined. The multi-harmonic nonlinear frequency response functions (FRFs) of the structure under test are then generated by a multi-sinusoidal excitation. Model calibration is conducted by minimizing the difference between the experimental multi-harmonic nonlinear FRFs and their analytical counterparts. The uncertainties of the estimated parameters are assessed by a k-fold cross validation, which confirm that the uncertainties of the estimated parameters are small when the optimal configuration is applied.
|
|
| 14. |
|
|
| 15. |
- Gibanica, Mladen, 1988, et al.
(författare)
-
A reduced interface component mode synthesis method using coarse meshes
- 2017
-
Ingår i: Procedia Engineering. - : Elsevier BV. - 1877-7058 .- 1877-7058. ; 199, s. 348-353
-
Konferensbidrag (refereegranskat)abstract
- Component mode synthesis is a technique to simplify the analysis of complicated finite element models. A structure is split into substructures from which reduced order models can be generated and subsequently assembled. A model reduction performance gain can be limited if the component interfaces contain many degrees of freedom, which is often the case for high resolution models. In this paper a substructuring framework with interface reduction is presented. The method first splits a detailed model into substructures. The substructures’ fine mesh is then coarsened on the internal region, while keeping the boundary mesh intact. Thereafter a Guyan reduction is performed on each coarse mesh substructure. The Guyan computations are cheap due to the reduced size of the linear equation system necessary to solve for the coarse mesh system. After synthesis of the statically reduced systems, a reduction basis for the interface degrees of freedom is computed. Thereafter a Craig-Bampton reduction is performed on each fine mesh substructure using projections with the reduced interface degrees of freedom and fixed interface modes. The method is verified on a dense mesh plate model consisting of two substructures.
|
|
| 16. |
- Gibanica, Mladen, 1988, et al.
(författare)
-
Calibration, Validation and Uncertainty Quantification of Nominally Identical Car Subframes
- 2016
-
Ingår i: Conference Proceedings of the Society for Experimental Mechanics Series. - Cham : Springer International Publishing. - 2191-5644 .- 2191-5652. ; 3, s. 315-326
-
Konferensbidrag (refereegranskat)abstract
- In this paper a finite element model, with over half a million degrees-of-freedom, of a car front subframe has been calibrated and validated against experimental MIMO data of several nominally identical components. The spread between the individual components has been investigated and is reported. Sensor positioning was performed with an extended effective independence method, using system gramians to reject sensors with redundant information. The Fisher information matrix was used in the identification of the most significant model calibration parameters. Validation of the calibrated model was performed to evaluated the difference between the nominal and calibrated model, and bootstrapping used to investigate the validity of the calibrated parameters. The parameter identification, calibration, validation and bootstrapping have been performed using the open-source MATLAB tool FEMcali.
|
|
| 17. |
|
|
| 18. |
- Gibanica, Mladen, 1988, et al.
(författare)
-
Multifidelity component interface reduction and modal truncation augmentation
- 2019
-
Ingår i: International Journal for Numerical Methods in Engineering. - : Wiley. - 0029-5981 .- 1097-0207. ; 120:1, s. 105-124
-
Tidskriftsartikel (refereegranskat)abstract
- A Hurty‐Craig‐Bampton reduced order component can have unnecessarily large dimension if it contains many interface degrees‐of‐freedom. This is often the case for high spatial resolution models. Furthermore, for such high fidelity models their static constraint modes can be expensive to compute. To overcome these problems, a component mode synthesis method with interface reduction is developed using multifidelity models. The interface reduction basis is computed from the assembled system by coarsening each substructure's mesh, but keeping the model resolution at the interface intact. It is shown that such a mesh coarsening has a small effect on the interface reduction basis quality. Using this reduction basis the dimension of the static constraint modes problem can be reduced and the modes computed at a low cost. When few interface modes can be used without significant loss of accuracy, it is possible to enrich the Hurty‐Craig‐Bampton basis with modal truncation augmentation vectors to increase accuracy at a small extra cost. The accuracy of a procedure that utilises modal truncation augmentation vectors together with the multifidelity interface reduction is investigated. The method's performance and accuracy are illustrated on a planar problem and a more complex problem from industry.
|
|
| 19. |
- Gibanica, Mladen, 1988, et al.
(författare)
-
Parameter Estimation and Uncertainty Quantification of a Subframe with Mass Loaded Bushings
- 2017
-
Ingår i: Conference Proceedings of the Society for Experimental Mechanics Series. - Cham : Springer International Publishing. - 2191-5644 .- 2191-5652. ; 3, s. 61-76
-
Konferensbidrag (refereegranskat)abstract
- In the automotive industry components are often connected through rubber bushings. The bushingsʼ material properties are usually not well known. In computational models these properties are parametrised and their spread can be considerable. A good estimate of these parameters is important in various applications, including substructuring, and for uncertainty quantification of systems with connected components. This paper deals with the calibration of an industrial size finite element model of a car subframe with parametrised bushing models. Mass loading is used on the bushings to bring local bushing modes to a lower frequency region and impose a more realistic boundary condition in component testing. The model parameters can be calibrated in different ways. In this paper two approaches are considered. They are based on two test configurations, one with and one without mass loaded boundaries. In the first case only the bushing parameters are considered for the mass loaded boundary configuration. In the second case, consisting of two steps, the configuration without mass loaded boundaries is considered in which the bushing parameters are first fixed and other model parameters considered, and in the last step a subset of all parameters is considered. The calibration, validation and uncertainty quantification, using bootstrapping, have been performed using the open-source MATLAB tool FEMcali.
|
|
| 20. |
- Gibanica, Mladen, 1988, et al.
(författare)
-
Physically motivated rank constraint on direct throughput of state-space models
- 2018
-
Ingår i: IFAC-PapersOnLine. - : Elsevier BV. - 2405-8963. ; 51:15, s. 329-334
-
Konferensbidrag (refereegranskat)abstract
- In frequency range vibration testing a few outside band eigenmodes are often included in the system identification to compensate for residual mass and stiffness influences. It has been observed that, in particular, energy conjugate input-output pair transfer functions with strong outside band modes tend to render models with poor fit even after inclusion of mass and stiffness residuals. For such problems the inclusion of another complementary residual term has been found to improve the fit to data. In this paper, modal models identified from acceleration data with a subspace state-space method are considered. The residual mass influence is modelled with a state-space direct throughput while the stiffness and complementary residuals are modelled with extra states. Furthermore, for state-space models on accelerance form it is shown that the direct throughput matrix can be partitioned into a flexible and rigid motion partition. For systems with more inputs and outputs than rigid body modes it is shown that the rigid body motion partition has a bounded rank. The upper bound is equal to the number of rigid body modes. Therefore, for identified models on accelerance form this constraint must be enforced for physical consistency. The proposed method is applied on simulated finite element test data from an automotive component.
|
|
| 21. |
- Gibanica, Mladen, 1988, et al.
(författare)
-
Redundant Information Rejection in Sensor Localisation Using System Gramians
- 2016
-
Ingår i: Conference Proceedings of the Society for Experimental Mechanics Series. - Cham : Springer International Publishing. - 2191-5644 .- 2191-5652. ; 10, s. 325-333
-
Konferensbidrag (refereegranskat)abstract
- Sensors placement is important in vibration testing. The method of effective independence, recently extended to account for triaxial sensors, is widely used for this purpose in case a finite element model of the structure is available. In this paper a criteria is added to reject redundant information that usually arises in symmetric structures or finite element models with high candidate sensor density. A sensor placement strategy is proposed in which, initially, the method of effective independence is used to select the best sensors from a candidate set by maximising the Fisher information matrix determinant. Next, the gramians of a balanced realisation is used to compare the information between systems consisting of only previously added sensors to the final set with systems of the previous and candidate sensors. Sensors with redundant information will have negligible effect on the gramian and can be rejected. The method is fast, as gramians of systems with only one or two outputs are evaluated. It is sub-optimal in the sense that all possible sensor placement combinations are not evaluated for optimality. A test case, consisting of a rectangular plate, is presented, but the method has been used on a large scale industrial model with good results.
|
|
| 22. |
- Gibanica, Mladen, 1988, et al.
(författare)
-
Residual States for Modal Models Identified from Accelerance Data
- 2019
-
Ingår i: Conference Proceedings of the Society for Experimental Mechanics Series. - Cham : Springer International Publishing. - 2191-5644 .- 2191-5652. ; 9, s. 195-206
-
Konferensbidrag (refereegranskat)abstract
- Residual stiffness and mass terms are often employed in frequency response synthesis to compensate for outside band eigenmodes in the identification of modal models from test data. For structures that have strongly participating modes above the test frequency band, it has been observed that in particular direct accelerances with strong outside-band modal contribution tend to render modal models that give poor fit to test data. For such problems it may be insufficient to just add residual mass and stiffness terms to the accelerance modal series to get a sufficiently improved fit. For accelerance, such residual terms are constant and quadratic in frequency. Another, residual term that is quasi-linear over the frequency range of interest has been found to augment the identified model. In this paper that complementary term is added to the constant and quadratic terms in a state-space model identification with a subspace state-space identification method. A comparison is performed to an alternative residualisation method. The methods' results are compared on simulated finite element test data from of an automotive component.
|
|
| 23. |
- Johansson, Anders, 1982, et al.
(författare)
-
A method for improving test rig performance using passive components
- 2015
-
Ingår i: Mechanical Systems and Signal Processing. - : Elsevier BV. - 0888-3270 .- 1096-1216. ; 52-53:1, s. 614-627
-
Tidskriftsartikel (refereegranskat)abstract
- The time waveform replication (TWR) algorithm is presently used in industry for calculating the actuation force needed to replicate a certain reference sensor output in a test rig. Power and force rate limitations restrict the feasible range of that actuation force. If the input force distribution of the reference test cannot be replicated in the test rig, the required test rig input force magnitudes may be large or the replication properties poor due to lack of controllability. To circumvent this, a theory of passive components to improve replication and limit the input force demands of dynamic test rigs is developed. The theory fits within the framework of the TWR algorithm.
|
|
| 24. |
- Khorsand Vakilzadeh, Majid, 1984, et al.
(författare)
-
Approximate Bayesian Computation by Subset Simulation for model selection in dynamical systems
- 2017
-
Ingår i: Procedia Engineering. - : Elsevier BV. - 1877-7058 .- 1877-7058. ; 199, s. 1056-1061
-
Konferensbidrag (refereegranskat)abstract
- Approximate Bayesian Computation (ABC) methods are originally conceived to expand the horizon of Bayesian inference methods to the range of models for which only forward simulation is available. However, there are well-known limitations of the ABC approach to the Bayesian model selection problem, mainly due to lack of a sufficient summary statistics that work across models. In this paper, we show that formulating the standard ABC posterior distribution as the exact posterior PDF for a hierarchical state-space model class allows us to independently estimate the evidence for each alternative candidate model. We also show that the model evidence is a simple by-product of the ABC-SubSim algorithm. The validity of the proposed approach to ABC model selection is illustrated using simulated data from a three-story shear building with Masing hysteresis.
|
|
| 25. |
- Khorsand Vakilzadeh, Majid, 1984, et al.
(författare)
-
Approximate Bayesian Computation by Subset Simulation for Parameter Inference of Dynamical Models
- 2016
-
Ingår i: Conference Proceedings of the Society for Experimental Mechanics Series. - Cham : Springer International Publishing. - 2191-5644 .- 2191-5652. - 9783319297538 ; 3, s. 37-50
-
Konferensbidrag (refereegranskat)abstract
- A new multi-level Markov chain Monte Carlo algorithm for Bayesian inference, ABC-SubSim, has recently appeared that combines the principles of Approximate Bayesian Computation (ABC) with the method of subset simulation for efficient rare-event simulation. ABC-SubSim adaptively creates a nested decreasing sequence of data-approximating regions in the output space. This sequence corresponds to increasingly closer approximations of the observed output vector in this output space. At each stage, the approximate likelihood function at a given value of the model parameter vector is defined as the probability that the predicted output corresponding to that parameter value falls in the current data-approximating region. If continued to the limit, the sequence of the data-approximating regions would converge on to the observed output vector and the approximate likelihood function would become exact, but this is not computationally feasible. At the heart of this paper is the interpretation of the resulting approximate likelihood function. We show that under the assumption of the existence of uniformly-distributed measurement errors, ABC gives exact Bayesian inference. Moreover, we present a new optimal proposal variance scaling strategy which enables ABC-SubSim to efficiently explore the posterior PDF. The algorithm is applied to the model updating of a two degree-of-freedom linear structure to illustrate its ability to handle model classes with various degrees of identifiability.
|
|
| 26. |
- Khorsand Vakilzadeh, Majid, 1984, et al.
(författare)
-
Approximate Bayesian Computation by Subset Simulation using hierarchical state-space models
- 2017
-
Ingår i: Mechanical Systems and Signal Processing. - : Elsevier BV. - 0888-3270 .- 1096-1216. ; 84, s. 2-20
-
Tidskriftsartikel (refereegranskat)abstract
- A new multi-level Markov Chain Monte Carlo algorithm for Approximate Bayesian Computation, ABC-SubSim, has recently appeared that exploits the Subset Simulation method for efficient rare-event simulation. ABC-SubSim adaptively creates a nested decreasing sequence of data-approximating regions in the output space that correspond to increasingly closer approximations of the observed output vector in this output space. At each level, multiple samples of the model parameter vector are generated by a component-wise Metropolis algorithm so that the predicted output corresponding to each parameter value falls in the current data-approximating region. Theoretically, if continued to the limit, the sequence of data-approximating regions would converge on to the observed output vector and the approximate posterior distributions, which are conditional on the data approximation region, would become exact, but this is not practically feasible. In this paper we study the performance of the ABC-SubSim algorithm for Bayesian updating of the parameters of dynamical systems using a general hierarchical state-space model. We note that the ABC methodology gives an approximate posterior distribution that actually corresponds to an exact posterior where a uniformly distributed combined measurement and modeling error is added. We also note that ABC algorithms have a problem with learning the uncertain error variances in a stochastic state-space model and so we treat them as nuisance parameters and analytically integrate them out of the posterior distribution. In addition, the statistical efficiency of the original ABC-SubSim algorithm is improved by developing a novel strategy to regulate the proposal variance for the component-wise Metropolis algorithm at each level. We demonstrate that Self-regulated ABC-SubSim is well suited for Bayesian system identification by first applying it successfully to model updating of a two degree-of-freedom linear structure for three cases: globally, locally and unidentifiable model classes, and then to model updating of a two degree-of-freedom nonlinear structure with Duffing nonlinearities in its interstory force-deflection relationship. (C) 2016 Elsevier Ltd. All rights reserved.
|
|
| 27. |
- Khorsand Vakilzadeh, Majid, 1984, et al.
(författare)
-
Experiment design for improved frequency domain subspace system identification of continuous-time systems
- 2015
-
Ingår i: IFAC-PapersOnLine. - : Elsevier BV. - 2405-8963. ; 48:28, s. 886-891
-
Konferensbidrag (refereegranskat)abstract
- A widely used approach for identification of linear, time-invariant, MIMO (multi-input/multi output) systems from continuous-time frequency response data is to solve it in discrete-time domain using subspace based identification algorithm incorporated with a bilinear transformation. However, the bilinear transformation maps the distribution of the frequency lines from continuous-time domain to discrete-time domain in a non-linear fashion which may make identification algorithm to be ill-conditioned. In this paper we propose a solution to get around this problem by designing a dedicated frequency sampling strategy. Promising results are obtained when the algorithm is applied to synthetic data from a 6DOF mass-spring model.
|
|
| 28. |
- Khorsand Vakilzadeh, Majid, 1984, et al.
(författare)
-
Sequential gauss-newton MCMC algorithm for high-dimensional bayesian model updating
- 2017
-
Ingår i: Conference Proceedings of the Society for Experimental Mechanics Series. - Cham : Springer International Publishing. - 2191-5644 .- 2191-5652. ; 3 Part F2, s. 303-314
-
Konferensbidrag (refereegranskat)abstract
- Bayesian model updating provides a rigorous framework to account for uncertainty induced by lack of knowledge about engineering systems in their respective mathematical models through updates of the joint probability density function (PDF), the so-called posterior PDF, of the unknown model parameters. The Markov chain Monte Carlo (MCMC) methods are currently the most popular approaches for generating samples from the posterior PDF. However, these methods often found wanting when sampling from difficult distributions (e.g., high-dimensional PDFs, PDFs with flat manifolds, multimodal PDFs, and very peaked PDFs). This paper introduces a new multi-level sampling approach for Bayesian model updating, called Sequential Gauss-Newton algorithm, which is inspired by the Transitional Markov chain Monte Carlo (TMCMC) algorithm. The Sequential Gauss-Newton algorithm improves two aspects of TMCMC to make an efficient and effective MCMC algorithm for drawing samples from difficult posterior PDFs. First, the statistical efficiency of the algorithm is enhanced by use of the systematic resampling scheme. Second, a new MCMC algorithm, called Gauss-Newton MCMC algorithm, is proposed which is essentially an M-H algorithm with a Gaussian proposal PDF tailored to the posterior PDF using the gradient and Hessian information of the negative log posterior. The effectiveness of the proposed algorithm for solving the Bayesian model updating problem is illustrated using three examples with irregularly shaped posterior PDFs.
|
|
| 29. |
- Khorsand Vakilzadeh, Majid, 1984, et al.
(författare)
-
Stochastic finite element model calibration based on frequency responses and bootstrap sampling
- 2017
-
Ingår i: Mechanical Systems and Signal Processing. - : Elsevier BV. - 0888-3270 .- 1096-1216. ; 88, s. 180-198
-
Tidskriftsartikel (refereegranskat)abstract
- Chemical and topological parameters have been widely used for predicting the phase selection in high-entropy alloys (HEAs). Nevertheless, previous studies could be faulted due to the small number of available data points, the negligence of kinetic effects, and the insensitivity to small compositional changes. Here in this work, 92 TiZrHfM, TiZrHfMM, TiZrHfMMM (M = Fe, Cr, V, Nb, Al, Ag, Cu, Ni) HEAs were prepared by melt spinning, to build a reliable and sufficiently large material database to inspect the robustness of previously established parameters. Modification of atomic radii by considering the change of local electronic environment in alloys, was critically found out to be superior in distinguishing the formation of amorphous and crystalline alloys, when compared to using atomic radii of pure elements in topological parameters. Moreover, crystal structures of alloying element were found to play an important role in the amorphous phase formation, which was then attributed to how alloying hexagonal-close-packed elements and face-centered-cubic or body-centered-cubic elements can affect the mixing enthalpy. Findings from this work not only provide parametric studies for HEAs with new and important perspectives, but also reveal possibly a hidden connection among some important concepts in various fields.
|
|
| 30. |
- Khorsand Vakilzadeh, Majid, 1984, et al.
(författare)
-
Towards an Automatic Modal Parameter Estimation Framework: Mode Clustering
- 2015
-
Ingår i: Conference Proceedings of the Society for Experimental Mechanics Series. - Cham : Springer International Publishing. - 2191-5644 .- 2191-5652. - 9783319152509 ; 10, s. 243-259
-
Konferensbidrag (refereegranskat)abstract
- The estimation of modal parameters from a set of measured data is a highly judgmental task, with user expertise playing a significant role for distinguishing between physical and spurious modes. However, it can be very tedious especially in situations when the data is difficult to analyze. This study presents a new algorithm for mode clustering as a preliminary step in a multi-step algorithm for performing physical mode selection with little or no user interaction. The algorithm commences by identification of a high-order model from estimated frequency response functions to collect all the important characteristics of the structure in a so-called library of modes. This often results in the presence of spurious modes which can be detected on the basis of the hypothesis that spurious modes are estimated with a higher level of uncertainty comparing to physical modes. Therefore, we construct a series of data using a simple random sampling technique in order to obtain a set of linear systems using subspace identification. Then, their similar modes are grouped together using a new correlation criterion, which is called Modal Observability Correlation (MOC). An illustrative example shows the efficiency of the proposed clustering technique and also demonstrates its capability to dealing with inconsistent data.
|
|
| 31. |
- Khorsand Vakilzadeh, Majid, 1984, et al.
(författare)
-
Using approximate Bayesian computation by Subset Simulation for efficient posterior assessment of dynamic state-space model classes
- 2018
-
Ingår i: SIAM Journal of Scientific Computing. - 1064-8275 .- 1095-7197. ; 40:1, s. B168-B195
-
Tidskriftsartikel (refereegranskat)abstract
- Approximate Bayesian Computation (ABC) methods have gained in popularity over the last decade because they expand the horizon of Bayesian parameter inference methods to the range of models for which an analytical formula for the likelihood function might be difficult, or even impossible, to establish. The majority of the ABC methods rely on the choice of a set of summary statistics to reduce the dimension of the data. However, as has been noted in the ABC literature, the lack of convergence guarantees induced by the absence of a vector of sufficient summary statistics that assures intermodel sufficiency over the set of competing models hinders the use of the usual ABC methods when applied to Bayesian model selection or assessment. In this paper, we present a novel ABC model selection procedure for dynamical systems based on a recently introduced multilevel Markov chain Monte Carlo method, self-regulating ABC-SubSim, and a hierarchical state-space formulation of dynamic models. We show that this formulation makes it possible to independently approximate the model evidence required for assessing the posterior probability of each of the competing models. We also show that ABC-SubSim not only provides an estimate of the model evidence as a simple by-product but also gives the posterior probability of each model as a function of the tolerance level, which allows the ABC model choices made in previous studies to be understood. We illustrate the performance of the proposed framework for ABC model updating and model class selection by applying it to two problems in Bayesian system identification: a single-degree-of-freedom bilinear hysteretic oscillator and a three-story shear building with Masing hysteresis, both of which are subject to a seismic excitation.
|
|
| 32. |
- Larsson, Karl-Johan, 1985, et al.
(författare)
-
Calibration and cross-validation of a car component model using repeated testing
- 2015
-
Ingår i: Conference Proceedings of the Society for Experimental Mechanics Series. - Cham : Springer International Publishing. - 2191-5644 .- 2191-5652. ; 3
-
Konferensbidrag (övrigt vetenskapligt/konstnärligt)abstract
- © The Society for Experimental Mechanics, Inc. 2015. Repeated testing has been made both in the sense of testing multiple test pieces of the same type and in the sense of repeated tests on the same test piece by 13 testers at different occasions. Between the individual tests, the test subject has been dismounted from the test stand and the sensors have been re-calibrated. Statistical evaluation of these tests gives information about the spread that can be expected in modal tests. The SIMO test data and data statistics are used for traditional validation and cross-validation of a finite element model of the car component under test. Planning of the sensor layout has been made prior to testing with the use of the nominal finite element model. A model calibration is made prior to the model validation. The finite element model size is over 20,000 degrees-of-freedom and involves two calibration parameters. The work has been made as part of a structural dynamics model validation Master’s course. An open-source Matlab application has been used for calibration, validation and cross-validation
|
|
| 33. |
- Linderholt, Andreas, 1968-, et al.
(författare)
-
A Comparison of the Dynamic Behavior of Three Sets of the Ampair 600 Wind Turbine
- 2015
-
Ingår i: Dynamics of Coupled Structures. - Cham : Springer. - 2191-5644 .- 2191-5652. - 9783319152080 ; , s. 99-110
-
Konferensbidrag (övrigt vetenskapligt/konstnärligt)abstract
- The Ampair 600 wind turbine assembly has been modified to suit as a benchmark structure in the pursuit of finding best practices for experimental substructure coupling of structural dynamic systems. Seven such systems have been assembled in test laboratories in Europe and in the USA. We scrutinize the dynamic behavior of the total assembly of three of those by vibration testing and compare the test outcome from seemingly identical assemblies. The aim is to support future component synthesis activities with high fidelity data and support future model validation. Comparisons are made by evaluating deviations of measured frequency response functions, the differences of identified structural eigenfrequencies and the correlation between eigenvector estimates. The testing is made in two parts. First, a partly assembled system, not including the hub and blades, is tested. This constitutes one possible and logical subsystem splitting that is likely to appear in future substructure synthesis efforts. In the second part, the full system assembly is tested. The test procedure, the test setup, the obtained test data and test data statistics are presented.
|
|
| 34. |
- Linderholt, Andreas, 1968-, et al.
(författare)
-
The Dynamic Behavior of Three Sets of the Ampair 600 Wind Turbine
- 2016
-
Ingår i: Proceedings of the International Modal Analysis Conference (IMAC XXXIV), 2016. - : SEM Society of Experimental Mechanics.
-
Konferensbidrag (övrigt vetenskapligt/konstnärligt)abstract
- A modified version of the Ampair 600 wind turbine was defined by the SEM substructuring focus group to be a benchmark for use in the pursuit of developing methods for high precision experimental substructuring. Today, seven such, intended to be identical, systems have been assembled at test laboratories within Europe and the USA. This paper reports the experimentally found dynamic behavior of five out of the seven systems. The aim of the study is to provide high fidelity data to support future component synthesis and model validation activities. Comparisons are made by evaluating deviations of measured frequency response functions and identified structural eigenfrequencies together with the correlation between eigenvector estimates. The testing is made in two parts. First, partly assembled systems, not including the rotors or the blades, are tested. This constitutes one possible and logical subsystem splitting that is likely to appear in future substructure synthesis efforts. In the second part, the fully assembled benchmark systems are tested. The test procedure, the test setup, the obtained test data and test data statistics are presented.
|
|
| 35. |
- Linderholt, Andreas, 1968-, et al.
(författare)
-
Time Domain Dynamic Simulations of Locally Nonlinear Large-Scale Systems
- 2016
-
Ingår i: Presented at Aerospace technology congress.
-
Konferensbidrag (övrigt vetenskapligt/konstnärligt)abstract
- In practice, most mechanical systems show nonlinear characteristics within the operational envelope. However, the nonlinearities are often caused by local phenomena and many mechanical systems can be well represented by a linear model enriched with local nonlinearities. Conventional nonlinear response simulations are often computationally intensive; the problem which becomes more severe when large-scale nonlinear systems are concerned. Thus, there is a need to further develop efficient simulation techniques. In this work, an efficient simulation method for large-scale systems with local nonlinearities is proposed. The method is formulated in a state-space form and the simulations are done in the Matlab environment. The nonlinear system is divided into a linearized system and a nonlinear part represented as external nonlinear forces acting on the linear system; thus taking advantage in the computationally superiority in the locally nonlinear system description compared to a generally nonlinear counterpart. The triangular-order hold exponential integrator is used to obtain a discrete state-space form. To shorten the simulation time additionally, auxiliary matrices, similarity transformation and compiled C-codes (mex) to be used for the time integration are studied. Comparisons of the efficiency and accuracy of the proposed method in relation to simulations using the ODE45 solver in Matlab and MSC Nastran are demonstrated on numerical examples of different model sizes.
|
|
| 36. |
- Rahrovani, Sadegh, 1981, et al.
(författare)
-
Bayesian Treatment of Spatially-Varying Parameter Estimation Problems via Canonical BUS
- 2016
-
Ingår i: Conference Proceedings of the Society for Experimental Mechanics Series. - Cham : Springer International Publishing. - 2191-5644 .- 2191-5652. - 9783319297538 ; 3, s. 1-13
-
Konferensbidrag (refereegranskat)abstract
- The inverse problem of identifying spatially-varying parameters, based on indirect/incomplete experimental data, is a computationally and conceptually challenging problem. One issue of concern is that the variation of the parameter random field is not known a priori, and therefore, it is typical that inappropriate discretization of the parameter field leads to either poor modelling (due to modelling error) or ill-condition problem (due to the use of over-parameterized models). As a result, classical least square or maximum likelihood estimation typically performs poorly. Even with a proper discretization, these problems are computationally cumbersome since they are usually associated with a large vector of unknown parameters. This paper addresses these issues through Bayesian approach, via a recently developed stochastic simulation algorithm, called Canonical BUS. This algorithm is considered as a revisited formulation of the original BUS (Bayesian Updating using Structural reliability methods), that is, an enhancement of rejection approach that is used in conjunction with Subset Simulation rare-event sampler. Desirable features of the method and its performance to treat real-world applications has been investigated. The studied industrial problem originates from a railway mechanics application, where the spatial variation of ballast bed is of particular interest.
|
|
| 37. |
- Rahrovani, Sadegh, 1981, et al.
(författare)
-
Stability Limitations in Simulation of Dynamical Systems with Multiple Time-Scales
- 2016
-
Ingår i: Conference Proceedings of the Society for Experimental Mechanics Series. - Cham : Springer International Publishing. - 2191-5644 .- 2191-5652. - 9783319152219 ; 1, s. 93-105
-
Konferensbidrag (refereegranskat)abstract
- This paper focuses on the stability properties of a recently proposed exponential integrator particularly in simulation of highly oscillatory systems with multiple time-scales. The linear and nonlinear stability properties of the presented exponential integrator have been studied. We illustrate this with the Fermi–Pasta–Ulam (FPU) problem, a highly oscillatory nonlinear system known as a test benchmark for multi-scale time integrators. This example is also illustrative when studying the numerical resonance and algorithmic instability in the multi-time-stepping (MTS) methods, such as in exponential and/or trigonometric integration schemes, since it has no external input force and therefore no real physical resonance.
|
|
| 38. |
- Yaghoubi Nasrabadi, Vahid, 1985, et al.
(författare)
-
A Parallel Solution Method for Structural Dynamic Response Analysis
- 2015
-
Ingår i: Conference Proceedings of the Society for Experimental Mechanics Series. - Cham : Springer International Publishing. - 2191-5644 .- 2191-5652. ; 4, s. 149-161
-
Konferensbidrag (refereegranskat)abstract
- With the continuous improvements of technology in and around multi-core CPU:s and GPU:s there is a strong desire to exploit this technology in its full potential. For structural dynamics problems, the domain decomposition is a very mature technique that is well adapted to parallel computations in multi-core machines as it is almost trivially parallelizable. However, competing alternatives with model reduction without parallel computation has also reached an extremely high level of maturity and are thus highly competitive. In this paper, a domain decomposition method, in a procedure named the split-stitch-spread (S3) procedure, is proposed to do transient analysis of large finite element models in parallel. In the method, the structure splits into model substructures with elastic interfacial substructures coupling them together. Each of them can be sent to different computer cores to do time discretization. The model substructures stitch to each other by using interfacial forces and as a result, the systems’ state sequence will be obtained. The solution can then be spread into the substructures and response quantities can be evaluated in parallel processing. The method is applied to a multi-story building subjected to earthquake loading and the results are compared with mode displacement method as a model reduction method with focus on computational efficiency.
|
|
| 39. |
|
|
| 40. |
- Yaghoubi Nasrabadi, Vahid, 1985, et al.
(författare)
-
An efficient exponential predictor-corrector time integration method for structures with local nonlinearity
- 2016
-
Ingår i: Engineering Structures. - : Elsevier BV. - 1873-7323 .- 0141-0296. ; 128, s. 344-361
-
Tidskriftsartikel (refereegranskat)abstract
- Simulating the nonlinear behavior of complex systems requires significant computational effort. Despite the rapid progress in computing technology, the demand is still strong for more efficient simulation methods in diverse structural dynamics fields such as nonlinear system identification and nonlinear system reliability. In addition to efficiency, algorithmic stability and accuracy must be addressed in the development of new simulation procedures. In this paper, a method to treat localized nonlinearities in a structure efficiently and accurately is proposed. The method is conditionally stable. The system equations are separated into a state-invariant linear part and a state-dependent nonlinear part that is considered to be external pseudo-forces that act on the linear system. The response of the system is obtained by fixed point iterations in which the pseudo-forces are updated until convergence. In addition to the one time-step-ahead prediction form, the novel idea of multiple time-step-ahead prediction is proposed. The effect of this approach is investigated and shown to increase algorithm efficiency and stability. To perform the numerical integration, time-stepping schemes like the exponential first-order hold method can be used to the advantage of efficiency and accuracy. To increase the accuracy and stability of the method, a novel second-order hold equivalent is derived and implemented. The efficiency, stability, and accuracy of the method are demonstrated in numerical examples. Finally, the method is applied to the earthquake-induced motion of a 20-story building with local nonlinearities.
|
|
| 41. |
- Yaghoubi Nasrabadi, Vahid, 1985, et al.
(författare)
-
Automated modal parameter estimation using correlation analysis and bootstrap sampling
- 2018
-
Ingår i: Mechanical Systems and Signal Processing. - : Elsevier BV. - 0888-3270 .- 1096-1216. ; 100, s. 289-310
-
Tidskriftsartikel (refereegranskat)abstract
- The estimation of modal parameters from a set of noisy measured data is a highly judgmental task, with user expertise playing a significant role in distinguishing between estimated physical and noise modes of a test-piece. Various methods have been developed to automate this procedure. The common approach is to identify models with different orders and cluster similar modes together. However, most proposed methods based on this approach suffer from high-dimensional optimization problems in either the estimation or clustering step. To overcome this problem, this study presents an algorithm for autonomous modal parameter estimation in which the only required optimization is performed in a three-dimensional space. To this end, a subspace-based identification method is employed for the estimation and a non-iterative correlation-based method is used for the clustering. This clustering is at the heart of the paper. The keys to success are correlation metrics that are able to treat the problems of spatial eigenvector aliasing and nonunique eigenvectors of coalescent modes simultaneously. The algorithm commences by the identification of an excessively high-order model from frequency response function test data. The high number of modes of this model provides bases for two subspaces: one for likely physical modes of the tested system and one for its complement dubbed the subspace of noise modes. By employing the bootstrap resampling technique, several subsets are generated from the same basic dataset and for each of them a model is identified to form a set of models. Then, by correlation analysis with the two aforementioned subspaces, highly correlated modes of these models which appear repeatedly are clustered together and the noise modes are collected in a so-called Trashbox cluster. Stray noise modes attracted to the mode clusters are trimmed away in a second step by correlation analysis. The final step of the algorithm is a fuzzy c-means clustering procedure applied to a three-dimensional feature space to assign a degree of physicalness to each cluster. The proposed algorithm is applied to two case studies: one with synthetic data and one with real test data obtained from a hammer impact test. The results indicate that the algorithm successfully clusters similar modes and gives a reasonable quantification of the extent to which each cluster is physical.
|
|
| 42. |
- Yaghoubi Nasrabadi, Vahid, 1985, et al.
(författare)
-
Sparse polynomial chaos expansions of frequency response functions using stochastic frequency transformation
- 2017
-
Ingår i: Probabilistic Engineering Mechanics. - : Elsevier BV. - 0266-8920 .- 1878-4275. ; 48, s. 39-58
-
Tidskriftsartikel (refereegranskat)abstract
- Frequency response functions (FRFs) are important for assessing the behavior of stochastic linear dynamic systems. For large systems, their evaluations are time-consuming even for a single simulation. In such cases, uncertainty quantification by crude Monte-Carlo simulation is not feasible. In this paper, we propose the use of sparse adaptive polynomial chaos expansions (PCE) as a surrogate of the full model. To overcome known limitations of PCE when applied to FRF simulation, we propose a frequency transformation strategy that maximizes the similarity between FRFs prior to the calculation of the PCE surrogate. This strategy results in lower-order PCEs for each frequency. Principal component analysis is then employed to reduce the number of random outputs. The proposed approach is applied to two case studies: a simple 2-DOF system and a 6-DOF system with 16 random inputs. The accuracy assessment of the results indicates that the proposed approach can predict single FRFs accurately. Besides, it is shown that the first two moments of the FRFs obtained by the PCE converge to the reference results faster than with the Monte-Carlo (MC) methods.
|
|
| 43. |
- Yaghoubi Nasrabadi, Vahid, 1985, et al.
(författare)
-
Stochastic Finite Element Model Updating by Bootstrapping
- 2016
-
Ingår i: Conference Proceedings of the Society for Experimental Mechanics Series. - Cham : Springer International Publishing. - 2191-5644 .- 2191-5652. ; 3, s. 117-130
-
Konferensbidrag (refereegranskat)abstract
- This paper presents a new stochastic finite element model calibration framework for estimation of the uncertainty in model parameters, which combines the principles of bootstrapping with the technique of FE model calibration with damping equalization. The bootstrapping allows to quantify the uncertainty bounds on the model parameters by constructing a number of resamples, with replacement, of the experimental data and solving the FE model calibration problem on the resampled datasets. To a great extent, the success of the calibration problem depends on the starting value for the parameters. The formulation of FE model calibration with damping equalization gives a smooth metric with a large radius of convergence to the global minimum and its solution can be viewed as the initial estimate for the model parameters. In this study, practical suggestions are made to improve the performance of this algorithm in dealing with noisy measurements. The performance of the proposed stochastic calibration algorithm is illustrated using simulated data for a six degree-of-freedom mass-spring model.
|
|
