0
Research Papers

Multifidelity Recursive Cokriging for Dynamic Systems' Response Modification

[+] Author and Article Information
Luigi Bregant

Department of Engineering and Architecture,
University of Trieste,
Via A. Valerio 10,
Trieste 34127, Italy
e-mail: bregant@units.it

Lucia Parussini

Department of Engineering and Architecture,
University of Trieste,
Via A. Valerio 10,
Trieste 34127, Italy
e-mail: lparussini@units.it

Valentino Pediroda

Department of Engineering and Architecture,
University of Trieste,
Via A. Valerio 10,
Trieste 34127, Italy
e-mail: pediroda@units.it

Contributed by the Technical Committee on Vibration and Sound of ASME for publication in the JOURNAL OF VIBRATION AND ACOUSTICS. Manuscript received December 1, 2017; final manuscript received June 19, 2018; published online July 24, 2018. Assoc. Editor: Julian Rimoli.

J. Vib. Acoust 141(1), 011002 (Jul 24, 2018) (10 pages) Paper No: VIB-17-1522; doi: 10.1115/1.4040597 History: Received December 01, 2017; Revised June 19, 2018

In order to perform the accurate tuning of a machine and improve its performance to the requested tasks, the knowledge of the reciprocal influence among the system's parameters is of paramount importance to achieve the sought result with minimum effort and time. Numerical simulations are an invaluable tool to carry out the system optimization, but modeling limitations restrict the capabilities of this approach. On the other side, real tests and measurements are lengthy, expensive, and not always feasible. This is the reason why a mixed approach is presented in this work. The combination, through recursive cokriging, of low-fidelity, yet extensive, numerical model results, together with a limited number of highly accurate experimental measurements, allows to understand the dynamics of the machine in an extended and accurate way. The results of a controllable experiment are presented and the advantages and drawbacks of the proposed approach are also discussed.

FIGURES IN THIS ARTICLE
<>
Copyright © 2019 by ASME
Your Session has timed out. Please sign back in to continue.

References

Fernandez-Godino, M. G. , Park, C. , Kim, N.-H. , and Haftka, R. T. , 2017, “ Review of Multi-Fidelity Models,” e-print arXiv: 1609.07196. https://arxiv.org/abs/1609.07196
Friswell, M. I. , and Mottershead, J. E. , 1995, Finite Element Model Updating in Structural Dynamics. Solid Mechanics and Its Applications, Kluwer Academic Publishers Group, Dordrecht, The Netherlands. [CrossRef]
Lucifredi, A. , Mazzieri, C. , and Rossi, M. , 2000, “ Application of Multiregressive Linear Models, Dynamic Kriging Models and Neural Network Models to Predictive Maintenance of Hydroelectric Power Systems,” Mech. Syst. Signal Process., 14(3), pp. 471–494. [CrossRef]
Liu, X. , Li, M. , and Xu, M. , 2016, “ Kriging Assisted On-Line Torque Calculation for Brushless DC Motors Used in Electric Vehicles,” Int. J Automot. Technol., 17(1), pp. 153–164. [CrossRef]
Krige, D. G. , 1951, “ A Statistical Approach to Some Basic Mine Valuation Problems on the Witwatersrand,” J. Southern Afr. Inst. Min. Metall., 52(6), pp. 119–139.
Matheron, G. , 1963, “ Principles of Geostatistics,” Econ. Geol., 58(8), pp. 1246–1266. [CrossRef]
Sacks, J. , Welch, W. , J., Toby , M. , and Wynn, H. P. , 1989, “ Design and Analysis of Computer Experiments,” Stat. Sci., 4(4), pp. 409–423. [CrossRef]
Kennedy, M. C. , and O'Hagan, A. , 2000, “ Predicting the Output From a Complex Computer Code When Fast Approximations are Available,” Biometrika, 87(1), pp. 1–13. [CrossRef]
Gratiét, L. L. , and Cannamela, C. , 2015, “ Kriging-Based Sequential Design Strategies Using Fast Cross-Validation Techniques With Extensions to Multi-Fidelity Computer Codes,” Technometrics, 57(3), pp. 418–427. [CrossRef]
Gratiét, L. L. , 2012, “ Bayesian Analysis of Hierarchical Multi-Fidelity Codes,” SIAM J. Uncertainty Quantif., 1(1), pp. 244–269. [CrossRef]
Gratiét, L. L. , 2013, “ Multi-Fidelity Gaussian Process Regression for Computer Experiments,” Ph.D. thesis, University of Paris VII Denis-Diderot, Paris, France.

Figures

Grahic Jump Location
Fig. 1

Scheme of test system (bold lines represent the locus of points where the two masses can be positioned)

Grahic Jump Location
Fig. 2

Natural frequency 9 for different configurations of masses (x1, x2): (a) values measured for 11 different configurations (x1, x2) and (b) values computed by FEM code with the fine mesh in 2500 different configurations (x1, x2)

Grahic Jump Location
Fig. 5

Natural frequency 9: high-fidelity and low-fidelity training points of recursive cokriging model

Grahic Jump Location
Fig. 3

Natural frequency 9: comparison between numerical models and experimental values: (a) original FE model, (b) updated FE model, (c) kriging model, and (d) recursive cokriging model

Grahic Jump Location
Fig. 4

Kriging model of natural frequency 9: (a) mean and (b) standard deviation

Grahic Jump Location
Fig. 6

Recursive cokriging model of natural frequency 9: (a) mean and (b) standard deviation

Grahic Jump Location
Fig. 7

Recursive cokriging model of natural frequency 7: (a) mean and (b) standard deviation are computed using 8 high fidelity training points, (c) mean and (d) standard deviation are computed using 9 high fidelity training points

Grahic Jump Location
Fig. 8

(a) Picture and (b) scheme of system to optimize (in (b) the bold lines represent the points where the two masses can be positioned)

Grahic Jump Location
Fig. 9

Recursive cokriging model of natural frequencies 8 an 9: (a) f8 mean, (b) f8 standard deviation, (c) f9 mean, and (d) f9 standard deviation (black line: locus of points (x1, x2), where the natural frequency is equal to frequency fe1 or fe2; gray line: locus of points (x1, x2), where the natural frequency is equal to frequency fe1±2 Hz or fe2±2 Hz)

Tables

Errata

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In