Extending modal testing technology for model validation of engineering structures with sparse nonlinearities: A first case study

Arnaldo delli Carri, B. Weekes, Dario Di Maio, D. J. Ewins

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

Modal testing is widely used today as a means of validating theoretical (Finite Element) models for the dynamic analysis of engineering structures, prior to these models being used for optimisation of product design. Current model validation methodology is confined to linear models and is primarily concerned with (i) correcting inaccurate model parameters and (ii) ensuring that sufficient elements are included for these cases, using measured data. Basic experience is that this works quite well, largely because the weaknesses in the models are relatively sparse and, as a result, are usually identifiable and correctable. The current state-of-the-art in linear model validation has contributed to an awareness that residual errors in FE models are increasingly the consequence of some unrepresented nonlinearity in the structure. In these cases, additional, higher order parameters are required to improve the model so that it can represent the nonlinear behaviour. This is opposed to the current practice of simply refining the mesh. Again, these nonlinear features are generally localised, and are often associated with joints. We seek to provide a procedure for extending existing modal testing to enable these nonlinear elements to be addressed using current nonlinear identification methods directed at detection, characterisation, location and then quantification – in order to enhance the elements in an FE model as necessary to describe nonlinear dynamic behaviour. Emphasis is placed on the outcome of these extended methods to relate specifically to the physical behaviour of the relevant components of the structure, rather than to the nonlinear response characteristics that are the result of their presence.
LanguageEnglish
Pages97-115
Number of pages19
JournalMechanical Systems and Signal Processing
Volume84
Issue numberPart B
Early online date23 Apr 2016
DOIs
Publication statusPublished - 1 Feb 2017

Fingerprint

Testing
Product design
Dynamic analysis
Refining

Keywords

  • Modal testing
  • Model updating
  • Model upgrading
  • Model validation
  • Nonlinear structures

Cite this

Extending modal testing technology for model validation of engineering structures with sparse nonlinearities: A first case study. / delli Carri, Arnaldo; Weekes, B.; Di Maio, Dario; Ewins, D. J.

In: Mechanical Systems and Signal Processing, Vol. 84, No. Part B, 01.02.2017, p. 97-115.

Research output: Contribution to journalArticle

@article{9492bc336daf48e6925fef55858de08b,
title = "Extending modal testing technology for model validation of engineering structures with sparse nonlinearities: A first case study",
abstract = "Modal testing is widely used today as a means of validating theoretical (Finite Element) models for the dynamic analysis of engineering structures, prior to these models being used for optimisation of product design. Current model validation methodology is confined to linear models and is primarily concerned with (i) correcting inaccurate model parameters and (ii) ensuring that sufficient elements are included for these cases, using measured data. Basic experience is that this works quite well, largely because the weaknesses in the models are relatively sparse and, as a result, are usually identifiable and correctable. The current state-of-the-art in linear model validation has contributed to an awareness that residual errors in FE models are increasingly the consequence of some unrepresented nonlinearity in the structure. In these cases, additional, higher order parameters are required to improve the model so that it can represent the nonlinear behaviour. This is opposed to the current practice of simply refining the mesh. Again, these nonlinear features are generally localised, and are often associated with joints. We seek to provide a procedure for extending existing modal testing to enable these nonlinear elements to be addressed using current nonlinear identification methods directed at detection, characterisation, location and then quantification – in order to enhance the elements in an FE model as necessary to describe nonlinear dynamic behaviour. Emphasis is placed on the outcome of these extended methods to relate specifically to the physical behaviour of the relevant components of the structure, rather than to the nonlinear response characteristics that are the result of their presence.",
keywords = "Modal testing, Model updating, Model upgrading, Model validation, Nonlinear structures",
author = "{delli Carri}, Arnaldo and B. Weekes and {Di Maio}, Dario and Ewins, {D. J.}",
year = "2017",
month = "2",
day = "1",
doi = "10.1016/J.YMSSP.2016.04.012",
language = "English",
volume = "84",
pages = "97--115",
journal = "Mechanical Systems and Signal Processing",
issn = "0888-3270",
publisher = "Elsevier",
number = "Part B",

}

TY - JOUR

T1 - Extending modal testing technology for model validation of engineering structures with sparse nonlinearities: A first case study

AU - delli Carri, Arnaldo

AU - Weekes, B.

AU - Di Maio, Dario

AU - Ewins, D. J.

PY - 2017/2/1

Y1 - 2017/2/1

N2 - Modal testing is widely used today as a means of validating theoretical (Finite Element) models for the dynamic analysis of engineering structures, prior to these models being used for optimisation of product design. Current model validation methodology is confined to linear models and is primarily concerned with (i) correcting inaccurate model parameters and (ii) ensuring that sufficient elements are included for these cases, using measured data. Basic experience is that this works quite well, largely because the weaknesses in the models are relatively sparse and, as a result, are usually identifiable and correctable. The current state-of-the-art in linear model validation has contributed to an awareness that residual errors in FE models are increasingly the consequence of some unrepresented nonlinearity in the structure. In these cases, additional, higher order parameters are required to improve the model so that it can represent the nonlinear behaviour. This is opposed to the current practice of simply refining the mesh. Again, these nonlinear features are generally localised, and are often associated with joints. We seek to provide a procedure for extending existing modal testing to enable these nonlinear elements to be addressed using current nonlinear identification methods directed at detection, characterisation, location and then quantification – in order to enhance the elements in an FE model as necessary to describe nonlinear dynamic behaviour. Emphasis is placed on the outcome of these extended methods to relate specifically to the physical behaviour of the relevant components of the structure, rather than to the nonlinear response characteristics that are the result of their presence.

AB - Modal testing is widely used today as a means of validating theoretical (Finite Element) models for the dynamic analysis of engineering structures, prior to these models being used for optimisation of product design. Current model validation methodology is confined to linear models and is primarily concerned with (i) correcting inaccurate model parameters and (ii) ensuring that sufficient elements are included for these cases, using measured data. Basic experience is that this works quite well, largely because the weaknesses in the models are relatively sparse and, as a result, are usually identifiable and correctable. The current state-of-the-art in linear model validation has contributed to an awareness that residual errors in FE models are increasingly the consequence of some unrepresented nonlinearity in the structure. In these cases, additional, higher order parameters are required to improve the model so that it can represent the nonlinear behaviour. This is opposed to the current practice of simply refining the mesh. Again, these nonlinear features are generally localised, and are often associated with joints. We seek to provide a procedure for extending existing modal testing to enable these nonlinear elements to be addressed using current nonlinear identification methods directed at detection, characterisation, location and then quantification – in order to enhance the elements in an FE model as necessary to describe nonlinear dynamic behaviour. Emphasis is placed on the outcome of these extended methods to relate specifically to the physical behaviour of the relevant components of the structure, rather than to the nonlinear response characteristics that are the result of their presence.

KW - Modal testing

KW - Model updating

KW - Model upgrading

KW - Model validation

KW - Nonlinear structures

U2 - 10.1016/J.YMSSP.2016.04.012

DO - 10.1016/J.YMSSP.2016.04.012

M3 - Article

VL - 84

SP - 97

EP - 115

JO - Mechanical Systems and Signal Processing

T2 - Mechanical Systems and Signal Processing

JF - Mechanical Systems and Signal Processing

SN - 0888-3270

IS - Part B

ER -