Details
| Original language | English |
|---|---|
| Pages (from-to) | 1313-1330 |
| Number of pages | 18 |
| Journal | Structural health monitoring |
| Volume | 17 |
| Issue number | 5 |
| Publication status | Published - 29 Nov 2017 |
Abstract
This article introduces a new adaptive two-step optimization algorithm for finite element model updating with special emphasis on damage localization at supporting structures of offshore wind turbines. The algorithm comprises an enhanced version of the global optimization algorithm simulated annealing, the simulated quenching method that approximates an initial guess of damage localization. Subsequently, sequential quadratic programming is used to compute the final solution adaptively. For the correlation of numerical model and measurement data, both a measure based on eigenfrequencies and mode shapes and a measure employing time series are implemented and compared with respect to their performance for damage localization. Phase balance of the time signals is achieved using cross-correlation. The localization problem is stated as a minimization problem in which the measures are used in time and modal domain as the objective function subject to constraints. Furthermore, the objective function value of the adjusted model is used to distinguish correct from wrong solutions. The functionality is proven using a numerical model of a monopile structure with simulated damage and a lab-scaled model of a tripile structure with real damage.
Keywords
- damage localization, Model updating, sequential quadratic programming, simulated quenching
ASJC Scopus subject areas
- Biochemistry, Genetics and Molecular Biology(all)
- Biophysics
- Engineering(all)
- Mechanical Engineering
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: Structural health monitoring, Vol. 17, No. 5, 29.11.2017, p. 1313-1330.
Research output: Contribution to journal › Article › Research
}
TY - JOUR
T1 - A two-step approach to damage localization at supporting structures of offshore wind turbines
AU - Schröder, Karsten
AU - Gebhardt, Cristian Guillermo
AU - Rolfes, Raimund
N1 - Funding information: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article:The experiments at the Tripile structure were carried out in the European Union Seventh Framework Programme (FP7) project ‘‘Marinet’’ under grant agreement no. 262552. The methods were developed within the scope of the research project ‘‘Ventus efficiens’’ funded by the Lower Saxony Ministry of Science and Culture under FKZ ZN3024. The authors would like to thank Rick Damiani, Fabian Wendt, and Jason Jonkman of the National Renewable Energy Laboratories in Golden, Colorado, for their extraordinary support, consultation, and hospitality. Their input and guidance in the use of the software FAST and in the theory of time series comparison was very helpful for the development of this work. The authors sincerely acknowledge support to these projects by the Lower Saxony Ministry of Science and Culture and the European Union. The authors would like to thank Rick Damiani, Fabian Wendt, and Jason Jonkman of the National Renewable Energy Laboratories in Golden, Colorado, for their extraordinary support, consultation, and hospitality. Their input and guidance in the use of the software FAST and in the theory of time series comparison was very helpful for the development of this work. The authors sincerely acknowledge support to these projects by the Lower Saxony Ministry of Science and Culture and the European Union. The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article:The experiments at the Tripile structure were carried out in the European Union Seventh Framework Programme (FP7) project “Marinet” under grant agreement no. 262552. The methods were developed within the scope of the research project “Ventus efficiens” funded by the Lower Saxony Ministry of Science and Culture under FKZ ZN3024.
PY - 2017/11/29
Y1 - 2017/11/29
N2 - This article introduces a new adaptive two-step optimization algorithm for finite element model updating with special emphasis on damage localization at supporting structures of offshore wind turbines. The algorithm comprises an enhanced version of the global optimization algorithm simulated annealing, the simulated quenching method that approximates an initial guess of damage localization. Subsequently, sequential quadratic programming is used to compute the final solution adaptively. For the correlation of numerical model and measurement data, both a measure based on eigenfrequencies and mode shapes and a measure employing time series are implemented and compared with respect to their performance for damage localization. Phase balance of the time signals is achieved using cross-correlation. The localization problem is stated as a minimization problem in which the measures are used in time and modal domain as the objective function subject to constraints. Furthermore, the objective function value of the adjusted model is used to distinguish correct from wrong solutions. The functionality is proven using a numerical model of a monopile structure with simulated damage and a lab-scaled model of a tripile structure with real damage.
AB - This article introduces a new adaptive two-step optimization algorithm for finite element model updating with special emphasis on damage localization at supporting structures of offshore wind turbines. The algorithm comprises an enhanced version of the global optimization algorithm simulated annealing, the simulated quenching method that approximates an initial guess of damage localization. Subsequently, sequential quadratic programming is used to compute the final solution adaptively. For the correlation of numerical model and measurement data, both a measure based on eigenfrequencies and mode shapes and a measure employing time series are implemented and compared with respect to their performance for damage localization. Phase balance of the time signals is achieved using cross-correlation. The localization problem is stated as a minimization problem in which the measures are used in time and modal domain as the objective function subject to constraints. Furthermore, the objective function value of the adjusted model is used to distinguish correct from wrong solutions. The functionality is proven using a numerical model of a monopile structure with simulated damage and a lab-scaled model of a tripile structure with real damage.
KW - damage localization
KW - Model updating
KW - sequential quadratic programming
KW - simulated quenching
UR - http://www.scopus.com/inward/record.url?scp=85053689226&partnerID=8YFLogxK
U2 - 10.1177/1475921717741083
DO - 10.1177/1475921717741083
M3 - Article
AN - SCOPUS:85053689226
VL - 17
SP - 1313
EP - 1330
JO - Structural health monitoring
JF - Structural health monitoring
SN - 1475-9217
IS - 5
ER -