Details
| Original language | English |
|---|---|
| Title of host publication | Structural Health Monitoring 2019 |
| Subtitle of host publication | Enabling Intelligent Life-Cycle Health Management for Industry Internet of Things (IIOT) - Proceedings of the 12th International Workshop on Structural Health Monitoring |
| Editors | Fu-Kuo Chang, Alfredo Guemes, Fotis Kopsaftopoulos |
| Pages | 3441-3449 |
| Number of pages | 9 |
| ISBN (electronic) | 9781605956015 |
| Publication status | Published - 2019 |
| Event | 12th International Workshop on Structural Health Monitoring: Enabling Intelligent Life-Cycle Health Management for Industry Internet of Things (IIOT), IWSHM 2019 - Stanford, United States Duration: 10 Sept 2019 → 12 Sept 2019 |
Abstract
The central concept of vibration-based structural health monitoring (SHM) is the computation of damage sensitive features like natural frequencies or model residues. These parameters are typically observed and evaluated from a statistical point of view following the statistical pattern recognition paradigm. It is of common knowledge that those features are not merely effected by local structural alterations, but also by environmental or operational conditions (EOC). To distinguish or eliminate these, many data normalization strategies exist for linear and even non-linear cases. Besides damage detection under environmental and operational variability, the determination of actual damage positions is desirable for large and remote engineering structures such as onshore and offshore wind turbines or bridges. The recently introduced state projection estimation error (SP2E) method appears to be a promising approach towards damage localization of mechanical systems. This procedure is purely data-driven (output-only) and therefore does not rely on updated numerical physical models (e.g. FE models). It is based on identified, parametric, linear time-invariant (LTI) systems, state estimation and advanced projection techniques. The method significantly differs from conventional modal approaches e.g. the comparison of mode shapes or curvatures. Although laboratory experiments proved a high sensitivity against local structural changes for some systems, damage localization with SP2E under varying conditions has so far not been investigated. This article deals with this important issue. Therefore, simulations of a linear parameter varying (LPV) system were conducted in the healthy and locally altered state. The principal component analysis (PCA) was then applied to distinguish between local and global changes and successfully locate the damage positions. As an important step for validation, these investigations were complemented by a laboratory test. This research helps to improve the understanding of the sensitivity of the SP2E method towards varying environmental and operational conditions. Additionally the applicability of linear data normalization strategies with respect to damage localization was tested. It therefore lays a foundation for future field applications.
ASJC Scopus subject areas
- Computer Science(all)
- Computer Science Applications
- Health Professions(all)
- Health Information Management
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
Structural Health Monitoring 2019: Enabling Intelligent Life-Cycle Health Management for Industry Internet of Things (IIOT) - Proceedings of the 12th International Workshop on Structural Health Monitoring. ed. / Fu-Kuo Chang; Alfredo Guemes; Fotis Kopsaftopoulos. 2019. p. 3441-3449.
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - Damage localization with SP2E under changing conditions
AU - Wernitz, Stefan
AU - Pache, Dorian
AU - Grießmann, Tanja
AU - Rolfes, Raimund
PY - 2019
Y1 - 2019
N2 - The central concept of vibration-based structural health monitoring (SHM) is the computation of damage sensitive features like natural frequencies or model residues. These parameters are typically observed and evaluated from a statistical point of view following the statistical pattern recognition paradigm. It is of common knowledge that those features are not merely effected by local structural alterations, but also by environmental or operational conditions (EOC). To distinguish or eliminate these, many data normalization strategies exist for linear and even non-linear cases. Besides damage detection under environmental and operational variability, the determination of actual damage positions is desirable for large and remote engineering structures such as onshore and offshore wind turbines or bridges. The recently introduced state projection estimation error (SP2E) method appears to be a promising approach towards damage localization of mechanical systems. This procedure is purely data-driven (output-only) and therefore does not rely on updated numerical physical models (e.g. FE models). It is based on identified, parametric, linear time-invariant (LTI) systems, state estimation and advanced projection techniques. The method significantly differs from conventional modal approaches e.g. the comparison of mode shapes or curvatures. Although laboratory experiments proved a high sensitivity against local structural changes for some systems, damage localization with SP2E under varying conditions has so far not been investigated. This article deals with this important issue. Therefore, simulations of a linear parameter varying (LPV) system were conducted in the healthy and locally altered state. The principal component analysis (PCA) was then applied to distinguish between local and global changes and successfully locate the damage positions. As an important step for validation, these investigations were complemented by a laboratory test. This research helps to improve the understanding of the sensitivity of the SP2E method towards varying environmental and operational conditions. Additionally the applicability of linear data normalization strategies with respect to damage localization was tested. It therefore lays a foundation for future field applications.
AB - The central concept of vibration-based structural health monitoring (SHM) is the computation of damage sensitive features like natural frequencies or model residues. These parameters are typically observed and evaluated from a statistical point of view following the statistical pattern recognition paradigm. It is of common knowledge that those features are not merely effected by local structural alterations, but also by environmental or operational conditions (EOC). To distinguish or eliminate these, many data normalization strategies exist for linear and even non-linear cases. Besides damage detection under environmental and operational variability, the determination of actual damage positions is desirable for large and remote engineering structures such as onshore and offshore wind turbines or bridges. The recently introduced state projection estimation error (SP2E) method appears to be a promising approach towards damage localization of mechanical systems. This procedure is purely data-driven (output-only) and therefore does not rely on updated numerical physical models (e.g. FE models). It is based on identified, parametric, linear time-invariant (LTI) systems, state estimation and advanced projection techniques. The method significantly differs from conventional modal approaches e.g. the comparison of mode shapes or curvatures. Although laboratory experiments proved a high sensitivity against local structural changes for some systems, damage localization with SP2E under varying conditions has so far not been investigated. This article deals with this important issue. Therefore, simulations of a linear parameter varying (LPV) system were conducted in the healthy and locally altered state. The principal component analysis (PCA) was then applied to distinguish between local and global changes and successfully locate the damage positions. As an important step for validation, these investigations were complemented by a laboratory test. This research helps to improve the understanding of the sensitivity of the SP2E method towards varying environmental and operational conditions. Additionally the applicability of linear data normalization strategies with respect to damage localization was tested. It therefore lays a foundation for future field applications.
UR - http://www.scopus.com/inward/record.url?scp=85074268070&partnerID=8YFLogxK
U2 - 10.12783/shm2019/32505
DO - 10.12783/shm2019/32505
M3 - Conference contribution
AN - SCOPUS:85074268070
SP - 3441
EP - 3449
BT - Structural Health Monitoring 2019
A2 - Chang, Fu-Kuo
A2 - Guemes, Alfredo
A2 - Kopsaftopoulos, Fotis
T2 - 12th International Workshop on Structural Health Monitoring: Enabling Intelligent Life-Cycle Health Management for Industry Internet of Things (IIOT), IWSHM 2019
Y2 - 10 September 2019 through 12 September 2019
ER -