Damage detection of jacket type offshore platforms using rate of signal energy using wavelet packet transform

doi:10.1016/j.marstruc.2015.10.003

Marine Structures

Volume 45, January 2016, Pages 1-21

https://doi.org/10.1016/j.marstruc.2015.10.003 Get rights and content

Highlights

•
We build a prototype scaled offshore platform (two support condition) subjected to dynamic load.
•
We use rate of signal energy (RSE) of wavelet packet transform as a damage index.
•
We use measured accelerations of different sensors from experiments to compute RSE for all of damage scenarios.
•
It was observed that RSE has desirable value for sensors near location of damage and not significant value for other sensors.
•
By using proposed method and continuous monitoring, damage detection can be predicated with reasonable accuracy.

Abstract

Steel jacket type offshore platforms are highly susceptible to damage because of the severe environmental condition, therefore ensuring its proper performance and detection of probable damage is very important and undeniable. In this paper, by studying on experimental results of a prototype scaled offshore platform subjected to dynamic loading for different damage conditions, efficiency of Rate of Signal Energy using Wavelet Packet Transform for damage detection of this kind of structures has been carried out. Experiments results of two support conditions including hinge based and pile supported cases have been used. Wavelet packet analysis was used to determine the location of damage for different damage scenarios. The results demonstrate that proposed method can predict location of damage accurately. It can be concluded that this method can be used for prediction of damages similar to damages considered in this paper by using a continuous measurements and monitoring of platform responses.

Introduction

Integrity assessment of existing offshore platforms is necessary to ensure fitness for purpose of the structure in view of severe environment of sea, fatigue, high corrosive zone, scouring and vessels impact. Result of structural health monitoring of offshore platforms can be integrated into integrity assessment process of structures based on reliable and unbiased information. Objective data from continuous monitoring of offshore platforms increases quality of decisions and improve maintenance and repair strategies. By using structural health monitoring of offshore platforms, prediction of possible damages and determination of structure's health can be achieved which is valuable for such an important structures. Several structural health monitoring techniques have been used in the past for offshore platforms.

Kim and Stubbs (1995) [1] presented an algorithm to detect and locate damage in jacket-type offshore structures which only post-damage modal parameters was available for few modes of vibration. Researcher formulated a theory of damage localization and severity estimation and a method to identify baseline modal parameters of jacket-type offshore structures. They demonstrated the feasibility of the damage detection algorithm by using a numerical example of a jacket-type offshore platform with limited modal information. Nichols (2003) [2] explored the role of ambient excitation and empirical modeling in detecting damage in offshore structure. They excited two models of an articulated offshore structure. The resulting prediction error was increased when damage incurred. They concluded that this technique was effective for diagnosing the presence. Yang et al. (2004) [3] presented a newly developed damage localization method. This method was based on decomposing the modal strain energy. The method needed a small number of mode shapes identified from damaged and undamaged offshore platforms. They demonstrated that this method was capable of localizing damage for template offshore structures. Shi et al. (2007) [4] proposed a damage detection algorithm based on partial measurement of vibration for offshore jacket platforms. They demonstrated that presented algorithm was feasible and robust against identification error of baseline structure.

Yuan-sheng and Zhen (2008) [5] used time-domain data for detecting damage in offshore platform structures. They presented a new method that used time-domain response data under random loading. They concluded that the use of a few sensors' acceleration history data was capable of detecting damage efficiently and increasing in the number of sensors improved the diagnosis success rate. Elshafey et al. (2010) [6] investigated a combined method of random decrement signature and neural networks to detecting damage in offshore jacket platform subjected to random loads. They concluded that this method can be used to discover any changes in the shape of damage index. Mojtahedi et al. (2011) [7] developed a robust SHM method for offshore jacket platform using model updating and fuzzy logic system. They demonstrated that this technique was shown to be effective for diagnosing the presence of degradation and quantify it. Agrarian et al. (2009) [8] predicted damage location of jacket type offshore platforms by using change in modal strain energy ratio (MSER) of the elements. Compare to the method used in this paper, the accuracy of MSER is reduced when damage rates are low or several damaged elements exist at various locations of structure. Moreover application of this method for existing structures due to insufficient information about its elements properties may be limited.

In this paper by using experimental results of dynamic response of prototype offshore platform, suitability of “rate of signal energy using wavelet packet transform method” was studied for predicting damage location. Wavelet analysis is one of the most important methods used in structural damage detection. Signal analysis using wavelet derived from structural response can characterize the heterogeneity of the signal. Typical wavelet analysis has some disadvantages which has no benefit in the analysis of structural signals. Wavelet transform is capable of detecting location and time of occurrence with highlighting abnormal vibration in signal. Hence, wavelet analysis can detect structural damages by creating disturbance in the vibration signal. However, sensitivity of such an analysis to structural connections and boundary conditions can create disturbance in signals which makes problem to diagnose the damage.

Wavelet Packet Transform (WPT) is a kind of wavelet transform eliminates disadvantages of simple wavelet transform. Wavelet packet transform is a developed state of simple wavelet transform that makes complete separation level of the signal. This transform is composed of a linear combination of simple wavelet function. Therefore, wavelet packet transform can be expressed as steady and unsteady characteristics of the signal with arbitrary frequency-time resolution. The wavelet packet transform is a mathematical tool that has a special advantage over the traditional Fourier transform in analyzing non-stationary signals. It adopts redundant basis functions and hence can provide an arbitrary time-frequency resolution.

Yen and Lin (2000) [9] investigated feasibility of using continuous wavelet transform for signal vibration. Researchers defined wavelet packet energy index technique and concluded that the direct calculation energy of node in wavelet packet coefficient to classification is stronger. Proposed method which combined a wavelet-packet-based feature extractor and neural network classifier was tested on a Westland data set. Sun and Chang (2002) [10] proposed a wavelet packet transform-based method for the damage assessment of structures. Proposed method decomposes measured signals from structure into wavelet packet components. Then, component energies need to be calculated and to be used as inputs into neural network models for damage localization, occurrence, and severity. This method was applied to a three-span continuous bridge under impact load previously by Sun and Chang. They concluded that component energies are sensitive to structural damage and can be used for damage assessment of structures.

Sun and Chang (2004) [11] developed a statistical pattern classification method based on wavelet packet transform for structural health monitoring. In this method, measured signals of an excited structure were decomposed into wavelet packet components. Then signal energies were calculated in which small components were discarded and the remaining components were defined as a novel condition index namely wavelet packet signature (WPS). They defined two damage indicators to lump the discriminate extracted WPS information and thresholds for damage alarming. Developed method was applied to a cantilever I beam. It was concluded that proposed method is good for online structural health monitoring. Han, Ren and Sun (2005) [12] proposed “wavelet packet rate index (WPERI) for damage identification of steel beams. They concluded that the WPERI is a suitable index to identify local damage. Shahverdi et al. (2013) [13] used “Detail Signal Energy Rate Index (DSERI)” for damage assessment of Jacket type Offshore Platform.

In this paper, rate of signal energy using wavelet packet transform is proposed to detect damages in jacket type offshore platforms. In this regards, experimental dynamic response of a scaled steel jacket type offshore platform was used to calculate the amount of rate of signal energy for defined node. Associated signal energy rates values were reviewed to identify the location of damages.

There are significant studies using numerical or experimental studies on damage detection of the important infrastructures. For damage detection purposes, similar experimental studies which considered semi-real conditions and different boundary condition (soil-pile interaction and hinge-based) can be hardly found. Proposed method of this paper for damage detection has also advantages of using small and limited number of sensors to identify damage zones, however most of previous methods need more sensors in general and specific sensors near or at the damage location. In addition, most of vibration-based damage detection methods require modal characteristics that are obtained from measured signals through the system identification techniques. However, the modal properties such as natural frequencies and mode shapes are not such a good sensitive indication of structural damage.

The rate of signal energy (RSE) method is also an applicable method that can detect damages immediately after damage occurrence. Compared to other methods of damage detection, RSE method is not sensitive to boundary conditions, connection types, type of applied loads and damages intensity. Dynamic response of the real platform in sea state from forced or ambient vibration tests can also be used as an input for this method. In this paper, instead of using ambient vibration response in a real condition of the sea state, response of template platform from forced vibration was used.

Section snippets

Wavelet packet analysis

Wavelet packet decomposition (WPD) which is called as optimal sub band tree structuring (SB-TS) is a wavelet transform where signals are passed through more filters than the standard wavelet transform [17], [19]. In the wavelet packet decomposition, both the detail (1-D case and 2-D case) and approximation coefficients are decomposed to create the full binary tree. Each m level of decomposition the WPD generates 2^m different sets of coefficients. Wavelet packet function have three indices, $η_{b, c}^{a}$

Brief review of experimental studies

In this paper to evaluate efficiency of proposed method for damage detection of jacket type offshore platforms, result of experimental study on prototype scaled model of recently installed offshore platform in Persian Gulf have been used. Studied jacket is a welded-steel space frame and six legged which has horizontal and vertical braces and it has four bays, three levels deck and eight skirt piles (Fig. 1). The geometric scale is chosen 1:15, because available pipes and laboratory facilities

Application of rate of signal energy using wavelet packet transform to detect damage of a sample steel jacket type offshore platform

Dynamic responses of platforms for intact and damaged conditions have been analyzed using rate of signal energy using wavelet packet transform. In order to detect the location of damages, two sets of signal are required: signals from intact and damaged condition. Rate of signal energy was calculated for each of the nodes that sensors records were exist. From Analysis results, greater RSE represents damage location. Analysis results for both of boundary conditions have been presented hereafter.

Conclusion

In this paper, rate of signal energy using wavelet packet transform was proposed as an index for damage detection of jacket type offshore platforms. In this regards, measured accelerations of a prototype scaled platform for both hinge-based and pile-supported conditions have been used to investigate efficiency of proposed method for its application in offshore platforms. Measured accelerations of different sensors from experiments were used to compute rate of signal energy (RSE) for all of

Acknowledgement

The research employed herein was sponsored under POGC (Pars Oil and Gas Company) project No. 132 “Investigation of Structural Health Monitoring of Steel Jacket Offshore Platforms”. The financial support of POGC is gratefully acknowledged.

References (21)

J.M. Nichols
Structural health monitoring of offshore structures using ambient excitation
Appl Ocean Res
(2003)
A. Mojtahedi et al.
Developing a robust SHM method for offshore jacket platform using model updating and fuzzy logic system
Appl Ocean Res
(2011)
J.G. Han et al.
Wavelet packet based damage identification of beam structures
Int J Solids Struct
(2005)
J.T. Kim et al.
Damage detection in offshore jacket structures from limited modal information
Int J Offsh Polar Eng
(1995)
H. Yang et al.
Damage localization for offshore structures by modal strain energy decomposition method
Am Control Conf Boston
(July 2004)
X. Shi et al.
A damage detection algorithm based on partial measurement for offshore jacket platforms
C. Yuan-Sheng et al.
Detecting damage to offshore platform structures using the time-domain data
J Mar Sci Appl
(2008)
A.A. Elsafey et al.
Damage detection in offshore structures using neural networks
Mar Struct
(2010)
B. Asgarian et al.
Damage in jacket type offshore platforms using modal strain energy
Struct Eng Mech
(2009)
G.G. Yen et al.
Wavelet packet feature extraction for vibration monitoring
IEEE Trans Ind Electron
(2000)

There are more references available in the full text version of this article.

Cited by (84)

Structural health monitoring on offshore jacket platforms using a novel ensemble deep learning model
2024, Ocean Engineering
Monitoring health condition of offshore jacket platforms is crucial to prevent unexpected structural damages, where a prevailing challenge involves translating available feature information into structural damage patterns. Although the artificial neural network (ANN) models are popular in addressing this challenge, they often fail to capture the temporal correlations between the feature information and the damage patterns, which reduce their capability for discovering the laws governing the structural damage detection. To bridge this research gap, this study proposes a novel ensemble deep learning model to enhance the temporal feature extraction to improve the damage pattern identification. In this approach, a one-dimensional Convolutional Neural Network (CNN) extracts the spatiotemporal features from the structural vibration measurements. Simultaneously, a SENet attention mechanism is introduced to select the most informatic features. Subsequently, a bidirectional long short-term memory network (BiLSTM) is employed to learn the mapping between the extracted features and the structural damage patterns. Furthermore, the particle swarm optimization (PSO) algorithm is used to optimize the BiLSTM hyperparameters to enhance its stability and reliability. Both simulations and experiments are carried out to collect the vibration responses of the offshore jacket structure in different damage scenarios. The analysis results demonstrate that the proposed method produces remarkable improvement with respect to the accuracy and robustness in identifying the structural damages when compared with the ANNs. The overall detection accuracy of the proposed CNN-BiLSTM-Attention ensemble model is beyond 95%, which provides strong applicability to practical structural health monitoring of offshore platforms.
Time-varying damage detection in beam structures using variational mode decomposition and continuous wavelet transform
2024, Construction and Building Materials
Civil engineering structures in operation are likely to suffer damage. During the service life, structural damage evolves gradually from minor to severe. However, most current studies focus on damage localization and quantification of damage severity, and hence little attention has been paid to the detection of time-varying damage. This paper aims to propose a new approach for tracking the damage evolution of beam structures. In this approach, the wavelet threshold method is used at first to denoise the response signal. After that, the variational mode decomposition (VMD) is introduced to decompose the denoised response signal into several mono-components adaptively. A proposed index of wavelet total energy change (WTEC) is then applied to the new decomposed signals to localize damages of beam structures. On a basis of this, a time-varying damage index called wavelet energy change ratio (WECR) is established via the continuous wavelet transform and time window techniques and then applied to the response at the damaged position for tracking damage evolution. The proposed method is verified via a numerical example of a simply supported beam (its length and area of cross section are 5 m and 0.04 m², respectively) with a sudden and a linear stiffness reduction under 1940 El Centro ground acceleration record. In addition, an experiment on a 10-meters-long steel bridge with abrupt stiffness reduction under a moving load is investigated to demonstrate the effectiveness and accuracy of the proposed time-varying damage detection method. The results show that the index of WTEC is able to localize damages of beam structures effectively whatever they are a single damage position or multiple damage positions. Moreover, the damage evolution can be tracked by the proposed index of WECR accurately even though random noises and end effects pose a great impact on the time-varying damage detection results.
Vibration-based damage localization in 3D sandwich panels using an irregularity detection index (IDI) based on signal processing
2024, Measurement: Journal of the International Measurement Confederation
As 3D sandwich panels have become prevalent in construction projects, developing structural damage detection approaches is of great significance to ensure their integrity. This study aims to introduce a damage identification method to diagnose damaged regions in 3D sandwich panels using modal dynamic data. To accomplish this, an Irregularity Detection Index (IDI) is defined based on the horizontal, vertical, and diagonal detail coefficients of the wavelet analysis of an input signal consisting of primary and secondary mode shapes. The relative maximum values of irregularities on the plotted IDI signal indicate damaged locations. It is also shown that with an increase in the damage severity at a damaged location, its corresponding IDI_max increases, while changes in IDI_max for other undamaged locations are negligible. In fact, the detection or non-detection of a damaged location does not affect the damage localization process in other locations.
Shaking table test and numerical analysis of dynamic response and damage mechanism of a deposit slope with a weak interlayer reinforced by a pile–anchor structure
2024, Structures
The study of the cumulative damage characteristics of pile–anchor reinforced slopes under multiple earthquakes is important for evaluating their long-term stability. In this study, a large shaking table test was designed and completed for deposit slope with a weak interlayer reinforced by a pile–anchor structure. The cumulative damage characteristics and evolution law of the slope were determined using a transfer function and wavelet packet decomposition. Numerical simulations were conducted to extend the scope of this study. The modeling tests showed that the internal acceleration of a slope under seismic action has an elevation amplification effect and a converging surface effect. Transfer function calculations showed that the natural frequency and dynamic shear modulus ratio of the slope decrease gradually with an increase in the seismic action, and the damping ratio increases gradually with the enhancement of the seismic action. The wavelet packet transform revealed that with an increase in seismic action, the energy consumed at the slope location increases, resulting in a decrease in the energy amplification factor. Numerical calculations showed that the shear strain develops from the top of the slope to the interior under seismic action. The permanent displacement of the slope increases with increasing elevation.
Damage detection of offshore jacket structures using structural vibration measurements: Application of a new hybrid machine learning method
2023, Ocean Engineering
The artificial intelligence (AI) technologies, such as meta-heuristic computing and deep learning, have provides solid technical support for structural health monitoring (SHM) of offshore jackets. In this paper, a physics-enhanced AI method based on the parametric damage identification is developed for SHM of the offshore jacket structures. In this new method, a hybrid kernel function-based kernel extreme learning machine (HKELM) is proposed to construct an AI structure to enhance the SHM detection capacity on the structural modal parameters extracted by the parametric damage identification technique. Simulation analysis is carried out to verify the feasibility of the HKELM-based method using the response signals of a jacket structure under impact force, and the result demonstrates good damage location capability of the proposed method. Furthermore, to select proper parameters for the HKELM, the whale optimization algorithm (WOA) is applied to optimize the values of the regularization coefficient and kernel parameter array. Then, the wavelet denoising (WD) is introduced to preprocess the vibration signals to improve the damage detection ability of the WOA-HKELM. Lastly, experimental tests are performed to validate the effectiveness of the proposed method in utilizing the structural modal parameters for identifying the structural damage. The analysis results illustrate that the proposed method produces satisfactory damage location ability under the influence of actual noise in the vibration signals. Meanwhile, this method has broad application prospects in the SHM of other offshore structures.
Influence of mechanical faults in pitch, brake, controller systems on fragility of offshore wind turbine under aero-hydro loadings
2023, Ocean Engineering
Offshore wind turbines (OWTs) have undergone unprecedented development in recent years, and during their service life the control and mechanical systems faults have emerged one of the primary reasons to cause downtime of OWTs. This paper addresses the impact of mechanical faults in control systems on the structural damage risk under wind and wave loadings. Numerical simulations implemented by a combination of finite element simulation software are conducted to simulate nonlinear behaviors of support structural system with and without mechanical faults. Incremental dynamic analyses are carried out for three cases: (1) intact OWTs, (2) OWTs with only pitch system fault, and (3) OWTs with simultaneous faults at pitch and brake systems. The IDA results and structural damage risks of blades and support structural system under different cases are comprehensively analyzed and discussed. The results indicated that the pitch system fault results in a minor increase in the damage risk of OWTs in parked condition, while the presence of both pitch and brake systems fault or controller fault leads to a significant rise in the structural risk and even structural failure due to tower buckling and blade-tower strike.

View all citing articles on Scopus

View full text

Damage detection of jacket type offshore platforms using rate of signal energy using wavelet packet transform

Highlights

Abstract

Introduction

Section snippets

Wavelet packet analysis

Brief review of experimental studies

Application of rate of signal energy using wavelet packet transform to detect damage of a sample steel jacket type offshore platform

Conclusion

Acknowledgement

Appl Ocean Res

Appl Ocean Res

Int J Solids Struct

Damage detection in offshore jacket structures from limited modal information

Int J Offsh Polar Eng

Damage localization for offshore structures by modal strain energy decomposition method

Am Control Conf Boston

A damage detection algorithm based on partial measurement for offshore jacket platforms

Detecting damage to offshore platform structures using the time-domain data

J Mar Sci Appl

Damage detection in offshore structures using neural networks

Mar Struct

Damage in jacket type offshore platforms using modal strain energy

Struct Eng Mech

Wavelet packet feature extraction for vibration monitoring

IEEE Trans Ind Electron