Kalman Filter-Based Multitype Measurement Data Fusion for Stress Intensity Factor Evaluation in Marine Structures

Dai, Tong-Tong; Jia, Zi-Guang; Ren, Liang; Fu, Xing

doi:https://doi.org/10.1155/2023/2743309

Structural Control and Health Monitoring

On this page

Abstract Introduction Numerical Simulation Conclusion Data Availability Conflicts of Interest Acknowledgments References Copyright Related Articles

Research Article | Open Access

Volume 2023 | Article ID 2743309 | https://doi.org/10.1155/2023/2743309

Kalman Filter-Based Multitype Measurement Data Fusion for Stress Intensity Factor Evaluation in Marine Structures

Tong-Tong Dai,¹Zi-Guang Jia,²Liang Ren,¹and Xing Fu¹

Academic Editor: Yi-Qing Ni

Received08 Oct 2022

Revised01 Jan 2023

Accepted03 Jan 2023

Published09 Feb 2023

Abstract

Fracture is the main form of marine structural failure due to the action of waves and oceans. It is necessary to monitor the initiation and extension of cracks in a timely and accurate manner, and the stress intensity factor (SIF) is an important indicator for analysing the propagation of cracks. Based on existing research, this paper combines the calculation method of the SIF based on a single strain gauge (SSG) and maximum crack opening displacement (CMOD) and proposes a method based on the Kalman filter (KF). Numerical simulation and experimental verifications of this method have been carried out. The results show that the method can be used for a variety of configurations of precracked specimens. The KF-based method obtains more accurate results than the SSG-based and CMOD-based methods, and the relative error (RE) is not greater than 1.2%.

1. Introduction

The service status and failure modes of metallic marine structures are usually complex due to the action of alternating loads such as waves. The frequency and cost of maintenance greatly increase as the structure ages, and performance significantly decreases [1–6]. For example, during the inspection and investigation of a semisubmersible platform in the South China Sea, cracks were observed in the corner of the living building, the connecting structure of the living building, and the related transverse brace, as shown in Figure 1. Cracks were detected in the four corners of the living building shortly after the platform was put into operation and still appeared and spread despite treatment. According to the investigation and research, the crack was caused by alternating loads or extreme loads. The specific causes need further analysis. Failure to detect and evaluate in time may reduce the local and overall bearing capacity of the structure. Therefore, it is necessary to monitor the structure, assess the structural condition, and take effective measures in a timely manner for the maintenance and extension of the service life of marine structures [7–10]. Fatigue fracture is the main form of metal structure failure, accounting for more than 90% of the failure modes. The impact on structural performance is small when the crack size is small, but with the cyclic action of loads, a crack may further expand. To predict the crack growth rate and fracture strength of cracked members, it is necessary to accurately analyse the stress state.

Irwin [11] proposed the concept of the SIF (K) based on Griffith’s work [12] in 1957, which is a metric parameter of crack severity. The SIF is related to the size, stress, and geometry, but is independent of the material. In addition, the SIF represents the strength of the stress-strain field at the crack tip and shows the ability of crack propagation, which is widely utilized in the calculation of fatigue strength and crack propagation of cracked structures [13–15]. In the definition of (K), it is assumed that the mechanical behaviour of the material is linear elastic, and Hooke’s law is followed. Linear elastic fracture mechanics are adopted. The SIF can also be applied to estimate the crack growth life. Paris proposed an empirical formula for the crack growth rate: , where, a is the crack size, C and m are constants related to materials, N is the number of stress cycles, and is the variation range of the SIF. Therefore, determining the value of the SIF is necessary to assess the current state of the structure and predict the crack life. The calculation of the SIF includes numerical methods, analytical methods, and direct measurement methods. The analytical solution is the basis for the development of fracture mechanics, which is based on the basic equations of the stress field and displacement field near the crack tip and is derived by using mathematical and mechanical knowledge. Numerous numerical methods for the SIF, such as the stress method, finite element method (FEM) [13–18], boundary element method [19], and meshless method [20], have been proposed after years of research.

However, in practical engineering, the analytical method and numerical method will greatly reduce the calculation accuracy and cannot achieve long-term continuous monitoring of cracked structures due to the limitations of the structural form and boundary conditions. Therefore, it is necessary to develop a method to determine the SIF based on easily measured physical quantities, such as strain and displacement, to realize real-time online monitoring of structures.

For the SIF calculation based on the strain field near the crack tip, Dally and Sanford [21] derived the four-parameter solution of the SIF according to the Westergaard stress function and proposed the experimental method of measuring the SIF with a strain gauge. Some scholars have conducted further studies on the technology. Wei and Zhao [22] discovered that the SIF could be measured by two strain gauges via numerical and experimental studies. Sarangi et al. [23, 24] carried out theoretical research and experimental verification on the optimal strain gauge position and its importance in the accurate determination of the SIF. In terms of the calculation of the SIF based on the crack opening displacement (COD), Masksimenko and Tyagnii [25] proposed an experimental method to calculate the SIF and stress distribution along the crack line by using discrete displacement data of several points on the crack edge. Chen et al. [26] proposed a method to determine the SIF according to the CMOD and determined the applicable range of the method via finite element simulation. Then, the method for calculating the SIF with collinear double cracks and impulse load based on the CMOD was investigated [27, 28]. Saucedo-Mora et al. [29] proposed using digital image correlation (DIC) to measure the SIF of crack tips. The stress-strain field was obtained by identifying the crack tip and by measuring the full-field displacement; thus, the SIF and the energy release rate of each crack during shear and tensile tests were calculated. Chen and Qian [30] proposed a method to quantify the J integral of the through-thickness of planar strain samples by DIC and tested the constraint effect of DIC measurement. Then, the basic relationship between the average J value of the free surface and the full thickness is theoretically and experimentally established and verified [31]. However, the SIF determination method based on the strain is greatly affected by the position of the strain measurement points due to the large strain gradient near the crack tip. The CMOD-based method is proposed under the assumption of ideal boundary conditions, which will deviate in practical engineering due to the complex load form.

Currently, structural health monitoring (SHM) systems are deployed on large civil structures and typically contain various types of sensors to measure different types of responses of structures. Numerous studies relating to data fusion have investigated SHM in time and space [32–40]. Wu and Jahanshahi [41] reviewed data fusion methods for SHM and system identification. Park et al. [34] realized data fusion between vision-based displacement and acceleration by a complementary filter, which reduces vision-based displacement error and expands the feasible frequency range. This method accurately measures the displacement response in terms of noise level and viable frequency range. Jin et al. [39] adopted the traditional, multifidelity data fusion framework based on Gaussian regression to investigate the complementary data fusion of point strain sensors and distributed sensors and obtained an accurate strain distribution by combining their advantages. Qi et al. [40] proposed a model-based, inverse finite element algorithm by using two kinds of sensors to identify the crack length and carried out experimental verification. However, according to the investigation, there is no precedent for applying data fusion technology to SIF calculation.

The paper is aimed at combining the determination methods of SIF based on strain and CMOD and at adopting the data fusion framework based on KF to improve the estimation accuracy of the SIF. To verify the feasibility of the proposed method, numerical and experimental studies were carried out. As the study is aimed at the current state of cracks in the living building part of an offshore platform, the test pieces are steel plates. The structure of this paper is described as follows: the calculation method of the SIF based on the SSG and CMOD is deduced in Section 2; in Section 3, the principle of KF and the specific process employed in this study are introduced; in Section 4, numerical simulation is carried out for plates with different ratios of crack length to plate width, and the SIF is obtained by SSG-based, CMOD-based and KF-based methods and compared with real values; experiments on plates with prefabricated cracks in different configurations are introduced to verify the effectiveness of the proposed method.

2. Methodology

In this section, the method of SIF calculation by the SSG and CMOD is discussed, the calculation formula of SIF is derived, and then the basic method of classical KF modelling is introduced. According to the stress form of the crack, the propagation mode is divided into three types (Figure 2): opening crack (mode I), sliding crack (mode II), and tearing crack (mode III). In ocean engineering, the crack form is mainly the opening mode. Therefore, in this section, the calculation of the SIF of mode I is introduced.

(a)

(b)

(c)

2.1. SIF Determination Method Based on the SSG

The plane problem of elasticity, can be solved by the stress function , which is expressed as in polar coordinates and expressed as in the form of separated variables, satisfying the coordination equation . At one end of the finite crack shown in Figure 3, the origin of the coordinate system is selected at the crack tip. The axis is collinear with the crack. The strain gauge is placed on the line (“line ”) with angle with the axis, and the included angle between the strain gauge and the -axis is .

Considering the boundary condition, that is, the force on the crack surface is zero as follows:

The general solution is obtained as follows:

The stress function is obtained as follows:

The SIF is obtained by

The three terms of strain components and are obtained by Hooke’s law and the geometric equation

The strain component at the strain gauge position satisfies the following equation:

and can be eliminated when

When , is a constant when the material and load are determined. Taking the log of both sides of equation (9)that is, within the range of radial length satisfying equation (10), the SIF can be expressed as

2.2. SIF Determination Method Based on the CMOD

The stress function indicated by Westergaard is adopted in the model of an infinite plate with a penetrating crack under uniform tensile stress as follows:

The displacement is obtained as

If and , then

Among them,that is,

The polar coordinates of the CMOD are , where .

Therefore, the CMOD is .

For an infinite plate with a central penetrating crack, the SIF at the crack tip is under uniform tensile loading. Therefore, the SIF expressed by the CMOD is

3. Determination of the SIF Based on the KF

This section introduces the classical KF algorithm flow, and then introduces the specific process for determining SIF in the subsequent numerical simulation and experiment.

3.1. Kalman Filtering Model

The KF optimizes the state estimation of the system by two processes of “prediction” and “update.” The basic model of the KF will be introduced in this section [42].

Suppose that the discrete dynamic system is modelled aswhere is the system state at moment , and is the observed value at moment , that is, the measured value. is the state transition matrix, is the control input matrix, and is the observation matrix.

Since most systems are not strictly linear and time-varying and the structural system parameters may be uncertain, there will be a deviation in the estimated state value, which is characterized by process noise . For the state observation equation (18)-b, due to the existence of noise and interference signals or the error of the observation matrix, there will be a deviation between the observed value and the real value, which is characterized by . and are uncorrelated white noise with mean value of 0 and variances of and , respectively. The principle of the KF is to modify the state prediction value so that it is similar to the real value using the Kalman gain value.

The a priori estimate and measurement of the state as follows:

Superscripts “^” and “−” denote estimates and priors, respectively.

The posterior value is calculated from the prior value and measured value as follows:

The goal is to obtain the Kalman gain value , such that .

The posterior state error and prior state error are

Then, the variance matrix of the estimation error is

The with the smallest variance is obtained as follows:

Then, the error variance matrix is

Figure 4 summarizes the flow of KF.

3.2. SIF Determination Method Based on Data Fusion

The calculation process based on the KF in this study will be introduced in this section. This method includes the determination method of the SIF based on the SSG and CMOD. Therefore, the model of the system contains two observation equations, and the system variable is one-dimensional.

The model of the system iswhere is the process noise, and and represent the observation noise of two observation values, which are uncorrelated white noise with a mean value of 0 and variances of , , and . As there is no way to model the noise term during the modelling process, the Kalman filter process is generally divided into two parts: a priori estimation and correction.(1)The SIF at the current time is estimated based on the value of the previous step(2)The error covariance of the a priori estimation is calculated(3)The prior estimate is corrected, and the Kalman gain of the two measured values is calculated(4)The a priori estimation, Kalman gain, and two measured values are employed for optimal estimation(5)The error covariance is updated(6)If the termination conditions are met, the filtering process is ended. Otherwise, steps (1)–(5) are repeated (the termination condition is to reach the maximum filtering times in this study).

4. Numerical Simulation

As most marine structures are plates and shells, plates are selected as the research object in this paper. In this section, a plate with prefabricated cracks is numerically simulated, and the applicable radial length range of the measuring points in the SSG-based method is calculated. The SIF values are obtained by SSG-based, CMOD-based, and KF-based methods and compared with the real values.

4.1. Determination of the Radial Position of the Strain Gauge

As described in Section 2.1, the determination method of the SIF based on the SSG is not applicable to the whole “line ” that has angle with the -axis. The relationship between the strain component and the radial length satisfies equation (10), that is, the slope of curve is equal to −0.5.

In this section, the plate with prefabricated cracks shown in Figure 5 is simulated, where , the plate thickness , the elastic modulus , Poisson’s ratio , and can be calculated by equation (7).

Using , that is, , as an example, to obtain the exact value of the applicable range, the element on “line ” at both ends of the crack is refined, as shown in Figure 6. Figure 7 is the enlarged view of the crack tip. The applicable radial length of the SIF determination method based on the SSG for different values of is calculated.

A uniform tensile load is applied to the model, the coordinates and strain components of each node on “line ” are obtained, and the curve of is obtained. As shown in Figure 8, there is an obvious linear segment in the curve. A straight line with slope = −0.5 is inserted into the curve, and the range of radial length with an error less than 0.5% is obtained (i.e., the area between two dotted lines in Figure 8).

(a)

(b)

(c)

(d)

(e)

The minimum and maximum of the applicable radial length are listed in Table 1.

Research by Rosakis and Ravichandar [43] shows that the radial length from the strain measuring point to the crack tip should be greater than half of the plate thickness. Therefore, the minimum radial length should be larger between and .

4.2. Calculation of the SIF Based on the SSG and CMOD

Using the condition of as an example, the strain values within the applicable radial length (3.81–21.40 mm) are substituted into equation (11) for the SIF calculation. Selecting the SIF obtained by the extended finite element method (XFEM) as the real value, Figure 9 shows that the strain value within the applicable range can be used to obtain a more accurate result. During the whole loading process, the relative error (RE) of the SIF obtained by the single strain gauge at r = 12.58 mm is concentrated at 3.18%, while the result obtained by r = 23.05 mm is 15.00%, indicating that it is necessary to calculate the applicable radial length.

Next, the plates with prefabricated cracks of different are simulated, and the SIF by SSG-based and the CMOD-based methods are calculated. Table 2 shows the SIF for . The RE of the two methods is not greater than 8.5%.

Table 3 shows the RE of the SIF obtained by the two methods during the whole loading process when . The results obtained by the SSG-based method are concentrated at 3.2%, while those obtained by the CMOD-based method are concentrated at 3.8%.

To further improve the accuracy of SIF estimation, KF is used to fuse the SSG-based method and CMOD method, which will be introduced in Section 4.3.

4.3. Determination Method of SIF Based on the Kalman Filter

In this section, the KF is used for data fusion of the SSG-based method and CMOD-based method, and the principle of the KF is introduced in Section 3. The KF process contains two observations, and the number of iterations is 100. Similarly, using as an example, the SIF for each stress state is obtained by data fusion (subgraph of Figure 10). Figure 10 shows that in the initial stage of the iterative process, the RE of the SIF obtained by data fusion is larger, which gradually decreases as the number of iterations increases. The RE with is 0.14%.

The RE of the SIF obtained by the three methods (SSG, CMOD, and KF) with is listed in Table 4. The RE of the SIF estimation by the KF is not greater than 1%. For the SSG-based method, the RE may be caused by the influence of ambient noise, strain gradient, and the omission of higher order terms in the derivation process. For the CMOD-based method, the error may be caused by the sensor accuracy, the notion that the location of the sensor is not the CMOD position due to human operation and the omission of the higher order terms. For the KF-based method, since the method is based on the SSG and CMOD methods, the error is caused by the inaccuracy of the two methods and the selection of the noise variance.

4.4. Noise Influence

In the previous research, no noise is added to introduce additional errors to the strain data and crack opening displacement data. However, the influence of human operation and environmental noise cannot be disregarded in practical engineering. Therefore, in this section, different levels of Gaussian white noise are added to the original data to evaluate the robustness of KF.

Similarly, using as an example, Gaussian white noise with a level of 1%–5% is added to the simulated data. Figure 11 shows a comparison of the SIF obtained by the KF, SSG, and CMOD-based methods under the influence of a 5% noise level. During the whole process of loading, there is a large deviation between the SSG-based method and CMOD-based method with the real value due to the influence of noise. The result of the KF has an error in the initial stage, but with an increasing number of iterations, it basically coincides with the real value and is not affected by noise. Figure 12 shows the RE of the three methods during the whole process.

Table 5 shows the RE of the SIF obtained by the three methods under the influence of different levels of noise when . The RE of the SIF obtained by the KF is still at a low level, which indicates that the method based on data fusion can still obtain a more accurate SIF under the influence of noise.

5. Experimental Study

An experiment was designed to verify the reliability of the proposed method, including two working conditions: (1) central cracks subjected to uniform loads and (2) central cracks under nonuniform loads. The test was carried out on a universal testing machine using strain gauges and a displacement metre to obtain the changes in strain and maximum crack opening displacement during the loading process.

5.1. Experimental Device

The specimen used in the experiment is a steel plate with a prefabricated crack, the modulus of elasticity is 206 GPa, and Poisson’s ratio is 0.3. The width and length of the specimen are 200 mm and 300 mm. The test device is shown in Figure 13. The two ends of the plate are installed on the testing machine using T-shaped adapters. The position and direction of the strain gauge paste refer to the results calculated in Section 4.1. The loading method is hierarchical loading and the loading speed is 100 N/s. The load is maintained for 10 s every 4 kN, and the maximum loading force is 24 kN. The radial lengths of the strain gauges in the specimens are shown in Table 6, which are within the effective range calculated in Section 4.1.

5.2. Central Crack under Uniform Loads

The strain value and maximum crack opening displacement during the loading process were recorded, and the SIF obtained by the two methods was calculated. The empirical value is used as the real value and compared with the experimental results, whose calculation refers to equation as follows:where is the boundary correction coefficient, which is calculated using the empirical formula given by Tada et al. [44] as follows:

The value of SIF during the loading process with different is calculated and listed in Table 7. An approximate range of the force value in the holding phase is selected due to the fluctuation in the force value during the loading process.

The SIF obtained by the SSG, CMOD, and KF is compared with the true value, and the RE is calculated according to equation.where is the result obtained by the SSG, CMOD, and KF, and is the true value. The RE of cracks with different during the entire loading process is listed in Table 8. As the number of iterations increases, the error obtained by KF gradually decreases and is lower than that obtained by the other two methods. However, in the middle stage of the iteration, accurate results may be obtained using the CMOD-based and SSG-based methods. When the error increases, KF processing can still keep the error within a lower range, that is, for the KF-based method, a sufficient number of iterations is necessary to obtain accurate results.

Figure 14 shows the RE of the SIF of cracks with different values obtained by the KF when the loading is completed. The RE does not exceed 1.1%.

5.3. Central Cracks under Nonuniform Loads

In practice, the load borne by a structure may be nonuniform. Therefore, the validity of the calculation method for the SIF based on the KF for specimens under nonuniform loads is verified by experiments. The stress of the plate in this condition is shown in Figure 15. Experimental studies were carried out on the specimens with .

Figure 16 shows the RE during the whole loading process. The RE at the completion of the iteration is expressed in Table 9, and the KF-based method can still obtain a very accurate result even though the error obtained by the SSG and CMOD methods is large. The error does not exceed 1.2%.

6. Prospects

The work of this paper was carried out in view of cracks on a floating offshore platform in the South China Sea. Cracks were observed in the living building of the platform shortly after it was put into operation. As an indicator of crack severity, the SIF can be used in the calculation of fatigue strength and crack propagation of cracked structures; therefore, it is necessary to calculate the SIF of the crack tip of marine structures. Fatigue is one of the most important failure modes for marine structures; therefore, the proposed method will be applied in the evaluation of fatigue life assessment methods in future research.

A more accurate fatigue crack propagation model can be established using the data fusion-based method proposed in this paper during the propagation process. However, the specific parameters need to be determined and verified by subsequent numerical simulations and experiments.

7. Conclusion

Based on the single strain gauge and maximum opening displacement, a novel calculation method of the SIF based on a Kalman filter is proposed in this paper. Numerical simulations and experimental verifications were carried out on specimens with precracks in different configurations. The following conclusions are drawn:(1)It is necessary to calculate the effective radial range. Using the condition of as an example, the effective radial length ranges from 3.81 to 21.4 mm. Radial lengths of 12.58 mm and 23.05 mm were used for calculation, and the results were more accurate when r = 12.58 mm.(2)Simulation and experimental results show that the KF-based method will yield more accurate results as the number of iterations increases. For the simulation, the number of iteration steps is 100, while for the experiment, the number of iteration steps varies due to the different collection times. Sufficient sample size is necessary to obtain accurate SIF values.(3)The method is effective for calculating the SIF of specimens with prefabricated cracks of different lengths, and the RE does not exceed 1.1% when the iteration is completed.(4)The proposed method is also applicable to central cracks under nonuniform loads, and the RE of the results is not greater than 1.2%.

Data Availability

The raw/processed data required to reproduce these findings cannot be shared at this time as the data also forms part of an ongoing study.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

This work has been supported by the National Key R&D Program of China (2019YFB1504303), the National Natural Science Foundation of China (51608094), and the Fundamental Research Funds for the Central Universities (DUT20JC09). These grants are greatly appreciated.

References

X. L. Qing, W. Z. Li, Y. S. Wang, and H. Sun, “Piezoelectric transducer-based structural health monitoring for aircraft applications,” Sensors, vol. 19, no. 3, p. 545, 2019.
View at: Publisher Site | Google Scholar
M. M. Wang, C. Y. Wang, A. Hnydiuk-Stefan, S. Z. Feng, I. Atilla, and Z. Li, “Recent progress on reliability analysis of offshore wind turbine support structures considering digital twin solutions,” Ocean Engineering, vol. 232, Article ID 109168, 2021.
View at: Publisher Site | Google Scholar
G. Liu, Z. R. Li, Z. Y. Li, and Y. Huang, “A multi-axial fatigue-oriented strategy for fatigue damage monitoring and assessment of tubular joints,” Ocean Engineering, vol. 227, Article ID 108876, 2021.
View at: Publisher Site | Google Scholar
X. Zhang and R. Simons, “Experimental investigation on the structure of turbulence in the bottom wave-current boundary layers,” Coastal Engineering, vol. 152, Article ID 103511, 2019.
View at: Publisher Site | Google Scholar
Y. Yao, S. T. E. Tung, and B. Glisic, “Crack detection and characterization techniques-An overview,” Structural Control and Health Monitoring, vol. 21, no. 12, pp. 1387–1413, 2014.
View at: Publisher Site | Google Scholar
P. Dey, S. Talukdar, and D. J. Bordoloi, “Multiple-crack identification in a channel section steel beam using a combined response surface methodology and genetic algorithm,” Structural Control and Health Monitoring, vol. 23, no. 6, pp. 938–959, 2016.
View at: Publisher Site | Google Scholar
J. C. P. Kam and M. Birkinshaw, “Reliability-based fatigue and fracture-mechanics assessment methodology for offshore structural components,” International Journal of Fatigue, vol. 16, no. 3, pp. 183–192, 1994.
View at: Publisher Site | Google Scholar
J. Rajasankar, N. R. Iyer, and T. Rao, “Structural integrity assessment of offshore tubular joints based on reliability analysis,” International Journal of Fatigue, vol. 25, no. 7, pp. 609–619, 2003.
View at: Publisher Site | Google Scholar
S. Yuan, H. Zhang, L. Qiu, and W. Yang, “A fatigue crack growth prediction method based on particle filter,” Acta Aeronautica et Astronautica Sinica, vol. 34, no. 12, pp. 134–141, 2013.
View at: Google Scholar
J. M. Domínguez, A. J. Crespo, M. Hall et al., “SPH simulation of floating structures with moorings,” Coastal Engineering, vol. 153, Article ID 103560, 2019.
View at: Publisher Site | Google Scholar
G. Irwin, “Analysis of stresses and strains near end of a crack traversing a plate,” Journal of Applied Mechanics, vol. 24, 1956.
View at: Publisher Site | Google Scholar
A. Griffith, “The phenomena of rupture and flows in solids,” Philosophical Transactions of the Royal Society, vol. 221, 1921.
View at: Publisher Site | Google Scholar
E. Byskov, “The calculation of stress intensity factors using the finite element method with cracked elements,” International Journal of Fracture Mechanics, vol. 6, no. 2, pp. 159–167, 1970.
View at: Publisher Site | Google Scholar
J. Yang and Y. Lei, “The enriched finite element analysis of stress intensity factor for a crack normal to BI-material interface,” Engineering Mechanics, vol. 33, no. 2, pp. 59–65, 2016.
View at: Google Scholar
K. H. Lee and Z. C. Xuan, “Computing a-posteriori bounds for stress intensity factors in elastic fracture mechanics,” International Journal of Fracture, vol. 126, no. 2, pp. 123–142, 2004.
View at: Publisher Site | Google Scholar
I. S. Raju and J. C. Newman, “Stress-Intensity factors for a wide-range of semi-elliptical surface cracks in finite-thickness plates,” Engineering Fracture Mechanics, vol. 11, no. 4, pp. 817–829, 1979.
View at: Publisher Site | Google Scholar
J. C. Newman and I. S. Raju, “AN empirical stress-intensity factor equation for the surface crack,” Engineering Fracture Mechanics, vol. 15, no. 1-2, pp. 185–192, 1981.
View at: Publisher Site | Google Scholar
I. Newman and I. S. Raju, Stress Intensity Factor Equations for Cracks in Three-Dimensional Finite Bodies, 1983.
N. N. V. Prasad, M. H. Aliabadi, and D. P. Rooke, “The dual boundary-element method for thermoelastic crack problems,” International Journal of Fracture, vol. 66, no. 3, pp. 255–272, 1994.
View at: Publisher Site | Google Scholar
H. Zheng, J. Sladek, V. Sladek, S. K. Wang, and P. Wen, “Hybrid meshless/displacement discontinuity method for FGM Reissner’s plate with cracks,” Applied Mathematical Modelling, vol. 90, pp. 1226–1244, 2021.
View at: Publisher Site | Google Scholar
J. W. Dally and R. J. Sanford, “Strain-gage methods for measuring the opening-mode stress-intensity factor, KI,” Experimental Mechanics, vol. 27, no. 4, pp. 381–388, 1987.
View at: Publisher Site | Google Scholar
J. Wei and J. H. Zhao, “A two-strain-gage technique for determining mode I stress-intensity factor,” Theoretical and Applied Fracture Mechanics, vol. 28, no. 2, pp. 135–140, 1997.
View at: Publisher Site | Google Scholar
H. Sarangi, K. Murthy, and D. Chakraborty, “Optimum strain gage location for evaluating stress intensity factors in single and double ended cracked configurations,” Engineering Fracture Mechanics, vol. 77, no. 16, pp. 3190–3203, 2010.
View at: Publisher Site | Google Scholar
H. Sarangi, K. S. R. K. Murthy, and D. Chakraborty, “Experimental verification of optimal strain gage locations for the accurate determination of mode I stress intensity factors,” Engineering Fracture Mechanics, vol. 110, pp. 189–200, 2013.
View at: Publisher Site | Google Scholar
V. N. Maksimenko and A. V. Tyagnii, “Calculation of fracture parameters based on crack opening displacements,” in Proceedings of the 8th Korea/Russia International Symposium on Science and Technology, pp. 39–42, Tomsk, Russia, June 2004.
View at: Publisher Site | Google Scholar
J. Chen, Y. Huang, and G. Liu, “Research on calculation of stress intensity factors in a cracked plate,” Journal of Harbin Engineering University, vol. 32, no. 1, pp. 129–134, 2011.
View at: Google Scholar
J. J. Chen, Y. Huang, and Y. G. Li, “Asme, research on calculation of dsif based on maximum crack opening displacement,” in Proceedings of the 36th ASME International Conference on Ocean, Offshore and Arctic Engineering, Trondheim, Norway, June 2017.
View at: Publisher Site | Google Scholar
J. Chen, X. Fan, and Y. Huang, “A practical method of stress intensity factors for two collinear cracks,” J. Huazhong Univ. Sci. Technol., Nat. Sci. Ed. (China), vol. 45, no. 4, pp. 61–67, 2017.
View at: Google Scholar
L. Saucedo-Mora, M. C. Andrade Perdrix, C. López Hombrados, J. Barroso Sanchez, and A. Zamora Bragado, “Contactless safety evaluation of damaged structures through energetic criteria,” Structural Control and Health Monitoring, vol. 25, no. 1, Article ID e2060, 2018.
View at: Publisher Site | Google Scholar
C. Chen and X. Qian, “Quantifying through-thickness J for plane strain specimens using digital image correlation considering constraint effects,” Engineering Fracture Mechanics, vol. 267, Article ID 108430, 2022.
View at: Publisher Site | Google Scholar
C. Chen and X. Qian, “A non-contact approach to measure J-R curve for side-grooved specimens using digital image correlation,” Theoretical and Applied Fracture Mechanics, vol. 121, Article ID 103549, 2022.
View at: Publisher Site | Google Scholar
A. Smyth and M. L. Wu, “Multi-rate Kalman filtering for the data fusion of displacement and acceleration response measurements in dynamic system monitoring,” Mechanical Systems and Signal Processing, vol. 21, no. 2, pp. 706–723, 2007.
View at: Publisher Site | Google Scholar
C. D. Zhang and Y. L. Xu, “Structural damage identification via multi-type sensors and response reconstruction,” Structural Health Monitoring, vol. 15, no. 6, pp. 715–729, 2016.
View at: Publisher Site | Google Scholar
J. W. Park, D. S. Moon, H. Yoon, F. Gomez, B. F. Spencer Jr, and J. R. Kim, “Visual-inertial displacement sensing using data fusion of vision-based displacement with acceleration,” Structural Control and Health Monitoring, vol. 25, no. 3, Article ID e2122, 2018.
View at: Publisher Site | Google Scholar
G. Ravizza, R. Ferrari, E. Rizzi, and E. N. Chatzi, “Effective heterogeneous data fusion procedure via kalman filtering,” Smart Structures and Systems, vol. 22, no. 5, pp. 631–641, 2018.
View at: Google Scholar
M. F. Bado, G. Kaklauskas, and J. R. Casas, “Performance of distributed optical fiber sensors (DOFS) and digital image correlation (DIC) in the monitoring of RC structures,” IOP Conference Series: Materials Science and Engineering, vol. 615, Article ID 012101, 2019.
View at: Publisher Site | Google Scholar
H. P. Zhu, K. Gao, Y. Xia et al., “Multi-rate data fusion for dynamic displacement measurement of beam-like supertall structures using acceleration and strain sensors,” Structural Health Monitoring, vol. 19, no. 2, pp. 520–536, 2020.
View at: Publisher Site | Google Scholar
M. F. Bado, J. R. Casas, A. Dey, and C. G. Berrocal, “Distributed optical fiber sensing bonding techniques performance for embedment inside reinforced concrete structures,” Sensors, vol. 20, no. 20, p. 5788, 2020.
View at: Publisher Site | Google Scholar
S. S. Jin, S. T. Kim, and Y. H. Park, “Combining point and distributed strain sensor for complementary data-fusion: a multi-fidelity approach,” Mechanical Systems and Signal Processing, vol. 157, Article ID 107725, 2021.
View at: Publisher Site | Google Scholar
C. Qi, Y. Weixi, L. Jun, G. Heming, and M. Yao, “A research on fatigue crack growth monitoring based on multi-sensor and data fusion,” Structural Health Monitoring, vol. 20, no. 3, pp. 848–860, 2021.
View at: Publisher Site | Google Scholar
R. T. Wu and M. R. Jahanshahi, “Data fusion approaches for structural health monitoring and system identification: past, present, and future,” Structural Health Monitoring, vol. 19, no. 2, pp. 552–586, 2020.
View at: Publisher Site | Google Scholar
K. Erazo, D. Sen, S. Nagarajaiah, and L. M. Sun, “Vibration-based structural health monitoring under changing environmental conditions using Kalman filtering,” Mechanical Systems and Signal Processing, vol. 117, pp. 1–15, 2019.
View at: Publisher Site | Google Scholar
A. J. Rosakis and K. Ravichandar, “On crack-tip stress state: an experimental evaluation of three-dimensional effects,” International Journal of Solids and Structures, vol. 22, no. 2, pp. 121–134, 1986.
View at: Publisher Site | Google Scholar
H. Tada, P. C. Paris, and G. R. Irwin, The Stress Analysis of Cracks Handbook, The Stress Analysis Of Cracks Handbook, The American Society of Mechanical Engineers, New York, NY, USA, 1985.

Copyright

Copyright © 2023 Tong-Tong Dai et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PDF Download Citation

Download other formats

Order printed copies