M-LFM: a multi-level fusion modeling method for shape−performance integrated digital twin of complex structure

Abstract

As a virtual representation of a specific physical asset, the digital twin has great potential for realizing the life cycle maintenance management of a dynamic system. Nevertheless, the dynamic stress concentration is generated since the state of the dynamic system changes over time. This generation of dynamic stress concentration has hindered the exploitation of the digital twin to reflect the dynamic behaviors of systems in practical engineering applications. In this context, this paper is interested in achieving real-time performance prediction of dynamic systems by developing a new digital twin framework that includes simulation data, measuring data, multi-level fusion modeling (M-LFM), visualization techniques, and fatigue analysis. To leverage its capacity, the M-LFM method combines the advantages of different surrogate models and integrates simulation and measured data, which can improve the prediction accuracy of dynamic stress concentration. A telescopic boom crane is used as an example to verify the proposed framework for stress prediction and fatigue analysis of the complex dynamic system. The results show that the M-LFM method has better performance in the computational efficiency and calculation accuracy of the stress prediction compared with the polynomial response surface method and the kriging method. In other words, the proposed framework can leverage the advantages of digital twins in a dynamic system: damage monitoring, safety assessment, and other aspects and then promote the development of digital twins in industrial fields.

Graphical abstract

Keywords

shape−performance integrated digital twin (SPI-DT) / multi-level fusion modeling (M-LFM) / surrogate model / telescopic boom crane / data fusion

References

[1] Bloem J , van Doorn M , Duivestein S , Excoffier D , Maas E , van Ommeren E . The fourth industrial revolution. Sogeti VINT2014, 2014, 8: 11–15

[2] Xu M , David J M , Kim S H . The Fourth Industrial Revolution: opportunities and challenges. International Journal of Financial Research, 2018, 9(2): 90–95

[3] Cekus D , Kwiatoń P , Geisler T . The dynamic analysis of load motion during the interaction of wind pressure. Meccanica, 2021, 56(4): 785–796

[4] Trąbka A . Dynamics of telescopic cranes with flexible structural components. International Journal of Mechanical Sciences, 2014, 88: 162–174

[5] Guivarch D , Mermoz E , Marino Y , Sartor M . Creation of helicopter dynamic systems digital twin using multibody simulations. CIRP Annals, 2019, 68(1): 133–136

[6] Grieves M. Digital Twin: Manufacturing Excellence Through Virtual Factory Replication. White Paper, 2014

[7] Tuegel E J , Ingraffea A R , Eason T G , Spottswood S M . Reengineering aircraft structural life prediction using a digital twin. International Journal of Aerospace Engineering, 2011, 2011: 154798

[8] Tao F , Cheng J F , Qi Q L , Zhang M , Zhang H , Sui F Y . Digital twin-driven product design, manufacturing and service with Big Data. The International Journal of Advanced Manufacturing Technology, 2018, 94(9): 3563–3576

[9] Jiang H F , Qin S F , Fu J L , Zhang J , Ding G F . How to model and implement connections between physical and virtual models for digital twin application. Journal of Manufacturing Systems, 2021, 58: 36–51

[10] Ye Y M , Yang Q , Yang F , Huo Y Y , Meng S H . Digital twin for the structural health management of reusable spacecraft: a case study. Engineering Fracture Mechanics, 2020, 234: 107076

[11] He X W , Qiu Y M , Lai X N , Li Z H , Shu L M , Sun W , Song X G . Towards a shape−performance integrated digital twin for lumbar spine analysis. Digital Twin, 2021, 1(8): 8

[12] Ritto T G , Rochinha F A . Digital twin, physics-based model, and machine learning applied to damage detection in structures. Mechanical Systems and Signal Processing, 2021, 155: 107614

[13] Lai X N , Wang S , Guo Z G , Zhang C , Sun W , Song X G . Designing a shape–performance integrated digital twin based on multiple models and dynamic data: a boom crane example. Journal of Mechanical Design, 2021, 143(7): 071703

[14] Kapteyn M G , Knezevic D J , Huynh D B P , Tran M , Willcox K E . Data-driven physics-based digital twins via a library of component-based reduced-order models. International Journal for Numerical Methods in Engineering, 2022, 123(13): 2986–3003

[15] Luo W C , Hu T L , Ye Y X , Zhang C R , Wei Y L . A hybrid predictive maintenance approach for CNC machine tool driven by digital twin. Robotics and Computer-Integrated Manufacturing, 2020, 65: 101974

[16] Wang S , Lai X N , He X W , Qiu Y M , Song X G . Building a trustworthy product-level shape−performance integrated digital twin with multifidelity surrogate model. Journal of Mechanical Design, 2022, 144(3): 031703

[17] Kontaxoglou A, Tsutsumi S, Khan S, Nakasuka S. Towards a digital twin enabled multifidelity framework for small satellites. In: Proceedings of the 6th European Conference of the Prognostics and Health Management Society. 2021, 211–220

[18] Chetan M , Yao S L , Griffith D T . Multi-fidelity digital twin structural model for a sub-scale downwind wind turbine rotor blade. Wind Energy, 2021, 24(12): 1368–1387

[19] Al Zamzami I , Susmel L . On the accuracy of nominal, structural, and local stress based approaches in designing aluminium welded joints against fatigue. International Journal of Fatigue, 2017, 101: 137–158

[20] Nziu P K , Masu L M . Formulae for predicting stress concentration factors in flat plates and cylindrical pressure vessels with holes: a review. International Journal of Mechanical and Production Engineering Research and Development, 2019, 9(5): 753–770

[21] Song X G , Lv L Y , Li J L , Sun W , Zhang J . An advanced and robust ensemble surrogate model: extended adaptive hybrid functions. Journal of Mechanical Design, 2018, 140(4): 041402

[22] Lai X N , He X W , Wang S , Wang X B , Sun W , Song X G . Building a lightweight digital twin of a crane boom for structural safety monitoring based on a multi-fidelity surrogate model. Journal of Mechanical Design, 2022, 144(6): 064502

[23] Sun G Y , Song X G , Baek S , Li Q . Robust optimization of foam-filled thin-walled structure based on sequential kriging metamodel. Structural and Multidisciplinary Optimization, 2014, 49(6): 897–913

[24] Chakraborty S , Adhikari S , Ganguli R . The role of surrogate models in the development of digital twins of dynamic systems. Applied Mathematical Modelling, 2021, 90: 662–681

[25] Khuri A I , Mukhopadhyay S . Response surface methodology. Wiley Interdisciplinary Reviews: Computational Statistics, 2010, 2(2): 128–149

[26] Cressie N . The origins of kriging. Mathematical Geology, 1990, 22(3): 239–252

[27] Fuhg J N , Fau A , Nackenhorst U . State-of-the-art and comparative review of adaptive sampling methods for kriging. Archives of Computational Methods in Engineering, 2021, 28(4): 2689–2747

[28] Grieves M , Vickers J . Digital twin: mitigating unpredictable, undesirable emergent behavior in complex systems. In: Kahlen J, Flumerfelt S, Alves A, eds. Transdisciplinary Perspectives on Complex Systems. Cham: Springer, 2017, 85–113

[29] Tao F , Zhang H , Liu A , Nee A Y C . Digital twin in industry: state-of-the-art. IEEE Transactions on Industrial Informatics, 2019, 15(4): 2405–2415

[30] Bhosekar A , Ierapetritou M . Advances in surrogate based modeling, feasibility analysis, and optimization: a review. Computers & Chemical Engineering, 2018, 108: 250–267

[31] Savković M , Gašić M , Pavlović G , Bulatović R , Zdravković N . Stress analysis in contact zone between the segments of telescopic booms of hydraulic truck cranes. Thin-Walled Structures, 2014, 85: 332–340

[32] Yao J , Xing F , Fu Y Q , Qiu X M , Zhou Z P , Hou J H . Failure analysis of torsional buckling of all-terrain crane telescopic boom section. Engineering Failure Analysis, 2017, 73: 72–84

[33] Liao J B , Tang G W , Meng L B , Liu H G , Zhang Y J . Finite element model updating based on field quasi-static generalized influence line and its bridge engineering application. Procedia Engineering, 2012, 31: 348–353

[34] Liu P , Gong J H , Yu M . Visualizing and analyzing dynamic meteorological data with virtual globes: a case study of tropical cyclones. Environmental Modelling & Software, 2015, 64: 80–93

[35] Wang Q Y , Jiao W H , Zhang Y M . Deep learning-empowered digital twin for visualized weld joint growth monitoring and penetration control. Journal of Manufacturing Systems, 2020, 57: 429–439

[36] de Jesus A M P , Matos R , Fontoura B F C , Rebelo C , Simões da Silva L , Veljkovic M . A comparison of the fatigue behavior between S355 and S690 steel grades. Journal of Constructional Steel Research, 2012, 79: 140–150

[37] Dong Q , He B , Qi Q S , Xu G N . Real-time prediction method of fatigue life of bridge crane structure based on digital twin. Fatigue & Fracture of Engineering Materials & Structures, 2021, 44(9): 2280–2306

[38] Liu X T , Zhang M H , Wang H J , Luo J , Tong J C , Wang X L . Fatigue life analysis of automotive key parts based on improved peak-over-threshold method. Fatigue & Fracture of Engineering Materials & Structures, 2020, 43(8): 1824–1836

[39] Hashin Z , Rotem A . A cumulative damage theory of fatigue failure. Materials Science and Engineering, 1978, 34(2): 147–160

[40] Jin R C , Chen W , Simpson T W . Comparative studies of metamodelling techniques under multiple modelling criteria. Structural and Multidisciplinary Optimization, 2001, 23(1): 1–13

[41] Zhang J , Chowdhury S , Messac A . An adaptive hybrid surrogate model. Structural and Multidisciplinary Optimization, 2012, 46(2): 223–238

[42] Karve P M , Guo Y L , Kapusuzoglu B , Mahadevan S , Haile M A . Digital twin approach for damage-tolerant mission planning under uncertainty. Engineering Fracture Mechanics, 2020, 225: 106766

[43] Wojciechowski S , Maruda R W , Krolczyk G M , Niesłony P . Application of signal to noise ratio and grey relational analysis to minimize forces and vibrations during precise ball end milling. Precision Engineering, 2018, 51: 582–596

Nomenclature

Abbreviations
3D	Three-dimensional
DT	Digital twin
FEM	Finite element method
KRG	Kriging
M-LFM	Multi-level fusion modeling
NMAE	Normalized maximum absolute error
NRMSE	Normalized root mean square error
PRS	Polynomial response surface
RUL	Remaining useful life
SNR	Signal-to-noise ratios
SPI-DT	Shape–performance integrated digital twin
Variables
a	Unknown coefficient used to change the result size of the predicted model
A	Amplitude of the signal
Anoise	Amplitude of the noise
b	Scale factor
Df	Remaining useful life
e	Error between the training value and the predicted value at training point
E	Elastic modulus,
fi(∙) (i = 1,2,…,m)	Unknown approximate function
Fi (i = 1,2,…,5)	Thrust produced by the hydraulic cylinder and the pulling force produced by the transmission chain
Ffi (i = 2,3,4)	Fraction between segments
F	Force vector
g	gth cycle
H	Vector whose elements are ones
k	Number of measured data
Kt	Elastic stress concentration factor
L	Cost function
m	Number of input variables
Mg	Total cycle number
M, C, K	Mass matrix, damping matrix, and stiffness matrix, respectively
ng, Ng	Cycle number in practical and cycles to failure, respectively
Nf	Cycle number to failure
Nt	Test point number
P	Signal power
Pnoise	Noise power
\mathit{Q},\dot{\mathit{Q}},\ddot{\mathit{Q}}	Displacement, velocity, and acceleration vector, respectively
r	Related vector between the unknown predicted point and the trained point
R	Relationship function
R2	Coefficient of determination
R	Relationship matrix whose elements are the value of relationship function
t	Time
Ttest	Measured data
ui, vi, wi	Space coordinates in the ith point
uC, vC, wC	Space coordinates in the compensation point
xi (i = 1,2,…,m)	ith input variable
X	Input vector
\overline{y}	Mean value of the true output at all test points
\hat{y}	Predicted value of output variable
y_i^{\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{i}\mathrm{n}}	Response value in the ith training point
{\hat{\mathit{y}}}^{\mathrm{P}\mathrm{R}\mathrm{S}}	Predicted data of PRS model
{\hat{y}}_C^{\mathrm{P}\mathrm{R}\mathrm{S}},{\hat{y}}_C^{\mathrm{K}\mathrm{R}\mathrm{G}}	Predicted values of the PRS model and KRG model at the compensation point, respectively
{\mathit{y}}_{\mathrm{s}}	True response value at samples.
\mathit{Y}	Output vector
{\mathit{Y}}_{\mathrm{P}\mathrm{R}\mathrm{S}},{\mathit{Y}}_{\mathrm{K}\mathrm{R}\mathrm{G}}	Predicted vectors of the PRS model and the KRG model, respectively
\mathit{\text{z}}(\cdot )	Error function of KRG model
C{f}_i (i = 1,2,…,n)	Compensation factor of the influence domain
C{f}_{min}, C{f}_{max}	Minimum and maximum values of the compensation factor set, respectively
C{f}_i^{\mathrm{n}\mathrm{o}\mathrm{r}}	Normalized compensation factor
\mathit{C}{\mathit{f}}_{\mathrm{n}\mathrm{o}\mathrm{r}}	Compensation factor set
{\alpha }_i (i = 1,2,…,m)	Unknown weight coefficient
{\alpha }_i^{\delta },{\alpha }_i^u,{\alpha }_i^a	Unknown weight coefficient of different parameters model
\mathit{\alpha }	Vector of unknown coefficients
\hat{\beta }	Maximum likelihood estimation value of hyper-parameter
{\beta }_i	Unknown weighted coefficients of the KRG model
\delta	Parameter of the influence domain
\hat{\delta },\hat{\mu },\hat{a}	Predicted values of the model parameters under different sensor states
{\delta }_{\mathrm{l}},{\delta }_{\mathrm{u}}	Upper and lower bounds of the parameter \delta, respectively
δopt, {\mathit{\mu }}_{\mathrm{o}\mathrm{p}\mathrm{t}},{\mathit{a}}_{\mathrm{o}\mathrm{p}\mathrm{t}}	Optimal parameters of \delta, \mu, and a, respectively
\mathit{\epsilon }	Vector of error function
\mu	Deviation coefficient of the compensation point
{\mu }_{\mathrm{l}},{\mu }_{\mathrm{u}}	Upper and lower bounds of the parameter \mu, respectively
{\rho }_i (i = 1,2,…,n)	Euclidean distance between the compensation point and the surrounding points
{\rho }_{min}, {\rho }_{max}	Minimum and maximum values of the Euclidean distance set, respectively
{\rho }_i^{\mathrm{n}\mathrm{o}\mathrm{r}}	Normalized Euclidean distance
σ2	Variance of error function
σa	Stress amplitude
σm	Mean stress
σy	Tensile yield stress
σe(Soder)	Modified mean stress
σ	Performance vector
σs, σh	Predicted stresses in the extension and retraction processes, respectively
∆σnom	Nominal stress range
ψi (x)	Function correlated with the physical properties of problem itself
{\psi }_i^{\delta }(x), {\psi }_i^u(x), {\psi }_i^a(x)	Functions correlated with the physical properties of different PRS model

Acknowledgements

This work was supported by the National Key R&D Program of China (Grant No. 2018YFB1700704) and the National Natural Science Foundation of China (Grant No. 52075068).

Electronic Supplementary Materials

The supplementary materials can be found in the online version of this article at https://doi.org/10.1007/s11465-022-0708-0 and are accessible to authorized users.

RIGHTS & PERMISSIONS

2022 Higher Education Press