A phase field and deep-learning based approach for accurate prediction of structural residual useful life

Published on Sep 1, 2021in Computer Methods in Applied Mechanics and Engineering6.756
· DOI :10.1016/J.CMA.2021.113885
S.Z. Feng3
Estimated H-index: 3
(HEBUT: Hebei University of Technology),
Y. Xu1
Estimated H-index: 1
(Ocean University of China)
+ 2 AuthorsAtilla Incecik29
Estimated H-index: 29
(University of Strathclyde)
Source
Abstract
Abstract In this work, we proposed a novel approach for the prediction of residual useful life (RUL) of structures through appropriately combining the phase field method and deep-learning. In this new approach, the phase field method is firstly utilized to obtain the structural responses of crack growth, which are further preserved as images. Then, the convolutional neural network (CNN) is constructed to establish a predictive model . The proposed approach is a hybrid model of both physical and data-driven techniques, which can build a bridge between traditional computational fracture mechanics and deep learning algorithms. Several numerical cases are studied to evaluate the prediction performance of the proposed approach. The analysis results demonstrate that the present approach is able to predict the RUL of the structures with high level of accuracy.
References60
Newest
#1Adam Glowacz (AGH University of Science and Technology)H-Index: 27
Abstract Fault diagnosis enables to make savings related to maintenance. The presented work describes fault diagnosis method based on analysis of thermal images. An original method for feature extraction of thermal images BCAoID (Binarized Common Areas of Image Differences) is proposed. Thermal images of three electric impact drills (EID) were used for an analysis: healthy EID, EID with faulty fan (10 broken fan blades), EID with damaged gear train. Features of thermal images were extracted usin...
Source
This work presents the development of novel convolutional neural network (NCNN) for effective identification of bearing defects from small samples. For effective feature learning from small training data, cost function of convolution neural network (CNN) is modified by adding additional sparsity cost in the existing cost function. A novel trigonometric cross-entropy function is developed to compute the sparsity cost. The proposed cost function introduces sparsity by avoiding unnecessary activati...
Source
#1Duc-Kien Thai (Sejong University)H-Index: 13
#2Tran Minh Tu (National University of Civil Engineering)H-Index: 9
Last. T.T. Bui (University of Lyon)H-Index: 10
view all 4 authors...
This paper proposed a new approach in predicting the local damage of reinforced concrete (RC) panels under impact loading using gradient boosting machine learning (GBML), one of the most powerful techniques in machine learning. A number of experimental data on the impact test of RC panels were collected for training and testing of the proposed model. With the lack of test data due to the high cost and complexity of the structural behavior of the panel under impact loading, it was a challenge to ...
Source
#1S.Z. Feng (HEBUT: Hebei University of Technology)H-Index: 3
#2Xu Han (HEBUT: Hebei University of Technology)H-Index: 58
Last. Zhen Li (Ocean University of China)H-Index: 1
view all 5 authors...
Abstract Fatigue crack growth analysis using extended finite element method (XFEM) is an efficient way to predict the residual life of structures; however, when the structure parameters vary stochastically, it will be very hard to make accurate predictions. To bridge this research gap, this work proposed a data-driven learning algorithm to improve the prediction capacity of fatigue life by considering stochastic parameters of structures. In this new algorithm, the XFEM was firstly employed to ge...
Source
#1Zuozhou Pan (Yanshan University)H-Index: 4
#2Zong Meng (Yanshan University)H-Index: 5
Last. Ying Shi (Yanshan University)H-Index: 2
view all 5 authors...
Abstract Rolling-element bearing is one of the main parts of rotating equipment. In order to avoid the mechanical equipment damage caused by the sudden failure of rolling-element bearings, it is necessary to monitor the condition of bearing and predict its life. Therefore, a two-stage prediction method based on extreme learning machine is proposed to predict the remaining useful life of rolling-element bearings quickly and accurately. This method uses the relative root mean square value (RRMS) t...
Source
#1Anil Kumar (Wenzhou University)H-Index: 17
#2C.P. Gandhi (Rayat Bahra University)H-Index: 7
Last. Jiawei Xiang (Wenzhou University)H-Index: 29
view all 5 authors...
Abstract An improved CNN is proposed for the diagnosis of defects in components of a centrifugal pump. The improvement is attained by modifying the cost function of CNN. For defect identification, first, grey scale acoustic images are obtained by processing acoustic signals using analytical wavelet transform (AWT). Second, a new entropy based divergence function, a type of regularization function is introduced in the cost function of CNN which avoids redundant activation of hidden layer in CNN, ...
Source
#2Wei Zhou (WHU: Wuhan University)H-Index: 23
Abstract Phase-field models have become popular to simulate cohesive failure problems because of their capability of predicting crack initiation and propagation without additional criteria. In this paper, new phase-field damage model coupled with general softening law for cohesive fracture is proposed based on the unified phase-field theory. The commonly used quadratic geometric function in the classical phase-field model is implemented in the proposed model. The modified degradation function re...
Source
Abstract Different fracture patterns can be observed because of different material properties, even the geometry and loading are the same. However, most of the known phase-field fracture models have only considered the tensile failure and may not be directly applicable to the shear fracture. In this paper, a phase-field model for mixed-mode fracture is proposed based on a unified tensile fracture criterion. The proposed model is developed from the unified phase-field theory and the original unif...
Source
#1Fan Fei (HKU: University of Hong Kong)H-Index: 3
#2Jinhyun Choo (HKU: University of Hong Kong)H-Index: 17
Geologic shear fractures such as faults and slip surfaces involve marked friction along the discontinuities as they are subjected to significant confining pressures. This friction plays a critical role in the growth of these shear fractures, as revealed by the fracture mechanics theory of Palmer and Rice decades ago. In this paper, we develop a novel phase-field model of shear fracture in pressure-sensitive geomaterials, honoring the role of friction in the fracture propagation mechanism. Buildi...
Source
#1Prasun Chandra Tripathi (IITs: Indian Institutes of Technology)H-Index: 3
#2Soumen Bag (IITs: Indian Institutes of Technology)H-Index: 10
Abstract Magnetic Resonance Images (MRI) are often contaminated by rician noise at the acquisition time. This type of noise typically deteriorates the performance of disease diagnosis by a human observer or an automated system. Thus, it is necessary to remove the rician noise from MRI scans as a preprocessing step. In this letter, we propose a novel Convolutional Neural Network (CNN), viz. CNN-DMRI, for denoising of MRI scans. The network uses a set of convolutions to separate the image features...
Source
Cited By1
Newest
This website uses cookies.
We use cookies to improve your online experience. By continuing to use our website we assume you agree to the placement of these cookies.
To learn more, you can find in our Privacy Policy.