Application of deep learning method in web crippling strength prediction of cold-formed stainless steel channel sections under end-two-flange loading

Abstract This paper proposes a deep-learning framework, specifically, a Deep Belief Network (DBN), for studying the web crippling performance of cold-formed stainless steel channel sections (lipped and unlipped as well as fastened and unfastened) with centered and offset web holes under the end-two-flange loading condition. G430 ferritic, S32205 duplex and 304 austenitic stainless steel grades are considered. A total of 17,281 data points for training the DBN are generated from an elasto plastic finite element model, validated from 69 experimental results reported in the literature. When a comparison was made against a further 53 experimental results reported in the literature, the DBN predictions were found to be conservative by around 10%. When compared with Backpropagation Neural Network (a typical shallow artificial neural network) and linear regression model based on PaddlePaddle, it was found that the proposed DBN outperformed these two methods, using the same big training data generated in this study. Using the DBN predictions, a parametric study is then conducted to investigate the effect of web holes, from which unified strength reduction factor equations are proposed. Finally, a reliability analysis is conducted, which shown that the proposed equations can predict the web crippling strength of cold-formed stainless steel channel sections under the end-two-flange loading condition.

[1]  Zhaohua Hu.  Research and Implementation of Railway Technical Specification Question Answering System Based on Deep Learning , 2020, 2020 IEEE 5th Information Technology and Mechatronics Engineering Conference (ITOEC).

[2]  Krishanu Roy,et al.  Cold-formed steel channel sections under end-two-flange loading condition: Design for edge-stiffened holes, unstiffened holes and plain webs , 2020, Thin-Walled Structures.

[3]  James B. P. Lim,et al.  Web crippling design of cold-formed ferritic stainless steel unlipped channels with fastened flanges under end-two-flange loading condition , 2019, Journal of Constructional Steel Research.

[4]  Mahen Mahendran,et al.  Web Crippling Tests of Cold-formed Steel Channels under Two Flange Load Cases , 2015 .

[5]  George Papazafeiropoulos,et al.  Prediction of buckling coefficient of stiffened plate girders using deep learning algorithm , 2020 .

[6]  Asraf Uzzaman,et al.  Web crippling behaviour of cold-formed steel channel sections with web holes subjected to interior-one-flange loading condition – Part II: parametric study and proposed design equations , 2017 .

[7]  Nuno Silvestre,et al.  A neural network based closed-form solution for the distortional buckling of elliptical tubes , 2011 .

[8]  Asraf Uzzaman,et al.  Effect of offset web holes on web crippling strength of cold-formed steel channel sections under end-two-flange loading condition , 2013 .

[9]  Ben Young,et al.  Web Crippling of Cold-Formed Stainless Steel Tubular Sections , 2008 .

[10]  Zia ul Rehman Tahir,et al.  Artificial neural network prediction of buckling load of thin cylindrical shells under axial compression , 2017 .

[11]  Reza Attarnejad,et al.  Assessment of load carrying capacity of castellated steel beams by neural networks , 2011 .

[12]  Mahen Mahendran,et al.  Experimental Studies of Lipped Channel Beams Subject to Web Crippling under Two-Flange Load Cases , 2016 .

[13]  Ben Young,et al.  Cold-formed stainless steel sections subjected to web crippling , 2006 .

[14]  Asifullah Khan,et al.  Wind power prediction using deep neural network based meta regression and transfer learning , 2017, Appl. Soft Comput..

[15]  Mahen Mahendran,et al.  Experimental studies on web crippling behaviour of hollow flange channel beams under two flange load cases , 2014 .

[16]  Haibo Xu,et al.  Deep belief network-based support vector regression method for traffic flow forecasting , 2019, Neural Computing and Applications.

[17]  Victor L. Brailovsky,et al.  On global, local, mixed and neighborhood kernels for support vector machines , 1999, Pattern Recognit. Lett..

[18]  Gao Fan,et al.  Vibration signal denoising for structural health monitoring by residual convolutional neural networks , 2020 .

[19]  Amir M. Yousefi,et al.  Cold-formed ferritic stainless steel unlipped channels with web openings subjected to web crippling under interior-two-flange loading condition – Part I: Tests and finite element model validation , 2017 .

[20]  Nancy Baddoo,et al.  Designing Structural Stainless Steel Members To Eurocode 3 , 2009 .

[21]  Dinh Van Thuat,et al.  A deep learning-based procedure for estimation of ultimate load carrying of steel trusses using advanced analysis , 2019, Journal of Science and Technology in Civil Engineering (STCE) - NUCE.

[22]  Amir M. Yousefi,et al.  Web bearing capacity of unlipped cold-formed ferritic stainless steel channels with perforated web subject to end-two-flange (ETF) loading , 2017 .

[23]  Heng Liu,et al.  Image-driven structural steel damage condition assessment method using deep learning algorithm , 2019, Measurement.

[24]  Amir M. Yousefi,et al.  Web crippling strength of cold-formed stainless-steel lipped channels with web perforations under end-two-flange loading , 2017 .

[25]  Sajjad Tohidi,et al.  Load-carrying capacity of locally corroded steel plate girder ends using artificial neural network , 2016 .

[26]  Esther Real,et al.  Study of web crippling in ferritic stainless steel cold formed sections , 2013 .

[27]  Geoffrey E. Hinton,et al.  Reducing the Dimensionality of Data with Neural Networks , 2006, Science.

[28]  G. J. van den Berg,et al.  Web crippling of stainless steel cold-formed beams , 1995 .

[29]  Hong Hao,et al.  Structural damage identification by sparse deep belief network using uncertain and limited data , 2020, Structural Control and Health Monitoring.

[30]  F Zhou,et al.  Cold-Formed High-Strength Stainless Steel Tubular Sections Subjected to Web Crippling , 2007 .

[31]  Xu Li,et al.  Improved Touch-screen Inputting Using Sequence-level Prediction Generation , 2020, WWW.

[32]  R. Rokade,et al.  Determination of Modelling Error Statistics for Cold-Formed Steel Columns , 2020 .

[33]  Ponnuthurai Nagaratnam Suganthan,et al.  Empirical Mode Decomposition based ensemble deep learning for load demand time series forecasting , 2017, Appl. Soft Comput..

[34]  James B.P. Lim,et al.  Deep learning-based axial capacity prediction for cold-formed steel channel sections using Deep Belief Network , 2021 .

[35]  Dinar Camotim,et al.  Computational modelling of flange crushing in cold-formed steel sections , 2014 .

[36]  Asraf Uzzaman,et al.  Effects of edge-stiffened circular holes on the web crippling strength of cold-formed steel channel sections under one-flange loading conditions , 2017 .

[37]  Yu Chen,et al.  Experimental and finite element analysis research on cold-formed steel lipped channel beams under web crippling , 2015 .

[38]  James B. P. Lim,et al.  Web crippling strength of perforated cold-formed ferritic stainless steel unlipped channels with restrained flanges under one-flange loadings , 2019, Thin-Walled Structures.

[39]  Krishanu Roy,et al.  Web crippling behaviour of cold-formed steel channel sections with edge-stiffened and unstiffened circular holes under interior-two-flange loading condition , 2020 .

[40]  Asraf Uzzaman,et al.  Cold-formed steel sections with web openings subjected to web crippling under two-flange loading conditions — Part II: Parametric study and proposed design equations , 2012 .

[41]  Asraf Uzzaman,et al.  Effect of web holes on web crippling strength of cold-formed steel channel sections under end-one-flange loading condition - Part II: Parametric study and proposed design equations , 2016 .

[42]  Asraf Uzzaman,et al.  Effect of web holes on web crippling strength of cold-formed steel channel sections under end-one-flange loading condition – Part I: Tests and finite element analysis , 2016 .

[43]  James B.P. Lim,et al.  Web crippling behaviour of cold-formed steel channel sections with offset web holes subjected to interior-two-flange loading , 2012 .

[44]  Konstantinos Daniel Tsavdaridis,et al.  Neural Network-Based Formula for the Buckling Load Prediction of I-Section Cellular Steel Beams , 2018, Comput..

[45]  Jianqiang Yi,et al.  Deep Belief Network Based Hybrid Model for Building Energy Consumption Prediction , 2018 .

[46]  Young-Jin Cha,et al.  Subsurface damage detection of a steel bridge using deep learning and uncooled micro-bolometer , 2019, Construction and Building Materials.

[47]  Asraf Uzzaman,et al.  Cold-Formed Steel Sections with Web Openings Subjected to Web Crippling under Two-Flange Loading Conditions — Part I : Tests and Finite Element Analysis , 2012 .

[48]  Xuan Jing,et al.  An efficient three-step search algorithm for block motion estimation , 2004, IEEE Transactions on Multimedia.

[49]  Asraf Uzzaman,et al.  Web crippling behaviour of cold-formed steel channel sections with web holes subjected to Interior-one-flange loading condition-part I: experimental and numerical investigation , 2017 .

[50]  Wenhao Huang,et al.  Deep Architecture for Traffic Flow Prediction: Deep Belief Networks With Multitask Learning , 2014, IEEE Transactions on Intelligent Transportation Systems.

[51]  James B. P. Lim,et al.  Web Crippling Behavior of Unlipped Cold-Formed Ferritic Stainless Steel Channels Subject to One-Flange Loadings , 2018, Journal of Structural Engineering.

[52]  James B.P. Lim,et al.  Deep learning-based procedure for structural design of cold-formed steel channel sections with edge-stiffened and un-stiffened holes under axial compression , 2021 .

[53]  Yanjun Ma,et al.  PaddlePaddle: An Open-Source Deep Learning Platform from Industrial Practice , 2019 .

[54]  Yitao Liu,et al.  Deep learning based ensemble approach for probabilistic wind power forecasting , 2017 .