LSTM-RNN-based defect classification in honeycomb structures using infrared thermography

Abstract Honeycomb-structured materials are widely used in commercial and military aircraft. Manufacturing defects and damage during operation have become primary safety threats. This has increased the demand for non-destructive testing (NDT) for damage and flaws during aircraft operation and maintenance. Characterizing, or classifying defects, in addition to detecting them, is important. Classifying the liquids trapped in aircraft honeycomb cells is an example. A small amount of ingressed water is often tolerable, whereas a small amount of hydraulic oil may be an early warning of hydraulic system malfunction. This paper proposes an infrared thermography-based NDT technique and a long short term memory recurrent neural network (LSTM-RNN) model which automatically classifies common defects occurring in honeycomb materials. These including debonding, adhesive pooling, and liquid ingress. This LSTM-based algorithm has a greater than 90% sensitivity in classifying water, and hydraulic oil ingress. It has a greater than 70% sensitivity in classifying debonding and adhesive pooling.

[1]  Clemente Ibarra-Castanedo,et al.  Reliability assessment of pulsed thermography and ultrasonic testing for impact damage of CFRP panels , 2019, NDT & E International.

[2]  Luis S. Rosado,et al.  Defect Characterization With Eddy Current Testing Using Nonlinear-Regression Feature Extraction and Artificial Neural Networks , 2013, IEEE Transactions on Instrumentation and Measurement.

[3]  Jin-Hwe Kweon,et al.  Testing and modeling of Nomex™ honeycomb sandwich Panels with bolt insert , 2014 .

[4]  Clemente Ibarra-Castanedo,et al.  Impact Modelling and A Posteriori Non-destructive Evaluation of Homogeneous Particleboards of Sugarcane Bagasse , 2018, Journal of Nondestructive Evaluation.

[5]  Jürgen Schmidhuber,et al.  LSTM recurrent networks learn simple context-free and context-sensitive languages , 2001, IEEE Trans. Neural Networks.

[6]  Sebastian Dudzik,et al.  Two-stage neural algorithm for defect detection and characterization uses an active thermography , 2015 .

[7]  Clemente Ibarra-Castanedo,et al.  Optical and Mechanical Excitation Thermography for Impact Response in Basalt-Carbon Hybrid Fiber-Reinforced Composite Laminates , 2018, IEEE Transactions on Industrial Informatics.

[8]  Xavier Maldague,et al.  Optical excitation thermography for twill/plain weaves and stitched fabric dry carbon fibre preform inspection , 2018 .

[9]  Luis Volnei Sudati Sagrilo,et al.  MFL signals and artificial neural networks applied to detection and classification of pipe weld defects , 2006 .

[10]  Yunze He,et al.  Induction Infrared Thermography and Thermal-Wave-Radar Analysis for Imaging Inspection and Diagnosis of Blade Composites , 2018, IEEE Transactions on Industrial Informatics.

[11]  Ping Hu,et al.  Elastic constants for adhesively bonded corrugated core sandwich panels , 2017 .

[12]  Clemente Ibarra-Castanedo,et al.  Automated defect classification in infrared thermography based on a neural network , 2019, NDT & E International.

[13]  Jingmin Dai,et al.  Quantitative detection of defects based on Markov–PCA–BP algorithm using pulsed infrared thermography technology , 2016 .

[14]  Guy Lapalme,et al.  A systematic analysis of performance measures for classification tasks , 2009, Inf. Process. Manag..

[15]  Walid Tizani,et al.  Experimental and numerical analysis of dynamic compressive response of Nomex honeycombs , 2018, Composites Part B: Engineering.

[16]  Xavier Maldague,et al.  Neural network based defect detection and depth estimation in TNDE , 2002 .

[17]  Michał Sałaciński,et al.  The Effect of Environmental Flight Conditions on Damage Propagation in Composite Sandwich Structure , 2015 .

[18]  Jian Wang,et al.  Mechanical behavior of novel GFRP foam sandwich adhesive joints , 2017 .

[19]  A. Masnata,et al.  Neural network classification of flaws detected by ultrasonic means , 1996 .

[20]  Yunze He,et al.  Optically and non-optically excited thermography for composites: A review , 2016 .

[21]  Ronald Sterkenburg,et al.  Investigating the effects of fluid intrusion on Nomex® honeycomb sandwich structures with carbon fiber facesheets , 2018, Composite Structures.

[22]  Baldev Raj,et al.  An artificial neural network for eddy current testing of austenitic stainless steel welds , 2002 .

[23]  Yunlong Sheng,et al.  An experimental and analytical study of micro-laser line thermography on micro-sized flaws in stitched carbon fiber reinforced polymer composites , 2016 .

[24]  S. Sambath,et al.  Automatic Defect Classification in Ultrasonic NDT Using Artificial Intelligence , 2011 .

[25]  Yunze He,et al.  Shared Excitation Based Nonlinear Ultrasound and Vibrothermography Testing for CFRP Barely Visible Impact Damage Inspection , 2018, IEEE Transactions on Industrial Informatics.

[26]  Giovanni Maria Carlomagno,et al.  Impact damage in GFRP: New insights with infrared thermography , 2010 .

[27]  Jürgen Schmidhuber,et al.  LSTM can Solve Hard Long Time Lag Problems , 1996, NIPS.

[28]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[29]  Clemente Ibarra-Castanedo,et al.  ThermoPoD: A reliability study on active infrared thermography for the inspection of composite materials , 2012 .

[30]  Yasuo Suga,et al.  Application of artificial neural network to discrimination of defect type in automatic radiographic testing of welds , 1999 .

[31]  Salim Belouettar,et al.  Modelling the fatigue behaviour of composites honeycomb materials (aluminium/aramide fibre core) using four-point bending tests , 2010 .

[32]  Xavier Maldague,et al.  Fiber orientation assessment on randomly-oriented strand composites by means of infrared thermography , 2015 .

[33]  Xavier Maldague,et al.  Defect characterization in infrared non-destructive testing with learning machines , 2009 .

[34]  Yunze He,et al.  Impact evaluation in carbon fiber reinforced plastic (CFRP) laminates using eddy current pulsed thermography , 2014 .

[35]  Xavier Maldague,et al.  A study of defect depth using neural networks in pulsed phase thermography: modelling, noise, experiments , 1998 .

[36]  Vladimir P. Vavilov,et al.  Modelling, detecting and evaluating water ingress in aviation honeycomb panels , 2016 .