scholarly journals Jensen-Shannon Divergence for Non-Destructive Incipient Crack Detection and Estimation

IEEE Access ◽  
2020 ◽  
Vol 8 ◽  
pp. 116148-116162
Author(s):  
Xiaoxia Zhang ◽  
Claude Delpha ◽  
Demba Diallo
2014 ◽  
Vol 658 ◽  
pp. 261-268
Author(s):  
Jean Louis Ntakpe ◽  
Gilbert Rainer Gillich ◽  
Florian Muntean ◽  
Zeno Iosif Praisach ◽  
Peter Lorenz

This paper presents a novel non-destructive method to locate and size damages in frame structures, performed by examining and interpreting changes in measured vibration response. The method bases on a relation, prior contrived by the authors, between the strain energy distribution in the structure for the transversal vibration modes and the modal changes (in terms of natural frequencies) due to damage. Using this relation a damage location indicator DLI was derived, which permits to locate cracks in spatial structures. In this paper an L-frame is considered for proving the applicability of this method. First the mathematical expressions for the modes shapes and their derivatives were determined and simulation result compared with that obtained by finite element analysis. Afterwards patterns characterizing damage locations were derived and compared with measurement results on the real structure; the DLI permitted accurate localization of any crack placed in the two structural elements.


Metals ◽  
2018 ◽  
Vol 8 (8) ◽  
pp. 612 ◽  
Author(s):  
Jue Hu ◽  
Weiping Xu ◽  
Bin Gao ◽  
Gui Tian ◽  
Yizhe Wang ◽  
...  

Eddy Current Pulsed Thermography is a crucial non-destructive testing technology which has a rapidly increasing range of applications for crack detection on metals. Although the unsupervised learning method has been widely adopted in thermal sequences processing, the research on supervised learning in crack detection remains unexplored. In this paper, we propose an end-to-end pattern, deep region learning structure to achieve precise crack detection and localization. The proposed structure integrates both time and spatial pattern mining for crack information with a deep region convolution neural network. Experiments on both artificial and natural cracks have shown attractive performance and verified the efficacy of the proposed structure.


2019 ◽  
Vol 278 ◽  
pp. 03006
Author(s):  
Björn Torsten Salmen ◽  
Marina Knyazeva ◽  
Frank Walther

Due to the increasing volume of traffic, bridges are exposed to higher loads as it was considered during the planning phase. Therefore, a regular inspection is necessary in order to detect cracks at very early stages. The use of weathering structural steel in bridges, as well as in composite bridge constructions is an alternative to conventional bridges, not only from an economic but also from an ecological point of view, since it is not necessary to apply a corrosion protection layer and renew it during the lifetime of the bridge. Unfortunately, conventional visual inspection or magnetic particle inspection on the weathering steel bridge are hindered by the protective patina and requires development of new test methods. Within the framework of this project, a combined crack detection technique using non-destructive inspection by means of Active Thermography and by Electro-Magnetic Acoustic Transducer (EMAT) were evaluated in laboratory environments and in real conditions on bridge structures made of weathering structural steel.


Sign in / Sign up

Export Citation Format

Share Document