| Title |
Fault Detection Method of Tapered Cantilever Pipe-type Beam |
| DOI |
http://dx.doi.org/10.5659/JAIK_SC.2015.31.2.13 |
| Keywords |
Crack Identification ; Committee of Neural Networks ; Modal Property ; Tapered Cantilever Beam |
| Abstract |
A crack identification method using an equivalent bending stiffness for cracked beam and committee of neural networks is presented. Modal properties of tapered cantilever pipe-type beam is identified by applying the boundary conditions to a general solution for tapered beam. A bending stiffness for cracked beam based on an energy method is used to identify natural frequencies and mode shapes of tapered cantilever thin-walled pipe, which has a through-the-thickness crack, subjected to bending. The identified modal properties of the cracked beam are used in constructing training patterns of neural networks. Then crack location and size are identified using a committee of the neural networks. Crack detection was carried out for an example beam using the proposed method, and the identified crack locations and sizes agree reasonably well with the exact values. |