Loading...

Predicting the Fatigue Life of Repaired Specimens by Composite Patch Exposed to Corrosive Environments Using Artificial Neural Network and Finite Element Method

Bakhshiyan, Amir Hossein | 2021

551 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 54031 (08)
  4. University: Sharif University of Technology
  5. Department: Mechanical Engineering
  6. Advisor(s): Farrahi, Gholamhossein
  7. Abstract:
  8. In this research, the application of composite patch in the repair of pipes damaged by corrosion has been investigated. Numerical modeling, artificial neural network and Taguchi algorithm are used for this purpose. In the numerical modeling section, the accuracy of modeling performance has been verified by experimental results of other researchers. Then, the effect of various parameters such as depth and, angle of corroded area, fiber orientation in the composite patch and angle of composite patch have been investigated. The depth and the angle of the corroded area and the angle of orientation of the fiber have been shown to have a large effect on the growth life of fatigue cracks. For example, when the depth of the corroded area increases from 32% of the pipe thickness to 65% of the pipe thickness, the fatigue crack growth life to reach a 48 mm crack length in the repaired pipe is reduced by about 30%. Also, repairing a damaged pipe to a depth of 32% of the pipe thickness will increase the fatigue crack growth life by 3.5 times compared to a pipe without repair, which shows the efficiency of this method in repairing damaged pipes.
    In the modeling, the separation between the composite patch and the pipe is modeled using the cohesive element. Subroutine has also been used to consider separation due to fatigue loading in the adhesive layer due to the lack of this feature in the software.In the fatigue life prediction section, Taguchi algorithm is used to design experiments and artificial neural network and Taguchi algorithm is used to predict fatigue crack growth life. It has been shown that both of these methods correctly predict the life of fatigue crack growth with acceptable accuracy and an error of less than 10% compared to the performed simulation
  9. Keywords:
  10. Artificial Neural Network ; Finite Element Modeling ; Fatigue Crack Growth Rate ; Composite Patch Repair ; Metallic Structures ; Fatigue-Corrosion ; Metal Structure Repair ; Fatigue Life Reduction ; Taguchi Method

 Digital Object List

 Bookmark

No TOC