Title: ANN-based prediction of ultimate strength of carbon/epoxy tensile specimen using acoustic emission RMS data
Authors: Kalidasan Krishnamoorthy; T. Sasikumar
Addresses: Center for Research, Lord Jegannath College of Engineering and Technology, Kanyakumari District, Tamilnadu State, India ' Department of Mechanical Engineering, Lord Jegannath College of Engineering and Technology, Kanyakumari District, Tamilnadu State, India
Abstract: Acoustic emission (AE) is a phenomenon very widely used to predict the ultimate strength of fibre reinforced plastic composites. The ultimate strength of the carbon/epoxy tensile specimens was predicted, using the artificial neural network (ANN). The 15 numbers of carbon/epoxy composite specimens were fabricated as per ASTM D 3039 standards. These specimens were loaded with a 10 TON capacity universal tensile machine. AE data were collected up to 70% of the failure load. AE parameters like amplitude, duration, energy, count and RMS values were collected. The RMS value corresponding to the amplitude ranges obtained during tensile testing were used to predict the failure load of a similar specimen subjected to uniaxial tension well before its failure load.
Keywords: artificial neural networks; ANNs; back propagation; acoustic emission; carbon-epoxy tensile specimens; RMS values; root mean square; ultimate strength; fibre reinforced plastic composites; FRPC; failure load; amplitude ranges; tensile testing.
DOI: 10.1504/IJMPT.2016.076374
International Journal of Materials and Product Technology, 2016 Vol.53 No.1, pp.61 - 70
Received: 23 Jun 2014
Accepted: 23 Sep 2015
Published online: 05 May 2016 *