DETERMINING THE BURST PRESSURE OF COMPOSITE PRESSURE BOTTLES USING ACOUSTIC EMISSION RESPONSE
Keywords:
Structural Integrity, Burst Pressure, Acoustic Emission, Neural Network, Composite Pressure VesselsAbstract
Acoustic emission (AE) Nondestructive testing was carried out during the hydrostatic loading of five identical glass fiber reinforced pressure bottles. The AE data acquired upto 50% of the theoretical burst pressure was recorded; the bottles were pressurized till failure. The Amplitude frequency distribution of AE data, maximum dilation and fiber strain at various locations were given as the inputs and the corresponding burst pressures were given as the targeted output for the supervised back propagation neural network. Architecturally 64-16-16-1, net work was able to map the patterns present in the AE signals, which lead to the burst failure of the pressure vessels. The network trained with the data generated from three bottles of the maximum, minimum and average burst pressures was able to predict the burst pressure of the remaining two bottles with a worst case prediction error of 3.49 % well within the desired goal of ±5 percent.
Downloads
References
Eric V K Hill, James L Walker and Ginger H Rowell (1996), “Burst Pressure Prediction in Graphite/Epoxy Pressure Vessels using Neural Networks and Acoustic Emission Amplitude Data”, Material Evaluation, Vol.54, 744-748
Marcus E Fisher and Eric V K Hill (1998), “Neural Network Burst Pressure Prediction in Fiber Glass Epoxy Pressure Vessels using Acoustic Emission”, Material Evaluation, Vol. 56, 1395-1401.
American society for Testing Materials (ASTM E 1316), “Standard Terminology for Non- Destructive Examinations, in Annual Book of ASTM standards, Non-Destructive Testing, ASTM, Philadelphia”, Vol.3.
Sivanandam and Sumathi S (2007), “Introduction to Neural Networks Using MATLAB 6.0”, The Mc Graw-Hill Publishing Company ltd., New Delhi.
Kalloo and Frederick R (1988), “Predicting Burst Pressure in Filament Wound Composite Vessels using Acoustic Emission Data”, M.S Thesis, Embry-Riddle Aeronautical University.
Walker and James L (1990), “Composite Structure Ultimate Strength Prediction from Acoustic Emission Amplitude Data”, M.S Thesis, Embry-Riddle Aeronautical University.
Walker, James L and Eric V K Hill (1991), “Composite Ultimate Strength Prediction from Acoustic Emission Amplitude data”, Third conference on Non-destructive evaluation for Aerospace requirements, Huntsville, Alabama, June 4-6.
Sasikumar T and Rajendra Boopathy S (2008), “Artificial Neural Network Prediction of Ultimate Strength of Unidirectional T300/914 Tensile Specimens using Acoustic Emission Response”, Journal of Non-destructive Evaluation, Vol.12, 127-133.
Sathish kumar (2004), “Neural Networks- A class room Approach”, Tata Mc Graw-Hill publishing company Ltd., New Delhi.
