The purpose of this research was to investigate the effectiveness of artificial neural networks (ANNs) in predicting the compression after impact (CAI) load of graphite/epoxy laminates from acoustic emission (AE) nondestructive testing (NDT) data. Thirty-four 24-ply bidirectional woven cloth laminate coupons were constructed and impacted at various energy levels ranging from 8 to 20 Joules, generating barely visible impact damage (BVID). Acoustic emission data were acquired as the coupons were compressed to failure. Not having been analyzed by previous experimenters, several noise tests were also performed to determine the impact of external noise on acoustic emission data during testing. Once the noise and other erroneous data were filtered out, several investigations were conducted using ANNs. First, a Kohonen self-organizing map (SOM) neural network was constructed and optimized in order to separate the AE data into noise plus the various failure mechanisms thought to be experienced by composite laminates undergoing compression. It was hoped that by quantifying the failure mechanisms, more accurate ultimate load predictions could be made. Secondly, a backpropagation neural network (BPNN) was constructed, which analyzed the AE amplitude distributions directly as inputs in order to arrive at accurate CAI load predictions. The BPNN was trained using twenty-four of the samples, systematically optimized, and then tested on the remaining ten samples. The relatively large sample size allowed both the SOM and the BPNN to experience a wide variety of failure scenarios, thus leading to very good classification and prediction results. The worst case error from the prediction results was found to be -11.53%, which was within the desired prediction error range of ±15%. Microscopy and C-scans were also an important part of the project in order to analyze the extent of damage created by impact and compression after impact. It was hoped that these methods would allow a better understanding of the failure mechanisms in the material.
","abstract_html":"<p>The purpose of this research was to investigate the effectiveness of artificial neural networks (ANNs) in predicting the compression after impact (CAI) load of graphite/epoxy laminates from acoustic emission (AE) nondestructive testing (NDT) data. Thirty-four 24-ply bidirectional woven cloth laminate coupons were constructed and impacted at various energy levels ranging from 8 to 20 Joules, generating barely visible impact damage (BVID). Acoustic emission data were acquired as the coupons were compressed to failure. Not having been analyzed by previous experimenters, several noise tests were also performed to determine the impact of external noise on acoustic emission data during testing. Once the noise and other erroneous data were filtered out, several investigations were conducted using ANNs. First, a Kohonen self-organizing map (SOM) neural network was constructed and optimized in order to separate the AE data into noise plus the various failure mechanisms thought to be experienced by composite laminates undergoing compression. It was hoped that by quantifying the failure mechanisms, more accurate ultimate load predictions could be made. Secondly, a backpropagation neural network (BPNN) was constructed, which analyzed the AE amplitude distributions directly as inputs in order to arrive at accurate CAI load predictions. The BPNN was trained using twenty-four of the samples, systematically optimized, and then tested on the remaining ten samples. The relatively large sample size allowed both the SOM and the BPNN to experience a wide variety of failure scenarios, thus leading to very good classification and prediction results. The worst case error from the prediction results was found to be -11.53%, which was within the desired prediction error range of ±15%. Microscopy and C-scans were also an important part of the project in order to analyze the extent of damage created by impact and compression after impact. It was hoped that these methods would allow a better understanding of the failure mechanisms in the material.</p>","abstract_has_math":false,"creators":["Gunasekera, Anthony Michael"],"institution":null,"degree_name":"Master of Aerospace Engineering","degree_level":"Thesis - Open Access","degree_discipline":"Aerospace Engineering","degree_department":null,"school":null,"contributors":["Eric v. K. Hill","Yi Zhao","David J. Sypeck"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2009,"date_issued":"2009-10-01T07:00:00Z","date_published":"2009-10-01T07:00:00Z","updated_at":"2026-07-27T19:25:45Z","subjects":["compression","impact load","graphite/epoxy","laminates","acoustic emission","artificial neural networks","Aerospace Engineering","Structures and Materials"],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://commons.erau.edu/db-theses/279","outbound_label":"Repository record","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Eric v. K. Hill","Yi Zhao","David J. Sypeck"]},{"key":"dc:creator","label":"Author","values":["Gunasekera, Anthony Michael"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"thesis:degree_discipline","label":"Discipline","values":["Aerospace Engineering"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Thesis - Open Access"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Master of Aerospace Engineering"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["compression","impact load","graphite/epoxy","laminates","acoustic emission","artificial neural networks","Aerospace Engineering","Structures and Materials"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://commons.erau.edu/db-theses/279"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["The purpose of this research was to investigate the effectiveness of artificial neural networks (ANNs) in predicting the compression after impact (CAI) load of graphite/epoxy laminates from acoustic emission (AE) nondestructive testing (NDT) data. Thirty-four 24-ply bidirectional woven cloth laminate coupons were constructed and impacted at various energy levels ranging from 8 to 20 Joules, generating barely visible impact damage (BVID). Acoustic emission data were acquired as the coupons were compressed to failure. Not having been analyzed by previous experimenters, several noise tests were also performed to determine the impact of external noise on acoustic emission data during testing. Once the noise and other erroneous data were filtered out, several investigations were conducted using ANNs. First, a Kohonen self-organizing map (SOM) neural network was constructed and optimized in order to separate the AE data into noise plus the various failure mechanisms thought to be experienced by composite laminates undergoing compression. It was hoped that by quantifying the failure mechanisms, more accurate ultimate load predictions could be made. Secondly, a backpropagation neural network (BPNN) was constructed, which analyzed the AE amplitude distributions directly as inputs in order to arrive at accurate CAI load predictions. The BPNN was trained using twenty-four of the samples, systematically optimized, and then tested on the remaining ten samples. The relatively large sample size allowed both the SOM and the BPNN to experience a wide variety of failure scenarios, thus leading to very good classification and prediction results. The worst case error from the prediction results was found to be -11.53%, which was within the desired prediction error range of ±15%. Microscopy and C-scans were also an important part of the project in order to analyze the extent of damage created by impact and compression after impact. It was hoped that these methods would allow a better understanding of the failure mechanisms in the material.
"]},{"key":"dc:title","label":"Title","values":["Compression After Impact Load Prediction in Graphite/Epoxy Laminates Using Acoustic Emission and Artificial Neural Networks"]}]}],"canonical_facts":{"dc:contributor":["Eric v. K. Hill","Yi Zhao","David J. Sypeck"],"dc:creator":["Gunasekera, Anthony Michael"],"dc:description.abstract":["The purpose of this research was to investigate the effectiveness of artificial neural networks (ANNs) in predicting the compression after impact (CAI) load of graphite/epoxy laminates from acoustic emission (AE) nondestructive testing (NDT) data. Thirty-four 24-ply bidirectional woven cloth laminate coupons were constructed and impacted at various energy levels ranging from 8 to 20 Joules, generating barely visible impact damage (BVID). Acoustic emission data were acquired as the coupons were compressed to failure. Not having been analyzed by previous experimenters, several noise tests were also performed to determine the impact of external noise on acoustic emission data during testing. Once the noise and other erroneous data were filtered out, several investigations were conducted using ANNs. First, a Kohonen self-organizing map (SOM) neural network was constructed and optimized in order to separate the AE data into noise plus the various failure mechanisms thought to be experienced by composite laminates undergoing compression. It was hoped that by quantifying the failure mechanisms, more accurate ultimate load predictions could be made. Secondly, a backpropagation neural network (BPNN) was constructed, which analyzed the AE amplitude distributions directly as inputs in order to arrive at accurate CAI load predictions. The BPNN was trained using twenty-four of the samples, systematically optimized, and then tested on the remaining ten samples. The relatively large sample size allowed both the SOM and the BPNN to experience a wide variety of failure scenarios, thus leading to very good classification and prediction results. The worst case error from the prediction results was found to be -11.53%, which was within the desired prediction error range of ±15%. Microscopy and C-scans were also an important part of the project in order to analyze the extent of damage created by impact and compression after impact. It was hoped that these methods would allow a better understanding of the failure mechanisms in the material.
"],"dc:identifier":["https://commons.erau.edu/db-theses/279"],"dc:subject":["compression","impact load","graphite/epoxy","laminates","acoustic emission","artificial neural networks","Aerospace Engineering","Structures and Materials"],"dc:title":["Compression After Impact Load Prediction in Graphite/Epoxy Laminates Using Acoustic Emission and Artificial Neural Networks"],"thesis:degree_discipline":["Aerospace Engineering"],"thesis:degree_level":["Thesis - Open Access"],"thesis:degree_name":["Master of Aerospace Engineering"]},"updated_at":"2026-07-27T19:25:45Z"}