In response to five failures since 2008 of the tail gearbox of multiple models of Sikorsky's H-60 helicopter, acoustic emission (AE) data collected from a rotating gearbox test stand at the Naval Air Station in Patuxtent River, MD, was used to monitor the initiation and propagation of a flaw from an electro-discharge machined (EDM) notch seeded on the face of a gear tooth. A period of testing was considered which spanned ~300,000 seconds or ~83 hours and culminates to a damage state such that a flaw has initiated on both ends of the EDM notch. AE data was analyzed for three separate channels which span a wide range of amplitude thresholds using clustering methods and verification algorithms developed at the Embry-Riddle Aeronautical University (ERAU) Structure Health Monitoring (SHM) and Nondestructive Evaluation (NDE) Laboratory. Energy, duration, amplitude, and average frequency of the AE signals were input into the Kohonen Self-Organizing Map (KSOM) artificial neural network (ANN) function in NeuralWorks Professional II/Plus software to separate cracking signals from other mechanisms such as noise and plastic deformation. Visual inspection and statistical analysis of the data in the AE plots created using the output ANN results was used to separate the cluster(s) which exhibited higher amplitude and energy, and lower duration and average frequency; hits typical to cracking. The similarities and differences in the progression of clusters sourced to cracking for each of the three channels is discussed. Cumulative testing time plots of AE parameters were compiled using both entire data sets and using clusters representative of cracking mechanisms. Replica cross sections which were taken throughout testing visually display, in chronological fashion, circumferential crack growth across gear splines adjacent to the spline with the EDM notch. Data analysis techniques are used in conjunction with replica cross sections to provide insight into the AE activity for crack initiation and crack propagation and define early damage state detection criterion for rotary components.
","abstract_html":"<p>In response to five failures since 2008 of the tail gearbox of multiple models of Sikorsky's H-60 helicopter, acoustic emission (AE) data collected from a rotating gearbox test stand at the Naval Air Station in Patuxtent River, MD, was used to monitor the initiation and propagation of a flaw from an electro-discharge machined (EDM) notch seeded on the face of a gear tooth. A period of testing was considered which spanned ~300,000 seconds or ~83 hours and culminates to a damage state such that a flaw has initiated on both ends of the EDM notch. AE data was analyzed for three separate channels which span a wide range of amplitude thresholds using clustering methods and verification algorithms developed at the <a href="http://search.proquest.com.ezproxy.libproxy.db.erau.edu/pqdtglobal/docview/1641123390/6DFEAD316474499APQ/1?accountid=27203#">Embry-Riddle Aeronautical University</a> (ERAU) Structure Health Monitoring (SHM) and Nondestructive Evaluation (NDE) Laboratory. Energy, duration, amplitude, and average frequency of the AE signals were input into the Kohonen Self-Organizing Map (KSOM) artificial neural network (ANN) function in NeuralWorks Professional II/Plus software to separate cracking signals from other mechanisms such as noise and plastic deformation. Visual inspection and statistical analysis of the data in the AE plots created using the output ANN results was used to separate the cluster(s) which exhibited higher amplitude and energy, and lower duration and average frequency; hits typical to cracking. The similarities and differences in the progression of clusters sourced to cracking for each of the three channels is discussed. Cumulative testing time plots of AE parameters were compiled using both entire data sets and using clusters representative of cracking mechanisms. Replica cross sections which were taken throughout testing visually display, in chronological fashion, circumferential crack growth across gear splines adjacent to the spline with the EDM notch. Data analysis techniques are used in conjunction with replica cross sections to provide insight into the AE activity for crack initiation and crack propagation and define early damage state detection criterion for rotary components.</p>","abstract_has_math":false,"creators":["Baum, Rudy L."],"institution":null,"degree_name":"Master of Science in Mechanical Engineering","degree_level":"Thesis - Open Access","degree_discipline":"Mechanical Engineering","degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2014,"date_issued":"2014-07-01T07:00:00Z","date_published":"2014-07-01T07:00:00Z","updated_at":"2026-07-27T19:26:22Z","subjects":["helicopters","acoustic emission","neural networks","Aerospace Engineering","Mechanical Engineering"],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://commons.erau.edu/edt/20","outbound_label":"Repository record","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:creator","label":"Author","values":["Baum, Rudy L."]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"thesis:degree_discipline","label":"Discipline","values":["Mechanical Engineering"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Thesis - Open Access"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Master of Science in Mechanical Engineering"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["helicopters","acoustic emission","neural networks","Aerospace Engineering","Mechanical Engineering"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://commons.erau.edu/edt/20"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["In response to five failures since 2008 of the tail gearbox of multiple models of Sikorsky's H-60 helicopter, acoustic emission (AE) data collected from a rotating gearbox test stand at the Naval Air Station in Patuxtent River, MD, was used to monitor the initiation and propagation of a flaw from an electro-discharge machined (EDM) notch seeded on the face of a gear tooth. A period of testing was considered which spanned ~300,000 seconds or ~83 hours and culminates to a damage state such that a flaw has initiated on both ends of the EDM notch. AE data was analyzed for three separate channels which span a wide range of amplitude thresholds using clustering methods and verification algorithms developed at the Embry-Riddle Aeronautical University (ERAU) Structure Health Monitoring (SHM) and Nondestructive Evaluation (NDE) Laboratory. Energy, duration, amplitude, and average frequency of the AE signals were input into the Kohonen Self-Organizing Map (KSOM) artificial neural network (ANN) function in NeuralWorks Professional II/Plus software to separate cracking signals from other mechanisms such as noise and plastic deformation. Visual inspection and statistical analysis of the data in the AE plots created using the output ANN results was used to separate the cluster(s) which exhibited higher amplitude and energy, and lower duration and average frequency; hits typical to cracking. The similarities and differences in the progression of clusters sourced to cracking for each of the three channels is discussed. Cumulative testing time plots of AE parameters were compiled using both entire data sets and using clusters representative of cracking mechanisms. Replica cross sections which were taken throughout testing visually display, in chronological fashion, circumferential crack growth across gear splines adjacent to the spline with the EDM notch. Data analysis techniques are used in conjunction with replica cross sections to provide insight into the AE activity for crack initiation and crack propagation and define early damage state detection criterion for rotary components.
"]},{"key":"dc:title","label":"Title","values":["Early Damage State Criterion from a Fault-Seeded Helicopter Gear Using Acoustic Emission and Neural Networks"]}]}],"canonical_facts":{"dc:creator":["Baum, Rudy L."],"dc:description.abstract":["In response to five failures since 2008 of the tail gearbox of multiple models of Sikorsky's H-60 helicopter, acoustic emission (AE) data collected from a rotating gearbox test stand at the Naval Air Station in Patuxtent River, MD, was used to monitor the initiation and propagation of a flaw from an electro-discharge machined (EDM) notch seeded on the face of a gear tooth. A period of testing was considered which spanned ~300,000 seconds or ~83 hours and culminates to a damage state such that a flaw has initiated on both ends of the EDM notch. AE data was analyzed for three separate channels which span a wide range of amplitude thresholds using clustering methods and verification algorithms developed at the Embry-Riddle Aeronautical University (ERAU) Structure Health Monitoring (SHM) and Nondestructive Evaluation (NDE) Laboratory. Energy, duration, amplitude, and average frequency of the AE signals were input into the Kohonen Self-Organizing Map (KSOM) artificial neural network (ANN) function in NeuralWorks Professional II/Plus software to separate cracking signals from other mechanisms such as noise and plastic deformation. Visual inspection and statistical analysis of the data in the AE plots created using the output ANN results was used to separate the cluster(s) which exhibited higher amplitude and energy, and lower duration and average frequency; hits typical to cracking. The similarities and differences in the progression of clusters sourced to cracking for each of the three channels is discussed. Cumulative testing time plots of AE parameters were compiled using both entire data sets and using clusters representative of cracking mechanisms. Replica cross sections which were taken throughout testing visually display, in chronological fashion, circumferential crack growth across gear splines adjacent to the spline with the EDM notch. Data analysis techniques are used in conjunction with replica cross sections to provide insight into the AE activity for crack initiation and crack propagation and define early damage state detection criterion for rotary components.
"],"dc:identifier":["https://commons.erau.edu/edt/20"],"dc:subject":["helicopters","acoustic emission","neural networks","Aerospace Engineering","Mechanical Engineering"],"dc:title":["Early Damage State Criterion from a Fault-Seeded Helicopter Gear Using Acoustic Emission and Neural Networks"],"thesis:degree_discipline":["Mechanical Engineering"],"thesis:degree_level":["Thesis - Open Access"],"thesis:degree_name":["Master of Science in Mechanical Engineering"]},"updated_at":"2026-07-27T19:26:22Z"}