Loading Frequency Classification in Shape Memory Alloys: A Machine Learning Approach DOI Creative Commons
Dmytro Tymoshchuk,

Oleh Yasniy,

Pavlo Maruschak

et al.

Computers, Journal Year: 2024, Volume and Issue: 13(12), P. 339 - 339

Published: Dec. 14, 2024

This paper investigates the use of machine learning methods to predict loading frequency shape memory alloys (SMAs) based on experimental data. SMAs, in particular nickel-titanium (NiTi) alloys, have unique properties that restore original after significant deformation. The significantly affects functional characteristics SMAs. Experimental data were obtained from cyclic tensile tests a 1.5 mm diameter Ni55.8Ti44.2 wire at different frequencies (0.1, 0.5, 1.0, and 5.0 Hz). Various used f (Hz) input parameters such as stress σ (MPa), number cycles N, strain ε (%), loading–unloading stage: boosted trees, random forest, support vector machines, k-nearest neighbors, artificial neural networks MLP type. 100–140 load–unload for four load training. dataset contained 13,365 elements. results showed network model demonstrated highest accuracy classification. trees forest models also performed well, although slightly below MLP. SVM method quite well. KNN worst among all models. Additional testing not included training (200th, 300th, 1035th cycles) retains high efficiency predicting frequency, gradually decreases later due accumulation structural changes material.

Language: Английский

Photodiode Signal Patterns: Unsupervised Learning for Laser Weld Defect Analysis DOI Open Access
Erkan Caner Ozkat

Processes, Journal Year: 2025, Volume and Issue: 13(1), P. 121 - 121

Published: Jan. 5, 2025

Laser welding, widely used in industries such as automotive and aerospace, requires precise monitoring to ensure defect-free welds, especially when joining dissimilar metallic thin foils. This study investigates the application of machine learning techniques for defect detection laser welding using photodiode signal patterns. Supervised models, including Support Vector Machine (SVM), k-Nearest Neighbors (kNN), Random Forest (RF), were employed classify weld defects into sound welds (SW), lack connection (LoC), over-penetration (OP). SVM achieved highest accuracy (95.2%) during training, while RF demonstrated superior generalization with 83% on validation data. The also proposed an unsupervised method a wavelet scattering one-dimensional convolutional autoencoder (1D-CAE) network anomaly detection. its effectiveness achieving accuracies 93.3% 87.5% training datasets, respectively. Furthermore, distinct patterns associated SW, OP, LoC identified, highlighting ability signals capture dynamics. These findings demonstrate combining supervised methods detection, paving way robust, real-time quality systems manufacturing. results indicated that could offer significant advantages identifying anomalies reducing manufacturing costs.

Language: Английский

Citations

0
Evolution of the Fatigue Failure Prediction Process from Experiment to Artificial Intelligence: A Review DOI Open Access
Cornel Samoilă, Doru Ursuțiu, Iuliana Tudorache

et al.

Materials, Journal Year: 2025, Volume and Issue: 18(5), P. 1153 - 1153

Published: March 4, 2025

An analysis of the time evolution fatigue break prediction shows increasingly shorter developmental stages. The experimental period was longest; combination more powerful mathematical methods led to a leap in and shortening implementation time. All rupture have proven limitations due multitude influencing factors insufficient number practical considered. Recently, attempts been made increase accuracy by combining based on physical mechanisms failure process with data-driven assisted artificial intelligence. We attempt present this herein. There are several review suitable for analyzing subject: systematic, semi-systematic, integrative. From these, semi-systematic integrative chosen precisely because two complement each other.

Language: Английский

Citations

0
Predicting high-cycle fatigue strength of precipitation-hardened Nickel-Based superalloys from transfer learning DOI
Zeyu Chen,

ZhaoJing Han,

ShengBao Xia

et al.

Engineering Fracture Mechanics, Journal Year: 2025, Volume and Issue: unknown, P. 111087 - 111087

Published: April 1, 2025

Language: Английский

Citations

0
Loading Frequency Classification in Shape Memory Alloys: A Machine Learning Approach DOI Creative Commons
Dmytro Tymoshchuk,

Oleh Yasniy,

Pavlo Maruschak

et al.

Computers, Journal Year: 2024, Volume and Issue: 13(12), P. 339 - 339

Published: Dec. 14, 2024

This paper investigates the use of machine learning methods to predict loading frequency shape memory alloys (SMAs) based on experimental data. SMAs, in particular nickel-titanium (NiTi) alloys, have unique properties that restore original after significant deformation. The significantly affects functional characteristics SMAs. Experimental data were obtained from cyclic tensile tests a 1.5 mm diameter Ni55.8Ti44.2 wire at different frequencies (0.1, 0.5, 1.0, and 5.0 Hz). Various used f (Hz) input parameters such as stress σ (MPa), number cycles N, strain ε (%), loading–unloading stage: boosted trees, random forest, support vector machines, k-nearest neighbors, artificial neural networks MLP type. 100–140 load–unload for four load training. dataset contained 13,365 elements. results showed network model demonstrated highest accuracy classification. trees forest models also performed well, although slightly below MLP. SVM method quite well. KNN worst among all models. Additional testing not included training (200th, 300th, 1035th cycles) retains high efficiency predicting frequency, gradually decreases later due accumulation structural changes material.

Language: Английский

Citations

0