Morphological assessment of cavitation caused damage of cordierite and zircon based materials using principal component analysis
Apstrakt
The pattern recognition approach, explored by this study, applies the principal component analysis (PCA) as the most widely used statistical method with the aim of assessing the initiation and propagation of the cracks and defects that appear on the surface of material exposed to the cavitation. The experiment was performed in four stages: (a) synthesis of two ceramic materials (cordierite and zircon); (b) subjecting the samples to the cavitation; (c) using image analysis software for collecting the data about morphological characteristics that describe defects; (d) principal component analysis as a pattern recognition tool in order to characterize the defects at the material surface. Besides that, according to standard, cavitation erosion was monitored by determining material mass loss during the cavitation. Large experimental datasets collected from morphological descriptors by image analysis are multivariate and difficult to interpret, thus are processed by principal component analy...sis as the most informative technique for extracting possible differences. The performed approach proved that this method has a great potential for better assessment of induced defects by proper distinguishing among them at different levels and that can be considered a very efficient and cost-effective one.
Ključne reči:
Principal component analysis (PCA) / Morphology analysis / Ceramics / CavitationIzvor:
Engineering Failure Analysis, 2023, 148Izdavač:
- Elsevier Ltd
Finansiranje / projekti:
- Ministarstvo prosvete, nauke i tehnološkog razvoja Republike Srbije, Ugovor br. 200135 (Univerzitet u Beogradu, Tehnološko-metalurški fakultet) (RS-200135)
- Ministarstvo prosvete, nauke i tehnološkog razvoja Republike Srbije, Ugovor br. 200287 (Inovacioni centar Tehnološko-metalurškog fakulteta u Beogradu doo) (RS-200287)
- Ministarstvo prosvete, nauke i tehnološkog razvoja Republike Srbije, Ugovor br. 200026 (Univerzitet u Beogradu, Institut za hemiju, tehnologiju i metalurgiju - IHTM) (RS-200026)
- Ministarstvo prosvete, nauke i tehnološkog razvoja Republike Srbije, Ugovor br. 200023 (Institut za tehnologiju nuklearnih i drugih mineralnih sirovina - ITNMS, Beograd) (RS-200023)
DOI: 10.1016/j.engfailanal.2023.107224
ISSN: 1350-6307
Scopus: 2-s2.0-85151488584
Institucija/grupa
Institut za tehnologiju nuklearnih i drugih mineralnih sirovinaTY - JOUR AU - Martinović, S. AU - Alil, A. AU - Milićević, Sonja AU - Živojinović, D. AU - Volkov-Husović, Tatjana PY - 2023 UR - https://ritnms.itnms.ac.rs/handle/123456789/655 AB - The pattern recognition approach, explored by this study, applies the principal component analysis (PCA) as the most widely used statistical method with the aim of assessing the initiation and propagation of the cracks and defects that appear on the surface of material exposed to the cavitation. The experiment was performed in four stages: (a) synthesis of two ceramic materials (cordierite and zircon); (b) subjecting the samples to the cavitation; (c) using image analysis software for collecting the data about morphological characteristics that describe defects; (d) principal component analysis as a pattern recognition tool in order to characterize the defects at the material surface. Besides that, according to standard, cavitation erosion was monitored by determining material mass loss during the cavitation. Large experimental datasets collected from morphological descriptors by image analysis are multivariate and difficult to interpret, thus are processed by principal component analysis as the most informative technique for extracting possible differences. The performed approach proved that this method has a great potential for better assessment of induced defects by proper distinguishing among them at different levels and that can be considered a very efficient and cost-effective one. PB - Elsevier Ltd T2 - Engineering Failure Analysis T1 - Morphological assessment of cavitation caused damage of cordierite and zircon based materials using principal component analysis VL - 148 DO - 10.1016/j.engfailanal.2023.107224 UR - conv_982 ER -
@article{ author = "Martinović, S. and Alil, A. and Milićević, Sonja and Živojinović, D. and Volkov-Husović, Tatjana", year = "2023", abstract = "The pattern recognition approach, explored by this study, applies the principal component analysis (PCA) as the most widely used statistical method with the aim of assessing the initiation and propagation of the cracks and defects that appear on the surface of material exposed to the cavitation. The experiment was performed in four stages: (a) synthesis of two ceramic materials (cordierite and zircon); (b) subjecting the samples to the cavitation; (c) using image analysis software for collecting the data about morphological characteristics that describe defects; (d) principal component analysis as a pattern recognition tool in order to characterize the defects at the material surface. Besides that, according to standard, cavitation erosion was monitored by determining material mass loss during the cavitation. Large experimental datasets collected from morphological descriptors by image analysis are multivariate and difficult to interpret, thus are processed by principal component analysis as the most informative technique for extracting possible differences. The performed approach proved that this method has a great potential for better assessment of induced defects by proper distinguishing among them at different levels and that can be considered a very efficient and cost-effective one.", publisher = "Elsevier Ltd", journal = "Engineering Failure Analysis", title = "Morphological assessment of cavitation caused damage of cordierite and zircon based materials using principal component analysis", volume = "148", doi = "10.1016/j.engfailanal.2023.107224", url = "conv_982" }
Martinović, S., Alil, A., Milićević, S., Živojinović, D.,& Volkov-Husović, T.. (2023). Morphological assessment of cavitation caused damage of cordierite and zircon based materials using principal component analysis. in Engineering Failure Analysis Elsevier Ltd., 148. https://doi.org/10.1016/j.engfailanal.2023.107224 conv_982
Martinović S, Alil A, Milićević S, Živojinović D, Volkov-Husović T. Morphological assessment of cavitation caused damage of cordierite and zircon based materials using principal component analysis. in Engineering Failure Analysis. 2023;148. doi:10.1016/j.engfailanal.2023.107224 conv_982 .
Martinović, S., Alil, A., Milićević, Sonja, Živojinović, D., Volkov-Husović, Tatjana, "Morphological assessment of cavitation caused damage of cordierite and zircon based materials using principal component analysis" in Engineering Failure Analysis, 148 (2023), https://doi.org/10.1016/j.engfailanal.2023.107224 ., conv_982 .