Datafication of material mechanisms in nuclear environment
Publikationen: Thesis / Studienabschlussarbeiten und Habilitationsschriften › Masterarbeit
Standard
2020.
Publikationen: Thesis / Studienabschlussarbeiten und Habilitationsschriften › Masterarbeit
Harvard
APA
Vancouver
Author
Bibtex - Download
}
RIS (suitable for import to EndNote) - Download
TY - THES
T1 - Datafication of material mechanisms in nuclear environment
AU - Reich, René
N1 - embargoed until null
PY - 2020
Y1 - 2020
N2 - Material selection relies on the information of feasible materials, especially in nuclear applications, where material tests are complex and require special attention to safety. A more profound material selection decision one can make with the increasing availability of the information. Peer-reviewed scientific articles present a vital source with high quality. This thesis studies the possibilities of information extraction and information summarization of the properties of metallic alloys. The following challenges for this purpose were identified: Fully automated information extraction by machine learning algorithms will become possible with the creation of materials science related corpus for natural language processing. Also, an automated plot digitizer would excavate enormous amounts of material data. With the increasing size of the database, its clarity decreases. So, the contained information should be summarized comprehensibly. The approach presented in this thesis is a material mechanism map. Based on the idea of Ashby maps, material mechanism maps visualize areas of material property changes under specific environmental conditions. As an example, a material mechanism map for the austenitic steel SS316, serving as nuclear fuel-cladding material, was computed. The map contains information about material hardening, recovery, irradiation embrittlement, swelling, creep, and precipitation formation depending on the irradiation dose and the homologous temperature.
AB - Material selection relies on the information of feasible materials, especially in nuclear applications, where material tests are complex and require special attention to safety. A more profound material selection decision one can make with the increasing availability of the information. Peer-reviewed scientific articles present a vital source with high quality. This thesis studies the possibilities of information extraction and information summarization of the properties of metallic alloys. The following challenges for this purpose were identified: Fully automated information extraction by machine learning algorithms will become possible with the creation of materials science related corpus for natural language processing. Also, an automated plot digitizer would excavate enormous amounts of material data. With the increasing size of the database, its clarity decreases. So, the contained information should be summarized comprehensibly. The approach presented in this thesis is a material mechanism map. Based on the idea of Ashby maps, material mechanism maps visualize areas of material property changes under specific environmental conditions. As an example, a material mechanism map for the austenitic steel SS316, serving as nuclear fuel-cladding material, was computed. The map contains information about material hardening, recovery, irradiation embrittlement, swelling, creep, and precipitation formation depending on the irradiation dose and the homologous temperature.
KW - material informatics
KW - information extraction
KW - material data extractor
KW - machine learning
KW - natural language processing
KW - information gathering
KW - material database
KW - information summarization
KW - information visualization
KW - Ashby map
KW - material mechanism map
KW - nuclear materials
KW - fuel-cladding material
KW - austenitic stainless steel
KW - SS316
KW - irradiation
KW - dose
KW - temperature
KW - hardening
KW - recovery
KW - embrittlement
KW - swelling
KW - creep
KW - precipitation formation
KW - Materialinformatik
KW - Informationsextraktion
KW - Materialdaten-Extraktor
KW - Maschinelles Lernen
KW - Verarbeitung natürlicher Sprache
KW - Informationsbeschaffung
KW - Materialdatenbank
KW - Informationsvisualisierung
KW - Informationszusammenfassung
KW - Ashby-Plot
KW - Materialmechanismen-Plot
KW - Werkstoffe in nuklearen Anwendungen
KW - Hüllrohrmaterial für Krennbrennstäbe
KW - austenitischer Edelstahl
KW - SS316
KW - Bestrahlung
KW - Dosis
KW - Temperatur
KW - Verfestigung
KW - Erholung
KW - Versprödung
KW - Materialschwellung
KW - Kriechen
KW - Ausscheidungsbildung
M3 - Master's Thesis
ER -