Details
| Originalsprache | Englisch |
|---|---|
| Titel des Sammelwerks | Proceedings of the 29th European Safety and Reliability Conference, ESREL 2019 |
| Herausgeber/-innen | Michael Beer, Enrico Zio |
| Seiten | 2693-2698 |
| Seitenumfang | 6 |
| ISBN (elektronisch) | 978-981-11-0745-0 |
| Publikationsstatus | Veröffentlicht - 2019 |
| Veranstaltung | 29th European Safety and Reliability Conference, ESREL 2019 - Leibniz University Hannover, Hannover, Deutschland Dauer: 22 Sept. 2019 → 26 Sept. 2019 |
Abstract
Within this work, a probability box approach is investigated to capture mixed aleatory and epistemic uncertainties within non-linear finite element method. The approach is applied to brittle damage problems regarding one and two input random fields. While random fields describe naturally aleatory uncertainty, the epistemic part is introduced by an interval-valued correlation length. The random field is discretized by Karhunen-Loève expansion. To avoid the truncation error affecting the probability box, the truncation error is kept constant with regard to the different assumed correlation lengths. Outcome of interest are the probability boxes of the local and the global damage of a four-point bending simulation of a concrete beam. It is shown that the correlation length mainly affects the standard deviation but not the mean value of the outcome. Furthermore, despite the non-linearity of the problem, it can be shown that the probability box is described by the correlation length interval bounds only for this example.
ASJC Scopus Sachgebiete
- Ingenieurwesen (insg.)
- Sicherheit, Risiko, Zuverlässigkeit und Qualität
- Sozialwissenschaften (insg.)
- Sicherheitsforschung
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
Proceedings of the 29th European Safety and Reliability Conference, ESREL 2019. Hrsg. / Michael Beer; Enrico Zio. 2019. S. 2693-2698.
Publikation: Beitrag in Buch/Bericht/Sammelwerk/Konferenzband › Aufsatz in Konferenzband › Forschung › Peer-Review
}
TY - GEN
T1 - Non-linear finite element analysis under mixed epistemic and aleatory uncertain random field input
AU - Dannert, Mona M.
AU - Fleury, Rodolfo M. N.
AU - Fau, Amelie
AU - Nackenhorst, Udo
N1 - Funding information: The demonstrated results arise from the project Sophisticated computational techniques for damage mechanics with mixed uncertain input fields (NA 330/12-1), a part of the Priority Programme SPP 1886. The funding by the German Research Foundation (DFG) is gratefully acknowledged.
PY - 2019
Y1 - 2019
N2 - Within this work, a probability box approach is investigated to capture mixed aleatory and epistemic uncertainties within non-linear finite element method. The approach is applied to brittle damage problems regarding one and two input random fields. While random fields describe naturally aleatory uncertainty, the epistemic part is introduced by an interval-valued correlation length. The random field is discretized by Karhunen-Loève expansion. To avoid the truncation error affecting the probability box, the truncation error is kept constant with regard to the different assumed correlation lengths. Outcome of interest are the probability boxes of the local and the global damage of a four-point bending simulation of a concrete beam. It is shown that the correlation length mainly affects the standard deviation but not the mean value of the outcome. Furthermore, despite the non-linearity of the problem, it can be shown that the probability box is described by the correlation length interval bounds only for this example.
AB - Within this work, a probability box approach is investigated to capture mixed aleatory and epistemic uncertainties within non-linear finite element method. The approach is applied to brittle damage problems regarding one and two input random fields. While random fields describe naturally aleatory uncertainty, the epistemic part is introduced by an interval-valued correlation length. The random field is discretized by Karhunen-Loève expansion. To avoid the truncation error affecting the probability box, the truncation error is kept constant with regard to the different assumed correlation lengths. Outcome of interest are the probability boxes of the local and the global damage of a four-point bending simulation of a concrete beam. It is shown that the correlation length mainly affects the standard deviation but not the mean value of the outcome. Furthermore, despite the non-linearity of the problem, it can be shown that the probability box is described by the correlation length interval bounds only for this example.
KW - Aleatory and epistemic uncertainty
KW - Non-linear finite element method
KW - Probability bound analysis
KW - Random fields
KW - Uncertain correlation length
KW - Uncertainty quantification
UR - http://www.scopus.com/inward/record.url?scp=85089192323&partnerID=8YFLogxK
U2 - 10.3850/978-981-11-2724-3_0286-cd
DO - 10.3850/978-981-11-2724-3_0286-cd
M3 - Conference contribution
AN - SCOPUS:85089192323
SP - 2693
EP - 2698
BT - Proceedings of the 29th European Safety and Reliability Conference, ESREL 2019
A2 - Beer, Michael
A2 - Zio, Enrico
T2 - 29th European Safety and Reliability Conference, ESREL 2019
Y2 - 22 September 2019 through 26 September 2019
ER -