TY - JOUR
T1 - Sensitivity analysis and probability modelling of the structural response of a single-layer reticulated dome subjected to an external blast loading
AU - Qi, Shao bo
AU - Huang, Guang yan
AU - Zhi, Xu dong
AU - Fan, Feng
N1 - Publisher Copyright:
© 2023 China Ordnance Society
PY - 2023/5
Y1 - 2023/5
N2 - The structural response of a single-layer reticulated dome to external explosions is shaped by many variables, and the associated uncertainties imply non-deterministic results. Existing deterministic methods for predicting the consequences of specific explosions do not account for these uncertainties. Therefore, the impact of the uncertainties associated with these input variables on the structures’ response needs to be studied and quantified. In this study, a parametric uncertainty analysis was conducted first. Then, local and global sensitivity analyses were carried out to identify the drivers of the structural dynamic response. A probabilistic structural response model was established based on sensitive variables and a reasonable sample size. Furthermore, some deterministic empirical methods for explosion-resistance design, including the plane blast load model of CONWEP, the curved blast load model under the 50% assurance level, and the 20% mass-increased method, were used for evaluating their reliability. The results of the analyses revealed that the structural response of a single-layer reticulated dome to an external blast loading is lognormally distributed. Evidently, the MB0.5 method based on the curved reflector load model yielded results with a relatively stable assurance rate and reliability, but CONWEP did not; thus, the 1.2MB0.5 method can be used for making high-confidence simple predictions. In addition, the results indicated that the structural response is very sensitive to the explosion parameters. Based on these results, it is suggested that for explosion proofing, setting up a defensive barrier is more effective than structural strengthening.
AB - The structural response of a single-layer reticulated dome to external explosions is shaped by many variables, and the associated uncertainties imply non-deterministic results. Existing deterministic methods for predicting the consequences of specific explosions do not account for these uncertainties. Therefore, the impact of the uncertainties associated with these input variables on the structures’ response needs to be studied and quantified. In this study, a parametric uncertainty analysis was conducted first. Then, local and global sensitivity analyses were carried out to identify the drivers of the structural dynamic response. A probabilistic structural response model was established based on sensitive variables and a reasonable sample size. Furthermore, some deterministic empirical methods for explosion-resistance design, including the plane blast load model of CONWEP, the curved blast load model under the 50% assurance level, and the 20% mass-increased method, were used for evaluating their reliability. The results of the analyses revealed that the structural response of a single-layer reticulated dome to an external blast loading is lognormally distributed. Evidently, the MB0.5 method based on the curved reflector load model yielded results with a relatively stable assurance rate and reliability, but CONWEP did not; thus, the 1.2MB0.5 method can be used for making high-confidence simple predictions. In addition, the results indicated that the structural response is very sensitive to the explosion parameters. Based on these results, it is suggested that for explosion proofing, setting up a defensive barrier is more effective than structural strengthening.
KW - Blast load uncertainty
KW - Dome structure
KW - Probabilistic investigation
KW - Reliability evaluation
KW - Sensitivity analysis
UR - http://www.scopus.com/inward/record.url?scp=85125443348&partnerID=8YFLogxK
U2 - 10.1016/j.dt.2022.02.004
DO - 10.1016/j.dt.2022.02.004
M3 - Article
AN - SCOPUS:85125443348
SN - 2096-3459
VL - 23
SP - 152
EP - 163
JO - Defence Technology
JF - Defence Technology
ER -