China
Africa
Explore chapters and articles related to this topic
Resilience-based design of reinforced masonry wall buildings under blast loading
Published in Claudio Modena, F. da Porto, M.R. Valluzzi, Brick and Block Masonry, 2016
S. Salem, W. El-Dakhakhni, M. Tait
Furthermore, man-made hazards (e.g terrorism) are characterized by their high uncertainty due to the possibility of adapting to different protection techniques. Different design standards addressing the design of blast resistant structures are based on a Design Basis Threat (DBT) (ASCE, 2011; CSA, 2012). This in turn is based on the assertion that the intelligence community (IC) is one of the best sources of information about potential attacks. The IC observes, collects and assesses terrorist activities, motivations, intent, and capabilities while the Department of Homeland Security's (DHS) [or its equivalent in each country] analysts translate the data obtained from IC to DBT (Ezell & Bennett, 2010). Moreover, blast loading represents highly uncertain loading conditions. Such uncertainties can be classified into two categories: epistemic uncertainty (parameter and model uncertainty) and aleatory uncertainty (inherent variability) (Fig. 1) (Stewart et al. 2006).?
Estimation of virgin state of stress and determination of final rock stress model
Published in Katsuhiko Sugawara, Yuzo Obara, Akira Sato, Rock Stress, 2020
Hakami et al. (2002) have defined the uncertainty in stress prediction into two categories corresponding to two different parameters, an uncertainty parameter (u-parameter) and a variability parameter (v-parameter). The two parameters will give two “spans” for each prediction like ± u and ± v, respectively. A stress estimation with maximum horizontal stress, S(H) = 20 MPa (u = 20%, v = 10%) means an average stress in the span 16–24 MPa and a variability of 10% gives 14.4–17.6 for the lower value and 21.6–26.4 for the higher value. Hence, the estimated total span considering both the uncertainty and variability gives the final stress interval 14.4–26.4. A similar approach can be applied to the BERSM and other stress models presented in this contribution.
Efficient numerical methods for the optimisation of large kinetic reaction mechanisms
Published in Combustion Theory and Modelling, 2022
Simret Kidane Goitom, Máté Papp, Márton Kovács, Tibor Nagy, István Gy. Zsély, Tamás Turányi, László Pál
Parameter vector P also contains optionally transformed parameters whose uncertainty after transformation can be characterised by a simple, symmetric distribution and symmetric uncertainty limits. For combustion kinetic problems, the main model parameters are the rate coefficients (), which, however, are temperature dependent, thus so is their uncertainty range (, ), which is usually taken symmetrically around the nominal value on logarithmic scale, therefore, its uncertainty parameter is defined [101] as This transformation implies the optimal transformations for the Arrhenius parameters, which parameterise the temperature dependence of with the extended Arrhenius equation, which takes the following form after logarithmic transformation: Here, we introduced transformed Arrhenius parameters (α, n, ε), which are used for composing parameter vector P. The prior joint uncertainty of the (α, n, ε) transformed parameter triplet of a reaction is indirectly defined by the temperature-dependent uncertainty parameter . It is usually assumed [102] that is proportional to , with a proportionality factor of 3: Regardless of whether the uncertainty limits have a probabilistic meaning or not, it was shown in [103] that it could be efficiently approximated and stored as a function of the covariance matrix of the Arrhenius parameters using the following equation: The covariance matrix of transformed Arrhenius parameters is defined as follows: Here , and are the standard deviations of the transformed Arrhenius parameters, and , , and are the correlation coefficients. The prior uncertainty parameter () with its definition temperature range () and the corresponding prior covariance matrix can be determined based on all direct measurements and theoretical determinations available in the literature according to the theory and protocol presented in Refs. [100,104].