Author(s): Dmitry A. Onishchenko; Sergey I. Shibakin
Linked Author(s): Dmitry Onishchenko
Keywords: No Keywords
Abstract: The paper deals with the question on the accuracy and reliability of ice feature thickness and ice load design values which are statistically derived from the field observation data covering short time periods, e. g. a few years in duration. A procedure of probability modelling is suggested which make it possible to find analytically a) the probability that the sampling design value of ice thickness is less than the 'true' value by quantity δ, and b) the probability that the load generated by an ice feature will exceed the calculated sampling value by ε% at least once during the nearest L years. As an example, model calculations are performed for the case of the Korzhavin type equation for sheet ice action. The calculation results for the case of five-year survey show that the reliability of the design values (for L = 50) obtained using the corresponding field data might be insufficient for such critical constructions as sea platforms in ice conditions.
Year: 2004