A method for the probabilistic modelling of ice pressure

- A probabilistic method for determining global ice pressure percentiles is presented.
- Lognormal distribution for ice pressure and regime parameters is often plausible.
- A positive correlation of the ice thickness and strength can lead to a decrease of design pressure.
- Powers in ice pressure design formula influence pressure moments and probabilities.

This paper presents a probabilistic model for global ice pressure based on empirical formulas. In recent studies, the Monte Carlo numerical simulation has become an important instrument for analysing information about ice pressure distributions in ice/structure interactions. The aim of this paper was to develop a method for obtaining theoretical probabilities of critical ice pressure values as well as other probabilistic characteristics of random ice pressure.The method presented in this paper provides an inexpensive and fast way to make estimations for ice pressure probabilistic parameters that can be used when the equation for pressure involves the product of parameters taken to fixed powers and the distributions for the parameters are either lognormal or uniform. An overview is presented of papers that show example cases where the use of lognormal or uniform distributions is appropriate.An example of the application of the developed model is provided. The relation of powers in the ice pressure formula recommended by ISO and a mathematical estimation of the resulting pressure are investigated. The results are discussed.