The paper considers the loads measured in indentation tests with cylindrical indenters 40 mm in diameter at 1 mm/s velocity. The common linear-like part of load growth within first 1.75 sec of the contact is revealed. An alternative autoregressive model of unit root for the load in this part is applied. The paper shows that adjustment of autoregressive model with Gaussian residuals may be used for localizing the beginning of ice-indenter interaction in raw ice load time series.


Ice load recordings made in model and in-situ experiments usually obtained as time series consisting of equidistant observations. Despite the fact that data are given as discrete values observed with some sampling rate, in analytical models an ice loads are usually interpreted as a continuous process of the stochastic nature. This approach was practiced in the early studies (Sundararajan and Reddy, 1973; Reddy et al., 1977), as well as was applied in the modern ones (Karna et al., 2007; Guo, 2012; Neumann et al., 2015; Hendrikse and Nord, 2019).

Processes of loads caused by ice on a structure are not studied well enough up to now. This is due to the general complexity of ice-structure interaction and difficulties in observing ice breaking and crushing processes. In-situ experiments are mostly focused on obtaining local loads. The corresponding measuring equipment can be positioned from the outside of a structure in the form of load panels, as well as from the inside in the form of strain gauges. For instance, the experimental setup established on Norstromsgrund lighthouse provided numerous measurements of local loads made by special measuring panels (Nord et al., 2016).

Global ice loads on structures are difficult to measure or estimate (Loset et al., 2010). Thus, global loads are usually considered in the context of model-scale experiments in an ice tank. There, experiments with inverse scheme of motion are usually performed: a model of a structure is being towed by a trolley, and the model ice sheet stands still. Dimensions of a model usually allow setting up a dynamometer in the place of its mounting to a trolley, which allows measuring global load caused by model ice on a model of a structure.

This content is only available via PDF.
You can access this article if you purchase or spend a download.