BEGIN:VCALENDAR VERSION:2.0 PRODID:-//Talks.cam//talks.cam.ac.uk// X-WR-CALNAME:Talks.cam BEGIN:VEVENT SUMMARY:Design and Assessment Methods for Ships in Ice - Claude Daley (Mem orial University of Newfoundland) DTSTART:20171107T100000Z DTEND:20171107T110000Z UID:TALK94570@talks.cam.ac.uk CONTACT:INI IT DESCRIPTION:Design and assessment of ships in ice is a topic that concerns a wide range of people\, from owners\, to designers\, builders\, regulato rs and the insurance industry\, with the public\, the environment and the economy being impacted by arctic shipping risks. The methods that are bein g used to design and evaluate ice class ships have evolved over many decad es and are currently continuing to change. In this discussion\, the focus is on the steel structure of ice-going ships\, including structurally impo rtant ice loads and the nature of the structural response. \;

As we seek to better understand the nature of the interaction between ship s and ice\, we are increasingly focusing on complex multi-body interaction s and non-linear behaviours. In such situations\, general solutions are no t only not currently available\, but may not be achievable at all. While w e can assemble specific solutions\, we must ask whether a general understa nding of complex nonlinear systems is possible\, and if so\, what mathemat ics can we use to develop them?

Concern for structural ice loads ha s\, until now\, mainly focused on a single situation\; the &ldquo\;design condition&rdquo\;. For that condition\, the structure behaves linearly or pseudo-linearly. The &ldquo\;design&rdquo\; loads have been largely develo ped empirically\, from a combination of measurement data and loads inferre d from past successful practice.

While much of the current industri al and regulatory practice still employs relatively simple models of load and response\, new sets of tools and approaches are taking shape and are b eing applied by the most sophisticated ship owners\, builders and operator s. These new approaches seek to directly model a growing set of complex is sues and scenarios\, with consequent improvements in the fidelity of ice l oads and structural response. The growing computational power and improvin g simulation tools are enabling these developments. To illustrate these ap proaches\, some of the author&rsquo\;s own efforts will be described.

One recent and developing technique is called safe speed analysis. Thi s method involves modeling a wide variety of discrete ship-ice interaction cases\, modelling the load and overload capacity of the vessel and then c ombining these results to produce a multi-parameter map of acceptable oper ations (speed\, ice size\, etc). The method makes use of available tools a nd solutions\, including &ldquo\;Popov&rdquo\; type collision models (alge braic solutions of two-body collisions)\, and explicit dynamic finite elem ent models (LS-Dyna) to capture interaction kinematics\, contact and a ful l range of structural responses (ductility\, dynamics\, stability).

A second\, but related technique is a model called GEM\, which uses simpl e event solutions and simple equations of motion to model large scale oper ations of vessels in ice. GEM\, while seeking a reasonable level of accura cy\, focuses on high simulation speeds and practical decision support rath er than on fidelity and universality.

After presenting these two me thods\, the presentation raises three mathematical challenges. The first c hallenge relates to the problem of interacting chaotic systems and wonders whether a calculus for such systems is possible. The second challenge que stions the application of probabilistic design to ice class ships. The thi rd challenge relates to the simulation of multi-body systems. Such systems are highly nonlinear and computationally costly to model. Could an asynch ronous or quasi-synchronous timestep algorithm yield improvements in overa ll speed? LOCATION:Seminar Room 1\, Newton Institute END:VEVENT END:VCALENDAR