All news Events

Probabilistic damage stability: enhanced algorithm for the determination of damage case dimensions

Posted on September 2, 2016

PIAS’ Probdam module is equipped with functions for damage case generation and for the automatic determination of damage boundaries. The algorithm for the latter has been enhanced for two reasons:

  • To accommodate damages over centerline (as already announced in the newsletter of October 2015).
  • To be able to handle non-foursided compartment shapes. With PIAS’ old compartment definition module Compart all subcompartments were limited to four sides. Newlay, the replacement module of Compart, supports a wider variety of shapes, including subcompartments having from three to twelve sides. Because all damage stability modules are being prepared to support this extended definition, the Probdam damage case management library must be updated accordingly.

 

The update of this algorithm has three consequences:

  • The found extent of damage can be slightly different than with the previous algorithm. By the way, the manual contains a short elucidation on the process of finding the damage boundaries.
  • Not each and every damage case is indeed possible, given the rules and constraints of SOLAS2009. This new algorithm has been further optimized to find matching damage boundaries, where with the previous algorithm it could have been concluded that the damage was impossible.
  • Although finding the extent of damage may be a time-consuming process, the old algorithm was highly optimized to do it as swift as possible. The new algorithm has to evaluate more complex cases – over CL, and non-foursided – which inevitably will result in occasionally longer processing times. However, please be assured that SARC has used every option to keep the performance high. In this respect contemporary computer hardware facilities can also be employed, with PIAS’ octothreading speed enhancer, which is scheduled to be released within some weeks.

A version of PIAS’ Probdam with this enhanced algorithm will be available for download from September 6, 2016.