The goal of the PhD activity was to conceive an innovative design process, easing the design of more efficient, robust, flexible, secure, and performing IPSs. At the same time, such process had to be able to limit the cost increase due to modifications a posteriori on the system, commonly caused by failures in requirements compliance found after vessels construction.
In Chapter 2.2.3 (“Ship design methodologies”, at page 33), the conventional design process (spiral design) has been described, together with other innovative methodologies conceived to optimize the ship design. The advantages that such methodologies may give to ships' design are undeniable, therefore their adoption is highly recommended. Actually, the shift towards collaborative concurrent design is already in progress, while design space exploration is still far to be applied (however, some applications to define naval vessels' concept design are done, mainly by US Navy). Due to that, it was deemed pointless defining a completely new process for ship design. However, the advantages that the aforementioned new tools are able to give to the design are clear; therefore, the decision has been to develop a sub-process that integrates such tools. This decision has been taken in order to make the innovative design process as general as possible: whatever the chosen design process will be, it will be possible to integrate into it the proposed design process as a sub-process. This allows achieving the pros given by
the use of both dependability theory concepts (fault events evaluation, objective comparison between designs through dependability attributes, etc.) and the possibility given by the software models for the dynamic simulation of electrical systems (evaluation of electromechanical transients after faults, reconfiguration procedures tests, etc.).
The new design process can be modeled with a circular structure (Figure 40) to be inserted within the main design process (of the IPS design, in its described embodiment). Several steps compose it, chosen in order to allow achieving significant advantages from the abovementioned new tools. The concepts on which the process is based are the following:
• Application of techniques given by dependability theory in order to assess which are the most frequent causes of a given top-event (single subsystems and elements faults) and what are the subsystems/elements that have the most impact on the given top-event occurrence (thus defining the most relevant changes in system's design and/or in components' reliability from a dependability point of view);
• Analysis of the system through time domain simulation (steady state and electromechanical transients), in order to obtain data on the dynamic evolution of the system needed to correctly define solutions to the issues highlighted through the dependability analysis, and to verify the correctness of the system design (in respect to regulations/owners requirements);
• Evaluation of the achieved improvements, using dependability theory techniques to assess if the solution is worth the adoption (or not) in terms of dependability indexes.
Figure 40 - Innovative design process, subroutine to be integrated into IPS's design
Dependabilityevaluation of the system
Dynamic simulationof the system in fault conditions
- Evaluation of electro-mechanic variables evolution during fault transients.
Proposalof solutions
- Modification of the design to solve issues highlighted in previous steps.
Figure 41 - Integration of innovative design process in IPS design
The circular process of Figure 40 has to be inserted in a particular point of the IPS design process, due to the information needed to apply the tools. Indeed, the design of the IPS has to be defined together with its main components in order to apply both dependability techniques and software simulations. Due to that, referring to Figure 12 (see page 39), the ideal moment in which applying the new process is right after Plant Configuration step. In this way, the information needed to apply the process are available, and the process can be used to modify the design before additional time-consuming steps are made. Nevertheless, some of the steps that follows Plant Configuration in IPS design process may be relevant for the new design process (mainly Cost, Ship Fit/Impact, Static System Analysis, and Power Quality).
In fact, the new design process allows defining solutions to issues emerging from the dependability study of the system, and choosing between them following a dependability based metric. However, in ship’s design also other constraints apply, such as costs and space/volumes issues. Due to that, a smart solution may be to perform Cost and Ship Fit/Impact analyses on each solution resulting from the new design process. The results of such analyses will allow selecting the most fitting solution considering all the impacting variables.
The impact in terms of human and time resources of these additional analyses has to be limited, in order to avoid increasing the design effort. Due to that, a higher grade of approximation in respect to main design process can be applied to evaluate the possible solutions, and detailed analysis can be made only on the chosen design. The resulting process may be something similar to what depicted in Figure 41, where the new design process is used to find issues in the design and propose solutions. These solutions can be then evaluated, and the best compromise can be chosen to be applied in the final ship design.
The proposed design process needs detailed information on the system to be able to give all its advantages, thus being ideally applicable from functional design onwards. However, its application it is not limited to such an advanced phase of the design. Indeed, preliminary design can greatly benefit from such a process, being possible to address main system issues in a phase in which high impact solutions can be implemented without excessive modification effort. Even though the information about the system is scarce in preliminary design phase, it may be sufficient to perform an approximated dependability analysis and to perform some simulations of system’s dynamic behaviour based on common components’ data. Such
analyses will be able to identify main design flaws, and guide the designers in the choice of the system architecture most suitable for the application.
The proposed approach seems to lead to an increase in design complexity due to both the additional design steps and techniques, but, in fact, the result is a simplification in the design.
Indeed, such a process allows to design relying on methodologies useful to systematically define: critical points of the system, redundant elements to be inserted or removed, best solutions to get the required response to fault events, and also to objectively demonstrate the quality of the design to all those concerned. Doing that, it is possible to avoid the "trial and error" procedure commonly used to find solutions in case of unforeseen issues, which is the most impacting activity to do during design in term of human, time, and financial resources.
Conversely, in an already well proven design devoting some effort to apply such an approach may lead to advantageous results, due to the possible design optimizations that can be deduced (thus increasing performance, decrease weights/volumes, decrease costs, and improve dependability attributes). In fact, it is demonstrable that the integration of dependable design and software simulation in such processes can be obtained with bearable effort, due to the possibility of using an already present ship design substrate. In particular, in [97] it is shown how dependability techniques can be used as a project management tool in ship design, and how these techniques can be integrated into present design tools in the least impacting way. Considering the most demanding AES application in merchant area, which are DP vessels, proposals to consider the adoption on dependability techniques during throughout all the design process are available in literature, such as in [96]. For what concerns dynamic simulations aid to system design, it has to be pointed out that such an application is already in study in most advanced IPS’s components suppliers, such as ABB [104]. Moreover, as previously affirmed the new IEC 61892-5 [99] will oblige to perform HIL testing on the main shipboard control systems to ensure they are suitable for the purpose.