The aim of this paper is to discuss the challenges associated with the early-stage design of novel and reliable vessels, and discuss the expected benefit of the application of multi-fidelity models in this specific problem. During early-stage design of novel or complex vessels, the naval architect aims to explore the multi-dimensional design space to identify the most promising solutions and identify key trade-offs. The goal is to understand the relationship between the design drivers and the key performance indicators (KPIs) which will lead to the formulation of the main design decisions. This analysis is typically performed with the help of bespoke early-stage design tools. Traditionally, early-stage design tools are computationally cheap, but lack in accuracy. These tools have been developed based on gained design knowledge such as empirical and semi-empirical methods, historical data, and experts’ opinions. However, for the design of novel vessels, these tools are not sufficient because they have typically been developed based on the experience gained from traditional vessels. The early-stage design of novel vessels is challenging because there is very limited existing knowledge about both the design drivers and the KPIs due to their inherent nature of being “novel”. The first part of the paper discusses the challenges associated with the design of novel vessels. Two novel hull designs were examined as qualitative case studies, namely the USS Zumwalt (DDG1000) and Littoral Combat Ship (LCS) – Independent variant, to identify the design drivers and KPIs connected with the design of novel vessels by following a bottom-up approach. Ensuring the vessel’s reliability is a major design aspect for all novel designs. To achieve this, analysis via high fidelity tools is traditionally required. Typically, high fidelity tools are used later on in the design process due to their high computational requirements and the detailed design information required. However, the application of multi-fidelity models (MFMs) in the design process is a promising technique in order to introduce high fidelity models earlier on in the design process. The second part of the paper focuses on an extended literature review on the application of the multi-fidelity models to the design of complex engineering systems. Finally, the most promising methods for this specific problem are identified and discussed.

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