A Comparative Study on the Seamlessness Between the Parametric Based Design Programmes SubParm and SubDes Compared to Build Submarine
Publish date: 2014-05-26
Report number: FOI-R--3887--SE
Pages: 84
Written in: English
Keywords:
- Functional Volumes
- Design Component
- Parametrical relations
- Seamless design
- Design variables
- Parameter sets
- Play-Cards
- Concepts
- Measure of Performance (MoP)
- Early knowledge and performance growth
- Design process
- Mission analysis
- Requirements
- Needs
Abstract
This report introduces a new approach to submarine design in the initial design phase as a tool to design requirements in the functions domain. Earlier Swedish parametric study programs were used in the initial design phases to explore the design space by varying design parameters to find a suitable design room from which a more refined design could be developed following the classic design spiral. These parametric study programs were simple and could not easily be adapted to changed conditions. Their output had systematic errors yielding too small size estimates. In subsequent design phases major adjustments were needed in order to get a submarine with correct size and balance. As a result this parametric program approach was abandoned in later submarine projects in favour of the standard "type ship" evolutionary design method. Contemporary packing algorithms, even if they initially use a functional breakdown structure, are used in the system domain to design concepts automatically. This approach puts great emphasis on very precise system data and requires large amounts of data in constant need of review and maintenance. As each and every component of importance has to be incorporated in the packing and there are different packing ratios in different sections and rooms, rigorous rules are needed to ensure that no components overlap and that appropriate service space and shock clearances are considered in the packing algorithm. Here functional volumes are introduced to represent rule-based and scalable volumes which include service space, shock clearance and general volumes. The size (and also the weight, cost, power consumption etc.) of a functional volume is determined from selected system principles and stated performances by parametric relations expressed in scripts, thus easy to change and refine. The functional volumes are packed into relevant rooms in a general arrangement, without a predefined layout but with a chosen style. In this fashion, it is possible to generate functional design representations of submarines, called Play-Cards, which are feasible, balanced and detailed enough to be useful for evaluation of initial design requirements. In the subsequent concept design phase, a few Play-Cards are selected for detailed arrangement of systems and installations. This approach enables the designer to rapidly explore vast regions of the design space without undue constraints from preconceived solutions. This open study, exemplified by four Swedish submarine designs with different styles, also shows that there exists a seamless transition from the functions domain with Play- Cards over to the installations domain via the systems domain with concepts, when using the SubAN toolbox. A German submarine design was also studied that initially gave a significant larger Play-Card. This was due to differences in design philosophy, packing and the more complex dual use of rooms. The study shows that, a) it is possible to design representations of single hull submarines, i.e. Play-Cards, in the functions domain with the use of the functional volume approach, b) when reengineering already built submarines of different styles, both in the functions domain and in the systems and installations domains, no significant errors, deficiencies, or change of characteristics occur during domain transition as long as the program package can accommodate the relevant design philosophies and related packing ratios in a coherent way. The ongoing development and validation for semi-double hull and double hull design representations of submarines indicate the same results. This result shows that there exists seamlessness and that the domain transition has sufficient precision and fidelity given that not only numerical data but also geometrical and topological relations are considered when modelling the functional volumes. In this study, the principal approach for designing a knowledge database in the functions domain is shown. Data has been deduced and normalised from several submarine designs since early 1950.