Trans RINA, Vol 154, Part C1, Intl J Marine Design, Jan - Jun 2012 irrespective of vessel length). The scatter is due to
changes in the other design parameters, with the variation about the mean giving some indication of the scope for possible optimisation of these parameters. The greater the scatter, the more influence the other parameters have on the performance.
4. PRACTICAL APPLICATION OF DESIGN- SPACE EXPLORATION RESULTS
Several typical scenarios where the design-space
knowledge, captured in the response surfaces, can help the design team have been identified. These are discussed in more detail below.
4.1 ISO-PARAMETRIC VARIATION
Iso-parametric lines and surfaces are useful for answering questions such as: “How is the required delivered power affected if I change the vessel length whilst keeping the other parameters constant?”. Because there is an infinite number of combinations of values of the design parameters that are kept fixed, it is normal to choose some sensible baseline design from which to make the single (or twin) parameter variations.
The sensitivity of performance measures to certain design parameters can be gauged by looking at the shape of the iso-parametric surfaces. For instance, for the example application, it can be seen in Figure 7, that the required delivered power is much more sensitive to length than it is to beam. Thus there is less of a trade-off in power if the beam needs to be changed (to improve another aspect, for instance) than to change the vessel length.
simple interpolation is required to obtain the performance measures (instead of direct analysis with computationally intensive simulation tools). Some typical optimisation scenarios are described below.
4.2 (a) Design Genesis
When performing an initial search for viable design candidates which will be possible solutions to the design brief, all design parameters are allowed to vary within the range covered in the design-space. Since multiple performance measures have been calculated, a multi- objective optimisation may be appropriate, leading to a Pareto-front
of optimum design alternatives with
different compromises between each of the performance objectives. By examining the Pareto-front solutions, a feel can be gained for the trade-offs that might be required to improve a preferred performance measure. This can help the designer answer questions such as “What will happen to the stability and sea-keeping comfort if I would like to reduce the vessel resistance?”
Within the FFW, it is straightforward to perform separate optimisations for each of the individual performance measures, starting from the same baseline vessel. Figure 8 shows how the design parameters differ when the vessel is optimised for the different performance measures. Five vessels are shown, the baseline vessel and then a vessel optimised for each of the individual performance measures of interest. The points of the star are the normalised values of each of the design parameters. It can be seen that the vessel optimised for stability is quite different
from the other vessels: a
shorter, beamier vessel with high midship area is favoured. (Referring back to Figure 6, it can be seen from the initial design-space investigation, that shorter vessels
generally have better stability than longer
vessels.) The vessels optimised for resistance and sea- keeping are longer and narrower (as might be expected). These results can be used to answer questions such as: “How does the design for minimum resistance compare to that which offers best comfort in a seaway?”.
Figure 7: Response surface for Power vs. Length and Beam.
4.2 OPTIMISATION
Automated optimisation has been well documented over recent years (see COMPIT conferences for example:
www.compit.info). The data obtained from the design- space exploration and captured in the response surfaces is ideally suited to such optimisation, because only a
Figure 8: Normalised Optimum Design Parameter Values for Different Performance Measures.
©2012: The Royal Institution of Naval Architects
C-23
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64