Reference designs

Let us consider the case of a ship owner who wishes to add a new cargo ship to his fleet. His prime interest is, of course, how much cargo (for which he gets paid) the ship can carry. He further specifies a minimum sustained speed and an action radius (distance the ship can sail without bunkering). How can the naval architect then generate a first design, i.e. establish the main dimensions of length, breadth, and draught of the ship?

The most important constraint in this case is the law of Archimedes: the weight of the ship, including cargo and fuel, equals the weight of the water displaced. This constraint is extremely difficult to cope with because the weight is the sum of a multitude of parts and components.

If the naval architect starts from scratch, i.e. an arbitrary first choice of main dimensions, he will invariably find that the 'design' does not satisfy the law of Archimedes. By variation of main dimensions he can finally arrive at a set of values which satisfies Archimedes' law. In each variation a complete weight calculation, i.e. establishing the sum of all weights involved, has to be done. This is an almost impossible and extremely cumbersome task, because most weight components are not known a priori.

The usual approach, therefore, is that the naval architect tries to find some existing ships which more or less satisfy the demands of the owner. He then studies these reference ships - or reference designs - and derives relevant ratios from their designs. For instance the ratio dead weight/displacement (dead weight = the weight of cargo and fuel). Once he has established this ratio for the class of ships concerned, he can apply it to his own design. From the owner's requirements he can establish the dead weight (the fuel can be estimated from the speed and action radius requirements). The dead weight/displacement ratio derived from the reference ships then provides a first estimate for the displacement. The block coefficient for the class of ships concerned, i.e. the displacement over the 'bounding box' around it, can be established in a similar way. Stability requirements determine the ratio breadth/draught, and the requirement of minimum wave resistance provides the ratio of speed over the square root of the length. From these ratios, the naval architect obtains his first set of main dimensions without carrying out any weight calculations. This procedure based on reference designs constitutes an enormous time-saver in the design process.

Similarly, the architect of a building can make use of reference projects by deriving certain ratios from them and applying these ratios to his particular design. For certain categories of buildings the latter has already been done for him. The resulting ratios are published in norm tables, for instance the REN: the Real Estate Norms. Such 'norms' are no more and no less than the result of an extensive regression analysis of a vast number of existing buildings. As in ship design, such empirical ratios are extremely useful for the efficiency of the design procedure. The architect should be aware, however, that these 'norms' are descriptive, not normative. They are no more than averages (weighted with the least squares method) of existing designs. As a corollary, the architect should not be afraid to deviate from these norms when his design problem so requires.

Equally important is to note that the architect can establish his own 'norms' from a limited number of reference designs by carrying out a regression analysis as described in Section 8.5. This is particularly recommendable when the design is unconventional and deviates substantially from the designs underlying the REN norms or other empirical ratios.

Example: Municipality office of The Hague*

RFPs (Request For Proposal) were directed towards selected construction firms for the municipality building of The Hague in The Netherlands. The design had been made by the American architect Meier. In the RFP, reference was made to several existing buildings designed by Meier, but the construction firm with the winning bid had not paid any particular attention to them.

When the construction firm later complained about excessive quality demands from Meier, the municipality could successfully refuse to pay any extra for this, because the contractor could have been aware of those demands if he had paid proper attention to

*F. Seyffert, private communication.

ratio cost price I size corrected for Meier effect ratio cost price I size corrected for Meier effect

Figure 6.4 Correction of empirical data by regression analysis of reference designs

the reference designs. What the contractor should have done is the following (described here in a simplified way):

• Establish the cost price from the ratio cost price/size according to the REN norms or the contractor's own empirical data (Fig. 6.4). Plot some existing 'Meier designs' in the Figure (indicated by crosses);

• Calculate a correction factor C on price for the 'Meier-effect' by a linear regression analysis of the reference designs: cost price 'Meier' = C x cost price according to REN norm (or own empirical data). Adjust the bidding price for the 'Meier-effect'.

By doing so, the contractor could have avoided his severe losses and a lot of fruitless dispute.

Best Investment Tips and Ideas

Best Investment Tips and Ideas

The dynamics of investing can be very emotional and stressful if not properly managed. When you are aware of what is all involved you give yourself the power to avoid those situations or at least manage them effectively. That will make your investments more exciting, rewarding, and enjoyable. Those positive factors will only lead to greater success in all that achieve with investments and life.

Get My Free Ebook

Post a comment