No model testing

Aug 2, 2007
Nowadays ship designers rarely are surprised in any way by the results of the speed and economy trials of a new ship. I've been involved in the development of several highly novel fast ships and have encountered some unpleasant surprises (engineering surprises are by definition unpleasant), but never regarding the power or fuel consumption versus speed; the resources for predicting ship resistance (i.e., drag) are adequate to almost any need.

It wasn't always that way. In the late 1860s the British Association for the Advancement of Science empanelled a group of experts under the chair of Charles W. Merrifield, F.R.S., to "report on the state of existing knowledge on the Stability, Propulsion and Seagoing Qualities of Ships, and as to the application it is desirable to make to Her Majesty's Government on these subjects." Their 1870 report, written by Merrifield, has an extended section which reviews the existing state of knowledge regarding ship resistance that's a marvel to read. It shows a great deal or knowledge and observation of the various phenomena involved, and virtually no understanding of how they do or do not contribute to resistance, nor any clear notion of how resistance might reliably be determined. It ends with a call to construct full-scale experimental ships of widely varying hullforms so their resistance may be measured.

One member of the panel, William Froude (1810-1879), submitted a dissenting view in which he cut through all the complexification to pick out the factors that truly mattered and proposed a clear and definite procedure for predicting resistance on the basis of tests of small (10' to 20') models. He included some rough test results that supported his theses.

On the strength of this the Admiralty declined the committee's proposals to build a myriad of experimental ships, but provided Froude with £2000 to build an experimental tank for towing models near his home in Torquay and sponsored a full-scale resistance test of an existing ship, HMS GREYHOUND, under Froude's supervision. The tests confirmed that Froude's methods could accurately predict full-scale resistance from model tests. The Torquay tank was dismantled in the early 1880s and Froude's son, Edmund, superintended the building and operation of its replacement at Haslar, opening in 1885. In the meantime, William Denny & Sons had built their own tank at the yard in Dumberton. The Haslar and Denny tanks were the only two in Britain for the next two decades, although tanks proliferated elsewhere. In 1906 John Brown opened a tank of their own at their Clydeside yard.

From at least the early 1900s there was considerable agitation amongst shipbuilders and naval architects for a national tank. The problem of course was money and it was not until Alfred F. Yarrow donated somewhere in the neighborhood of £20,000 that it was finally possible to build what became the William Froude National Tank, which opened at the then-new NPL facility in Teddington in 1911.

In all of this I cannot see any possibility of model testing of the OLYMPIC class. For the most part the Admiralty would not test merchant ships at Haslar; the only exceptions were ships of particular naval interest. The Denny and John Brown tanks were major competitive assets and would not have been rented out to a competitor. (H&W did test at Dumberton many years later, but that was under radically different competitive conditions.) And the Teddington tank did not open until after the design had been cast and OLYMPIC herself was in service. Thus it seems we can write off any possibility of tank testing.

Froude's methods permitted accurate predictions of speed vs. power based on trials data from previous ships of similar form (but different size and displacement). Nevertheless, tank testing usually more than paid for itself since it usually revealed how small changes could shave several percent off power requirements, thus permitting substantial savings on fuel and perhaps reduction in the number of boilers. Given the emphasis on economy of operation of the OLYMPICs, it's a bit surprising that H&W failed to pursue tank testing for them. In fact, given the business he was in, it's surprising that Pirrie never built his own tank facility.