1.
Introduction
The 'Age of Sail' (mid-sixteenth to mid-nineteenth century) saw a dramatic change in ship design, navigational practices, and technology in Europe. This rapidly increased the ability of European nations to explore, compete, and trade globally. Crucial to this was the development of new designs of ship hulls. Competition between European naval powers, such as the Dutch, British, and French Navies, created pressure on shipwrights to continually innovate so as to outperform their competition (Dodds / Moore 2022; Winfield 2017). This was a time of industrial espionage, copying, and updating, and a moment of tension between traditional ways of shipbuilding and new approaches (Ferreiro / McGee 2006). With almost no ships surviving from the period, it is difficult for historians to directly assess the performance of these vessels, how they compared to one another, and the innovation successive models introduced.
This paper outlines a novel approach which is currently being developed to study the sailing performance of Dutch and British ships during the long eighteenth century. Building in part on existing approaches for ship modelling and at sea performance, it introduces a technique that combines historical records with digital capture and engineering simulations. This consists of 3D capturing a representative subset of surviving ship models, built to scale, and using FEMs (finite element methods) and hydrodynamics analysis to assess their design performance.
2.
Data Collection
During the eighteenth century, British ships are thought to have radically outperformed Dutch ones. This difference has been attributed to the hull design constraints imposed by the shallow waters which Dutch ships had to navigate (Solar and de Zwart 2017). French ships, on the other hand, were often considered to be faster by British observers, who tried to copy the shape of prizes captured in war (Pritchard 1987: 2-3). These positions, however, are drawn from ship logs and printed accounts, and the careful discarding of other hypotheses, and maintain the flair of conjecture. Their testing is the basis of the study attempted here.
Although few ships from this period survive, an extensive body of wooden ship models, built to scale, has been preserved. These can be seen as almost accurate representations of the historical ships that were produced during the eighteenth and early nineteenth century and followed the same general construction principles of full-scale vessels (Gardiner 2013). In the case of Great Britain, for example, the Royal Navy required every model of a ship produced after 1716 to be preserved, many of which are now held in the
National Maritime Museum in Greenwich. Similarly, in the Netherlands, the
Maritime Museum in Rotterdam contains an equally sizeable collection.
A curated test of a specific design, 3D scanned in-situ, constitutes the basis of this pilot study. The general models used are limited to British and Dutch military vessels from the early and late eighteenth century selected for sharing similar maritime tasks. The models, captured via photogrammetry or laser scan, have varying degrees of accuracy that require some form of output standardisation to enable comparability.
3.
Method
The method to compare models is structured as a “pipeline” from 3D model to performance score (cf. Rudan et al. 2023; Indruszewski et al. 2004).
1
A digital model of the ship
(Figure 1.A.), once scanned, can be reduced to a 3D cloud of points.
2
From this, it is possible to create a simplified closed version of the ship hull, irrespective of the presence of ribs, empty space, or artifices on the ship model. This is achieved by means of an alpha shape closed over the cloud of points representation of the ship (Edelsbrunner et al. 1983). Additional cleaning is produced in
Blender if necessary. These steps can guide the creation of a 3D hull which maintains watertightness (i.e., clear in and out surfaces) and smoothness (no noise and artefacts) across scans (Figure 1.B.).
Once the watertight and smooth hull is modelled, it is inputted into a fluid dynamics software (such as MaxSurf and Orca 3D). The model is set at a specific, historically consistent, water level to account for its actual behaviour. The use of a standardised model in combination with a computational simulation of its water behaviour produces an estimate of the drag factor of the vessel, used in this project as its performance measure. Other measures of efficiency are possible and indeed necessary, but at this stage a well known performance measure allows for an agile comparison across models and time.
4.
Outcomes and Significance
As a pilot study, this paper satisfies two aims. Firstly, it provides an illustration of how questions of historical efficiency can be answered using a combination of historical and engineering techniques. In particular, the use of links between design and efficiency opens interesting opportunities for the study of technological innovation. Secondly, this study develops a technical methodology to effectively do so, which can be scaled to a larger corpus. This creates a direct connection between theory and practice in the study of digital material culture.
By combining 3D capture of historical ship models and historical research with engineering techniques of fluid dynamics simulation, this study opens new methodologies for the study of the past. At a higher level of abstraction, the procedure devised here can be extended to the digitisation and comparison of large collections of historical artefacts with important implications for the study of technological innovation across time and cultural boundaries.
