Grace's Guide To British Industrial History

Registered UK Charity (No. 115342)

Grace's Guide is the leading source of historical information on industry and manufacturing in Britain. This web publication contains 163,113 pages of information and 245,598 images on early companies, their products and the people who designed and built them.

Grace's Guide is the leading source of historical information on industry and manufacturing in Britain. This web publication contains 147,919 pages of information and 233,587 images on early companies, their products and the people who designed and built them.

Denny Ship Model Experiment Tank

From Graces Guide
JD 2022 06 Denny 3.jpg
Cast iron moulding tank which was lined with clay, shaped with the assistance of templates to form a mould into which wax would be poured after inserting a wooden core
Wax model shaping machine to William Froude's patent, made by John McDowall and Sons in 1884
The high speed milling cutters on the shaping machine
Drawing of hull contours attached to shaping machine, with tracer on far right hand side

Castle Street, Dumbarton, Dumbartonshire, G82 1QS.

The Museum

William Denny's ship model experimental tank facility has been preserved as the focus of the Dumbarton branch of the Scottish Maritime Museum (the other, much larger branch is at Irvine). It is a fascinating museum. Much of the equipment is original, dating from the 1880s, and provides a fitting tribute to the foresight of William Denny and to William Froude's ingenuity and meticulous scientific approach.

Much of the equipment present when the yard closed in 1963 has been preserved, including the main tank and its building, the wax-melting equipment, the moulding tank, the Froude-designed milling machine used to produce the profile of the hull models from contour drawings (made by John McDowall and Sons in 1884), the Drawing Office and draughtsmen's instruments, and even a graph paper printing machine. There is also a reconstruction of the manager's office, and a variety of displays relating the Denny's shipbuilding activities. Staff are on hand to explain the processes of producing the large wax hull models.

On display outside the museum is the first marine engine designed by Robert Napier, constructed in 1821.


The first such testing tank was designed by William Froude and commissioned in 1872. Froude's work provided the basis for the Dumbarton tank, which was completed in 1883/4 as the first to be constructed by a shipbuilder - the first commercial testing tank[1]. It was designed with the assistance of William Froude's son, Dr. Robert Edmund Froude, who was Superintendent of the Experimental Works at Torquay and later at Haslar. The tanks were built to enable designers to carry out hydrodynamic testing of scale wax models of ships' hulls to investigate the performance of hulls and propulsion systems. There was a long-felt need for information to assist designers with hull design and with the optimisation of propulsion systems.

Note: In the 1870s Bruno Joannes Tideman of the Royal Netherlands Navy had established a modest testing tank in Amsterdam, with some guidance from Froude[2]

William Denny had first visted Froude's Torquay tank in 1873. Construction of another tank required the permission of the Admiralty, and this was given to Denny in 1881, who also instructed R. E. Froude to facilitate the work at Dumbarton. Denny had earlier appointed Frank Purvis, a former assistant to William Froude, to establish a scientific department at the Leven Yard, and he was given the task of setting up the testing tank at Dumbarton. Purvis recruited another of William Froude's former assistants, Edwin Richard Mumford to the staff. Mumford became Superintendent in 1883, and remained in the post until his death 43 years later.[3]

The tank was 300 feet long, later increased to 350 feet.[4]. The first tests here were probably carried out in late 1883.

The Original Installation

The following information is partly condensed from an article by E. R. Mumford in The Engineer in February 1884[5]

Mumford's article hints at many of the challenges which had to be addressed by Froude to make the equipment viable. His objectives were far-sighted, the means he selected were well-founded, and his ingenuity and attention to detail were remarkable.

Froude recognised the need for numerous model hulls, with a requirement to produce them hulls accurately, quickly and cheaply. His solution was ingenious. He selected paraffin wax as the most suitable material. It has no grain, and is very easy to machine, requiring little power and involving minimal tool wear. It melts at about 54 degC, making it easy to cast a model hull approximately to the correct form. It does not absorb water and thus swell and become heavy, as might arise with wooden models. There is little waste of material, as the model can be broken up and remelted. A large iron trough is used to contain a female clay mould roughly to the size and shape of the model (see photo), and a wooden core is prepared. The models at Torquay and Dumbarton were usually from 3/4in. to l 1/4 in. thick, and from 12ft. to 20ft. long. The core is secured to prevent it rising when the liquid wax is run into the mould. Note: It would seem that the core itself would need to be accurately made in order to obtain the stated wall thickness.

'To prevent the core being crushed by the pressure due to the head of liquid wax, it is filled with water sufficiently to balance the outside pressure. When the model is cold the core is removed, and marks are put on to guide in the after process of cutting and shaping. The model is floated out of the mould by water pressure, and placed bottom up on the bed of the shaping machine.

'The shaping machine has a pair of revolving cutters which can cut out any water line on the model by using a tracer, connected to the cutters by a system of levers, to trace out the corresponding line on a suitable drawing, the drawing being fixed to a board which is geared to the bed of the machine. Machining leaves a series of terraces, which are faired by hand. The model is then accurately weighed and placed in the water, and the difference between the required displacement and the weight of the model is made up partly by the wooden frame or harness wh1ch must be fixed on the beams, but principally by canvas bags filled with small lead shot and by lead weights. The model was towed by means of a carriage drawn along a straight and level railway, 18in. above the water level, and any speeds between 50 ft. per minute and 1000 ft. per minute using a stationary steam engine, winding a steel wire rope attached to the carriage.'

The stated maximum speed of 1000 ft/min seems rather high, given that the tank was only 250 ft long!

Without illustrations, Mumford's descriptions of the towing carriage, the dynamometer used to measure the towing force, and the 'strophometer' which recorded the speed are difficult to comprehend.

Mumford described how the graphically-recorded speed and resistance of the model was converted into speed and horse power for an actual ship.

He also noted that in addition to measuring the speed, resistance, and trims of the models, the waves created by them were also measured and plotted. The information obtained was of great value, because of the importance when designing a vessel - especially a paddle steamer - to place the paddle or screw in such a position that it will do its work with the maximum efficiency, and the positions and heights of the waves created by the vessel have a great deal to do with that efficiency.

With each model at Torquay or at Denny's, a series of experiments was made with various displacements and various trims, and in each condition several tests were done at different speeds.

Mumford pointed out an important matter with regard to indicated horse-power curves, wherein 'an indicated horse-power curve need not be what is called "fair" but may consistently have a "hump." Cases have been known where indicated horse-power curves have been drawn in "fair," although the spots showed a decided hump, yet this hump has afterwards been proved to be genuine in cases where a model of the same ship has been tried, it being borne out by the effective horse-power curve deduced from the model experiments. The present Mr. Froude, of Torquay, has discovered the cause of these humps and treats of them in a paper which he read before the Institute of Naval Architects in 1881. The following explains the cause:-
When a vessel is moving through water at a given speed it creates a wave the length of which is a function of the speed, consequently at certain speeds a crest of the created wave will be at such a point of the stern of the ship that the positive pressure due to the head of water will be most effective in pushing the ship forward, therefore at these speeds the indicated horse-power curve is hollow. When the speed of the ship is such that a hollow of the created wave is at the same point, then it is easily seen that at that speed the indicated horse-power curve is round. Hence in designing a vessel to go at a certain speed due attention should be paid to this matter so that the hull of the ship may be efficient at the required speed.'

1886 Descripton of the shaping machine


MESSRS WM. DENNY & BROS., (DUMBARTON) SHIP MODEL SHAPING MACHINE. ‘Engineering’ of the 6th inst. contains the following notice of the ship model shaping machine which is in use in the experimental department of Messrs Wm. Denny & Bros.' shipyard. : —

‘This machine is a modification of the well-known apparatus invented by the late Dr Froude. It is so designed that it can be made to cut any number of water-lines on a model not exceeding 21ft. in length, 3ft. in breadth, 2ft. in depth.

'The machine consists of a horizontal bed, on which the model to be cut is fixed, in its proper position, by means of central pins passing through two wooden beams attached to the model. The latter is made of paraffin wax ; a substance that lends itself admirably to this description of work. The bed is supported on a travelling carriage by means of four screws. The carriage is made to move longitudinally by means of a handwheel. There are two revolving cutters, which run at high speed, for removing the material. These are held by two supports depending [descending] from the cross beam. Vertical adjustment of the model is obtained by means of the four screws referred to, which are all operated by one handle. The two cutters revolve at a speed of about 1500 revolutions a minute. They are attached to vertical spindles, which in turn are supported by two frames. These can be made to move by a second handwheel either towards or away from the middle line of the bed of the machine. A half-breadth drawing of the water lines of the model is placed in a vertical iron frame. The latter is geared to the bed of the machine in such a way that it moves parallel to it, the ratio of travel of the frame adjusted by means of gear wheels to the same ratio as the length of the drawing is to the length of the model. A tracer is brought to the lines on the drawing end connects with the revolving cutters by means of lever and bell-cranks in such a manner that, as it moves vertically on the face of the drawing, the ratio of travel of the tracer to the travel of the cutter is the same as the ratio of the breadth of the drawing is to the breadth if the model. This ratio can be regulated by means of an adjustable fulcrum. The tendency of the cutters to vibrate is checked by a cataract or “dash-pot” in the centre of the framing. The shape of the tracer corresponds to the circle described by each cutter. It is generally an ellipse, the major axis being vertical. The reason for this is that the breadth scale is for convenience, usually made greater than the length scale. The ellipse is drawn in ink on a gummed surface of glass, the line being very close to the drawing when the tracer is in position.

'The cutter spindles are driven in the same manner as in Mr Froude's original machine at Torquay except that spliced cotton bands are used instead of cat-gut. The object of the change is to get of jerks caused by the hooks in a cat gut band.

'To set the machine for working the height of the model is adjusted so that the position of the cutters with respect to it, will correspond to the position of a particular water line. When the model is at the required level, it can be seen by means of a pointer moving vertically on a scale, the scale being divided according to the water-lines of the model. The pointer is fixed to the bed of the machine, the scale is attached to the carriage which supports the bed. The tracer is adjusted by means of a right and left-handed screw, cut on the rod carrying it. When the cutter circles touch, the tracer is at the centre line of the drawing. To work the machine the operator moves the two handwheels on the right and left of the seat. By means of these the tracer is made to follow the lines on the drawing, and it will be easily seen, from the foregoing explanation, that the cutters will move in harmony with the tracer, and so reproduce the lines on the model.

'In order to insure greater accuracy in tracing, the stool, on which the operator is seated, is connected the moving levers, so that the vertical rise and fall of the stool is exactly the same as that of the tracer. In this way the operator keeps his eye always on the same level as the tracer. After cutting one line the height of the model is adjusted to another water-line position, and the corresponding water-line is cut. The operation is then repeated until the whole of the lines have been reproduced on the model. The machine, however, does not produce a smooth surface, but cuts the material away in a series of horizontal terraces. These have to be removed by hand, the operator working to the lines left by the machine, and only removing superfluous material. One person can work the machine, but it is found desirable to have an attendant at each handwheel, so that one controls the travel of the bed of the machine,' whilst the other keeps the tracer on the water-line.

'The late Mr Froude described the original machine in a paper before the meeting of the Institution of Mechanical Engineers held at Plymouth in 1873. Some particulars of the machine are also contained in the report of the committee appointed by the Association to consider the different kinds of instruments for measuring the speeds of ships.’

See Also


Sources of Information

  1. [1] BSHS
  2. [2] Bridging the Seas: The Rise of Naval Architecture in the Industrial Age, 1800–2000 (Transformations: Studies in the History of Science and Technology) by Larrie D. Ferreiro, 2020, p.124ff.
  3. Lennox Herald, 3 November 1995
  4. [3] NHMF
  5. The Engineer, 15 Feb 1884, p.128
  6. Dumbarton Herald and County Advertiser - Wednesday 18 August 1886