The effects of large-scale engineering schemes, such as harbour works, hydro-electric plant or river control, can be studied beforehand by the use of scale models, of which some remarkable examples have been made
REPRESENTATION OF THE RIVER MISSISSIPPI, showing, in the foreground, the Yazoo River area, in the United States Waterways Experiment Station at Vicksburg, Mississippi. The model of the river represents a total length of ninety-six miles in the region known as the Greenville bends. The horizontal scale is 1:4800 and experiments are carried out to determine the effects of cut-offs and other engineering proposals.
THE engineer uses scale models for many important purposes. Exact scale models, shaped by ingenious mechanisms, are used extensively for determining the best form to be adopted when designing ships. By the proper use of models of this type it is possible to save thousands of pounds when designing new hull forms, particularly those of large or fast vessels.
For instance, more than sixteen different waxen hull forms of the Queen Marywere tested, and thousands of experiments were carried out on them, by towing or propelling them through water, so that the best possible hull design could be evolved. An interesting series of tests was made on a scale model of the vessel to determine the height to which the funnels had to be built, so that there would be no possibility of the smoke from them reaching the after decks.
For this purpose the model was tested in a wind tunnel, similar to that used for testing model aircraft. A wind was created in the tunnel, and air was expelled from the funnels and drawn in through the ventilators in a manner corresponding exactly to conditions which would be expected when the vessel was travelling at full speed. The air from the funnels was passed over chemicals, so that a dense white smoke was formed, and the path of this could easily be seen. There was little difficulty in studying the required conditions beforehand; thus was saved the large expense which might have been entailed by alterations to the completed ship.
For many years harbour engineers have made considerable use of tidal models when studying the many problems presented by port and dock development. This method has been used for works in the Mersey, Severn and Humber estuaries, and in many instances abroad. One of the most remarkable models of this kind was that specially built by Sir Alexander Gibb and Partners for studying the approaches of the Port of Rangoon, Burma.
This is believed to be the largest model ever built for this purpose, and the information obtained from it was of the greatest possible value in compiling the report on the port development. Rangoon, the principal port of Burma, lies on the Rangoon River, over twenty miles from the sea. It deals with more than 90 per cent of the overseas trade of Burma and is one of the largest passenger ports in the world. In addition to this, it is the heart of a vast inland water transport system which involves about 1,000 miles of navigable waterways on the River Irrawaddy and its tributaries.
The whole of Burma, with its huge area of 260,000 square miles and its population of 14,000,000, forms the hinterland of the port. The exports flowing from it represent an average annual value of some £25,000,000. Thus the prosperity of Burma is to no small extent dependent on the prosperity of the Port of Rangoon.
The model used for these experiments was built in the basement of University College, London, and measured approximately 45 feet by 40 feet. The full size area represented measured about fifty-five miles from north to south, and about fifty miles from east to west. The exact scales to be used were all carefully worked out, and an ingenious arrangement of plungers simulated the effects of the tides. The main plunger measured 15 feet by 2 feet, and the tidal variation between spring tide and neap tide was reproduced by epicyclic gearing, which automatically varied the length of stroke given to the plunger.
Surveys of the haibour had shown that there was also a diurnal variation of the tide in the Gulf of Martaban, and for this purpose a smaller plunger was fitted as well. This plunger ran at half the speed of the main plunger. On the model the tidal period between two successive high waters was represented by seventy-six seconds, and one day by about two and a half minutes. It was thus possible to reproduce the tidal happenings of one year in the amazingly short space of some fifteen hours, and a week’s
continuous running of the model represented approximately ten years.
The foundation of the model consisted of more than twenty tons of concrete, in which were modelled the bed of the rivers and of the gulf. In addition to this, some fifteen tons of sand were required for moulding the bed of the gulf and the rivers. To imitate the conditions at Rangoon as nearly as possible, great pains were taken to find the sand which would be exactly similar to that of Rangoon, allowing for the scale of the model.
Movement of Sand Studied
The movement of the sand due to the tides was studied by introducing a special colouring matter into it; and different colours were used at different parts of the river so that the source of material forming the bars and shoals could be seen without any difficulty. The amount of water used when the model was operating normally averaged about 10,000 gallons a day.
One of the most important problems to be studied was the effect of silting. For this purpose two tanks were provided for the supply of silt-laden water, one for sea water and the other for river water. It was discovered in the course of these experiments that the sea water in the model had to be nine times as salt as the sea to have the correct precipitating effect on the silt.
From this wonderful model it has been possible to predict with fair accuracy the future tidal conditions in the Rangoon River and the estuary. It has also been possible to study the best methods for improving the conditions of the port, and for ensuring a permanent and adequate approach channel in all conditions.
A model of a similar type is used by the Waterways Experiment Station at Vicksburg, Mississippi, where a great deal of valuable work has been carried out for the purpose of solving river control problems. This particular model was made for studying the tortuous length of a section of the Mississippi River.
The territory represented by the model has a total length of forty-two and a half miles, whereas the distance by the river is more than double, about ninety-eight miles. This section of the river is known as the Greenville Bends area, the channel varying from wide shallow sections to deep sections across the pools of the bends.
The purpose of this model is to determine the exact effect of excavating cut-offs through the sharp bends of the river, and to provide information about the discharges and velocities. The question of silting has also been carefully studied, and it was found possible to study with reasonable accuracy the effects of scour.
U.S. WATERWAYS EXPERIMENT STATION at Vicksburg, Mississippi. In the groundj are large models of parts of the River Mississippi, in which the effects of proposed control undertakings are carefully studied. Water is admitted to the model from the reservoir seen at the left.
For technical reasons the horizontal and vertical scales of the model are different, the horizontal scale being 1:4800 and the vertical scale 1:360. An ingenious method was used for making this model, the first stage consisting of setting in place a large number of accurately cut galvanized-iron templets, the top edges of which represented the contours of the river bed and the banks.
These contours were then modelled in moist sand, the top of the sand being left about an inch below the tops of the templets. An even covering of cement grout was then filled in to the top of the templets, and all the necessary details were modelled in place. This work had to be done with great care to prevent seepage of water between the cement and the sand. Accurate gauges were then fitted F to the model, so that differences of level to the nearest thousandth of a foot could be read with ease.
The working of the model was checked by a number of test runs, by which comparison could be drawn between the records of the river flow and the flow measured on the model. In every instance it was found that there was remarkable conformity between the model results and those found in practice. Several hundreds of readings were taken over a period of some months.
It was discovered that there would be changes in river surface levels of as much as 10 feet at a distance of nearly 443 miles below Cairo, Illinois, if four bends were cut through at Greenville Bends, and only 2 feet at nearly 449 miles from the same place if only one bend was cut through. The discovery was made also that a cut-off on the river causes a general lowering of the water level above the cut-off, with little change below it.
Experiments were carried out also to determine the effects of erosion and the deposition of material in the river. It was found, for example, that when bars were built up below individual cut-offs there would be no permanent effect on the levels.
There is also another remarkable river model at Vicksburg, made for the purpose of studying the flood control plans of the lower Mississippi Valley, a matter of great importance to many thousands of American citizens. This is one of the largest hydraulic models ever built, and represents a length of some 602 miles of the main river, five tributaries in the area covered, and the by-pass to Atchafalaya Bay on the Gulf of Mexico, embracing a total area of some 16,000 square miles.
Such hydraulic models as these are of the greatest possible value, since by their use it is possible to save many thousands of pounds in the carrying out of river and harbour improvement schemes. The engineer is able to plan a complete dredging or reclamation plan which will be effective and economical.
For Testing New Inventions
One of the most important purposes for which engineering models are used is in connexion with new mechanical ideas and the testing of new inventions. In the past, before the basic principles of engineering practice were as well understood as they are to-day, the enthusiastic perpetual motion inventors were soon convinced by a model of the error of their ways. As a general rule, it can be assumed that if a model will work efficiently, the full-size machine will also be a success.
It was that fine craftsman James Watt who experimented with a model of Newcomen’s engine, and from it he evolved the idea of the condensing engine (see also the chapter “Origin of the Steam Engine”. His invention formed the basis of modern steam practice, and here is a classic example of the advantage of a model to an experimental engineer. Watt made a large number of models, from which he was able to invent a number of new mechanisms. Many of these models can be seen in the Science Museum, South Kensington.
Some remarkable model experiments were recently carried out in France in connexion with the design for a streamlined locomotive. The wind tunnel of the St. Cyr Aerodynamic Institute was used for this purpose, and an exact scale model of the locomotive was tested with air streams of different velocities, and with the air blowing upon it from different angles, to determine the effect of different wind conditions. The object of these experiments was to discover the effect of a special form of smoke deflector known as the Huet deviator.
It was found that when operating in calm air this form of deflector gave an economy in horse-power of 41 per cent, at a speed of 90 miles an hour, as against 32 per cent for complete streamlining. When the angle of the wind, however, was more than ten degrees it was found that more power was absorbed than with pure streamlining. With the wind at an angle of between 60 degrees and 160 degrees, the horse-power absorbed may be greater than that of a normal engine.
ONE-TENTH FULL SIZE, this model of a Coppee coke oven measures 7 feet long, 2 ft. 9 in. wide and 5 ft. 6 in. high. Sixty drawings were made in preparation of the model and hundreds of castings were prepared for the metal parts. Other parts are made of limewood. The model was exhibited at the Brussels Exhibition (1935).
Working models are useful for illustrating plant or apparatus which would otherwise require a large number of working drawings and cross-sections to make their operation clear.
An excellent example of this type of model is afforded by a model coke oven which was exhibited at the Brussels Exhibition (1935). It represents a Coppee coke oven and is carried out to a scale of one-tenth full size, measuring 7 feet long, 2 ft. 9 in. wide, and 5 ft. 6 in. high. The main brick structure of the oven is made of lime wood, and every brick joint is marked on the surface. The model was prepared from as many as sixty drawings.
In many instances it is impossible to discover from a drawing whether or not a mechanism or apparatus will work properly. An interesting example of this was afforded some time ago, when a new form of retractable undercarriage was being developed by an aircraft manufacturer.
A model of the undercarriage in question was made from rough sketches and instructions, as it was found to be impossible to transmit by a drawing the information needed. This initial model did not work as well as had been expected, and it was not until several modifications had been made to the model that success was finally achieved. The cost of such a model is insignificant compared with the cost which might be involved in the alteration of a design which had reached the production stage.
Illumination is a subject which is being studied with great care nowadays, and even here models can be used with advantage. A model of this kind was made a short time ago for studying the effects of various forms of lighting in a scale model of a classroom. The model was in the form of a large sealed cabinet, fitted at certain points with a number of small electric light bulbs, which could be controlled singly or in groups from a central switchboard.
The interior of the cabinet was viewed through an eyepiece, and a cube painted dead white was then placed within the cabinet on the floor. As the lights were operated, the effects of various arrangements of illumination could be clearly seen on the cube which, if evenly lighted over all its faces, appeared white all over. If the light fell more upon one side than upon another, then one side appeared to be grey. From this model it was possible to work out the best possible lighting system, with due regard to economy, effectively preventing shadows and unbalanced lighting.
Miniature Ship Canal
The engineer makes considerable use of models for illustrating technical points to the uninitiated in suits which come before the law courts. An excellent example of this type of model was built for illustrating the work involved in the water supply of the City of Birmingham. This was being discussed in the House of Lords, and the plans were opposed by certain authorities in Glamorganshire. Explanation of the true facts with the aid of the model caused the case against Birmingham to be dropped. The whole matter was settled in a few minutes, whereas without such an aid discussion and litigation might have gone on for a considerable time.
Models are used extensively for conveying a complete picture of large civil engineering schemes. For example, some years ago a complete model of the Manchester Ship Canal was made. This model was in relief and showed all the principal features involved in the scheme. The model was some 30 feet long and enabled the whole scheme to be studied at a glance and to be understood by the layman.
Models of this type are used also by structural engineers and architects, who are able to see exactly how the proposed building or structure will appear when it is completed. Alterations can be made on the model at a minute fraction of the cost which would be involved in modifications to the completed building.
One of the most remarkable examples of the use of a working model in planning a large engineering work was in connexion with a hydro-electric project in Canada. The work involved was unusual in the sense that a dam had to be built across the middle of a fast running river, and an ingenious construction was devised for this purpose. It was decided that the only possible way of building the dam, which was to be in mass concrete, was to erect it in a vertical position at the side of the river, and to cause it to fall into the water by a charge of dynamite placed under the portion resting on the bank.
Such an operation was extremely delicate to perform properly, as the slightest deviation from the required position would cause the dam to be badly placed in the river. To study the proposed method, an exact scale model of the dam and river bed was made, and a cinema film was taken of the model dam falling into the river under the action of an equivalent quantity of explosive corresponding to that which would be required in the full-size dam. The film showed that this idea was perfectly feasible, and the full-scale operation was carried out without the slightest difficulty.
MODEL OF THE PUMPING STATION of the Metropolitan Water Board at Hampton, Middlesex. The model contains representations of the turbines, pumps, condensing plant, electrical equipment, generators and instruments, and so forth.
