306 



FOREST AND STREAM. 



[April 6, 1893. 



5HiP M O PE L . A PP AR ATUS g 



Apparatus for Model Experiments. 



Most yachtsmen are aware of the extensive series of model experi- 

 ments which have been carried on for over twenty years under the 

 auspices of the British Admiralty, first by the late Wm. Froude, and 

 since his death by his son, B. Edmund Froude. The method of making 

 these experiments, and the curious and ingenious machinery that has 

 been gradually invented will, however, be new to most of our readers, 

 and we therefore reproduce entu-e from the pages of Engineering a 

 paper recently read by Mr. Froude before the Institute of Mechanical 

 Engineers, as follows: 



The principal feature of the present Admiralty Experiment Station 

 at Haslar, as of the former one at Torquay, consists in a large cov- 

 ered waterway, in which models of ships are towed to ascertain then- 

 resistances. The towing is done from a dynamometer carriage driven 

 at deflnite speeds by a stationary engine worliing a wire rope. The 

 models are made of hard paraffin, generally about 14ft. long, and 

 something upward of lin. in thickness when finished. They are cast 

 in a mold with an allowance of about ''/^in. for finishing accurately to 

 shape, which is done by means of guidance grooves cut in the paraf- 

 flne by a shaping machine specially devised for the purpose. 



A brief general description of the experimental tank and apparatus 

 of Torquay was given in a paper read to this Institution hy Blr. Eob- 

 ert Gordon in May, 1884, and a full description of the model shaping 

 machine as it then existed at Torquay, was given in an earlier paper 

 read by the late Mr. William Froude at the Cornwall meeting in July, 

 1873. The object of the present paper is to describe the principal nov- 

 elties of mechanical interest in the experimental apparatus now used 

 at Haslar. 



WcCterivay and Experimental Carriage. — The waterway at Haslar, 

 Pigs. 1 to 4, is nearly 400ft. long and of nearly uniform section 

 throughout. Instead of haMng only about 200ft. of uniform section, as 

 at Torquay. The sides are concrete and vertical, instead of asphalted 

 earth slopes, and the railway on which the dynamometer carriage 

 runs is bedded on the tops of the side walls of the waterway, instead 

 of being suspended over the water from the roof. The railway is ex- 

 tended beyond one end of the main waterway, so that the experiment 

 cari'iage can be brought over a dry pit for getting at the experimental 

 apparatus from beneath. 



The experiment carriage, which has to span the whole width of the 

 tank, and therefore to run on a i-ailway of nearly 21ft. gauge, is a 

 trussed structure, shown in perspective in Fig. 5. Its principal 

 peculiarity consists in the fact that the members of the several 

 trusses composing it are wooden trunks or boxes about 4in. squai-e in 

 cross section, made of %in. deal and put together with screws and 

 shellac varnish. 



At the joints formed by the intersection of the various members of 

 the trusses the sides of the boxes are made to overlap one another 



over a large area, providing a large surface for screwing and for the 

 adhesion of the shellac varnish. The dimensions of the boxes form- 

 ing the several members of the girders are assigned so as to bring the 

 sides of the boxes into the right planes to suit these overlaps. Views 

 of one of these intefsections, showing the method of overlapping, are 

 given in Figs. 6 and 8. The whole structure thus produced is remark- 

 ably rigid and hght. The general design of the carriage is arranged 

 so as to leave clear a sort of central alley provided with a railway, the 

 rails of which are close to the sides of the alley. The object of this 

 secondary railway is to carry the smaller carriages, on which are 

 mounted the actual experimental apparatus of different kinds, so 

 that these may be adjusted on this railway to any desired position 

 fore and aft on the main carriage, 



StaUonm-y Engine and Haiding Gear. — The wire rope by which the 

 experiment carriage is driven from the stationary engine is not wound 

 upon a barrel, as it was at Torquay, but is simply led over a grooved 

 sheave driven by the engine, the necessary tension of the wu-e rope 

 providing quite sufficient adhesion for the purpose. The engine is a 

 lOin. Tower spherical engine (Proceedings 1885, page 96), the ample 

 power of which is required only for starting the truck quickly for 

 high speed experiments. The ordinary speeds of experiment range 

 between 100ft. and 500ft. per minute; for some classes of models ex- 

 periments are occasionally made up to about 850ft. per minute, or 9J^ 

 miles an hour. The truck has been run at over 1,200ft. per minute, or 

 about 14 miles an hour. 



Governor. — The engine is regulated by a governor, acting on the 

 same principle as that used at Torquay and described in Mr. Robert 

 Gordon's paper, but somewhat improved in detail, having been origin- 

 ally designed in accordance with that principle, instead of beihg merely 

 adapted to it. as was the instrument at Torquay. Tlie arrangement of 

 the governor is shown in Figs. 9 to 11, and to a large scale in the de- 

 tails. Pigs. 12 to 14. The lower ends of the two symmetrical beU cranks 

 A A are attached to each other by the links B B, having slotted holes 

 which aUow the beU cranks a very small range of freedom of angular 

 motion When a certain speed of rotation is reached the centrifugal 

 force of the weights C C overcomes the tension of the spiral spring D, 

 which tension may be adjusted by the coupled screws B and P to any 

 desired amoimt within a considerable range. But before the bell 

 cranks, in yielding to the surplus centrifugal force of the weights, 

 reach the outer limits of their very small tether, b3' means of the 

 hooked rod E they put pressure on the stationary wheel G, pressing it 

 down upon the rotating wheel H above the top bearing of the gov- 

 ernor, and by so increasing the friction between the faces of the 

 wheels thej' cause the stationary wheel to attempt to turn with the 

 governor, and thereby to extend the spiral spring connected with the 

 throttle valve, and partljr to close the latter. The extension of the 

 spring, and the consequent distance of departure of the throttle valve 

 from its full open position, are proi^ortional to the frictional turning 



movement applied to the stationary wheel, which movement is in itself 

 proportional to the pressure brought to bear upon it by the bell 

 cranks; in other words, it is proportional to the excess of the speed 

 above that at which the centrifugal force of the weights just equals 

 the tension of the spiral spi-ing. To give greater sensitiveness of 

 action the beU cranks are hung not on pin joints but on flat springs S, 

 Figs. 13 and 14, after the fashion of a clock pendulum. The apparent 

 pin joints at K are only safeguards to prevent the instrument from 

 flying to pieces if any of the flat springs should break. While the 

 springs remain sound the pins do not touch the insides of the holes. It 

 was found advantageous to coat the rubbing face of the stationary 

 wheel G with wood veneer, to give steadier friction. 



General Arrangement of Model- Shaping Machine.—The present 

 model-shaping machine is shown in Pigs 15 to 17. As in the case of 

 the Torquay machine, described in the Proceedings of the Penzance 

 meeting, the machine shaping of the model consists in cutting a 

 series of grooves in the rough model, to the depth of the intended 

 finished surface. Pig. 15, at the successive levels of the series of 

 waterlines in the half-breadth plan, the surface between the grooves 

 being finished off by hand and eye. These grooves are cut by a pair 

 of revolving cutters, between which the model passes, traveling in a 

 fore-and-aft line, while the cutters are moved laterally, receding from 

 and approaching each other sjTnmetrically, in such accordance with 

 the longitudinal travel of the model as to trace in plan upon it the 

 intended horizontal section or waterline CFig- IT). This due accord- 

 ance of the lateral motion of the cutters with the longitudinal motion 

 of the model is accomplished by the operator so regulating the cutter 

 motion as to maintain a tracer in contact with the corresponding 

 waterline on the drawing. By suitable mechanism the drawing itself 

 is made to imitate the longitudinal travel of the model, while the tracer 

 imitates the lateral travel of the cutters. In the Torquay machine, at 

 the time of the Penzance meeting, tbe tracer was maintained in con- 

 tact with an adjustable template set to the curve of each waterline. 

 Afterward the tracer was made to foUow the line on the drawing 

 itself by eye, assisted by a magnifying glass on account of the small- 

 ness of scale of the drawings used in that machine. 



The principal features of difference between the shaping machine at 

 Haslar and that at Torquay are as follows: 



Firstly, the adjustment of relative level of model and cutters, re- 

 quisite for cutting the successive waterlines, is effected at Haslar by 

 raising and lowering the cutters themselves, instead of the model. 

 The mam advantage of this arrangement Ues in a firmer vertical sup- 

 port for the model throughout its entire length, since it now rests 

 directly upon every one of the six axles of the model table V, which 

 are spaced 3ft. Gin. apart; the arrangement also lessens the vertical 

 height required for the whole machine. The two cutter spindles R, 

 which run at 3,700 revolutions per minute, are mounted inside stout 

 steel tubes, as shown in Fig. 18; the cylindi-ical outer surfaces of the 



