In comparatively recent times the idea has been revived in a more elaborate form and the anti-rolling tank first projected by Frahm has been fitted in several ships. This tank is of U form extending from side to side of the ship athwartships and consists essentially of two vertical columns connected by a horizontal tube at the bottom ; the water fills the latter completely and about one-half of the vertical tanks. The upper parts of the vertical tanks are connected by an air pipe which is fitted with a valve or throttle by means of which the motion can be controlled. To prevent eddying during motion fixed baffles are arranged at the sides. One set of tanks is fitted near each end of the ship. It can be shown that once the shape of the tank is fixed the period is also more or less fixed and is practically independent of the amount of water.
For further information reference should be made to the paper by Frahm in Trans. I.N.A. 1911 Pt. I : and for the complete mathematical investigation see Mr. Woollard's paper Trans. I.N.A. 1913 Part II.
The general conclusions reached may be summarized as fol lows :—(1.) The "powerfulness" of the tank, or its capacity for affecting rolling depends principally on the proportion of the ship's metacentric height lost by virtue of free water surface in the tanks. There is also a relatively small hydro-dynamical effect from the inertia of the water in motion from side to side. (2.) The period of the tank depends on its shape, particularly the area of the constricted channel at the bottom ; it is almost independent of the amount of water contained and of the degree to which the flow is resisted. (3.) If a ship be likened to a pendulum, the ship and tank are equivalent to a double pendulum; to get the best results the period of the tank should be slightly less than that of the ship. (4.) The tank is effective in stopping heavy rolling but is useless or disadvantageous among very short waves and should then be put out of action. The U tank has been fitted in some liners and warships but as it does little to reduce moderate rolling it has not made much headway. More recently the adop tion of an anti-rolling tank, open to the sea, has been projected. The use of anti-rolling tanks in general has not found much favour and they are open to practical objections.
The idea of using a weight moving from side to side to prevent or reduce rolling at sea was tried by Sir John Thornycrof t in a yacht of about 23o tons displacement, the moving weight being 8 tons. The weight was worked automatically by apparatus con trolled by two pendulums, one of which—a long period pendulum —remained vertical and the other—a short period pendulum— placed itself perpendicular to the effective wave slope. The appa
ratus is heavy and cumbrous, but is sound in principle, and effected a reduction in rolling of about one-half in the yacht in which it was fitted (see Trans. I.N.A. 1892).
Diminution of rolling by the use of the gyroscope was investi gated by Dr. 0. Schlick and an arrangement fitted in the S.S. "See-bar." The ship was of 56 tons displacement having a metacentric height of 1.64 ft. and a period of roll of 4.14 sec onds. The flywheel of the gyroscope weighed 'Jo() lb. and was I metre in external diameter, revolving at 1,600 revolutions per minute. The axis of the flywheel was initially vertical, and the casing containing the wheel was capable of revolving about a hori zontal athwartship axis, the centre of gravity of the apparatus lying slightly below this axis. A brake was fitted to control the longitudinal oscillation of the casing. When the wheel was revolving and the axis held by the brake, no effect was produced on the motion of the ship ; but when the axis was allowed to oscillate freely in the middle-line plane the period of roll was lengthened to 6 seconds, but no other extinctive effect was ob tained. By suitably damping the longitudinal oscillations of the gyroscope, however, by means of the brake, a large extinctive effect resulted, and during the trials made, the apparatus prac tically stopped all rolling motion. (For detailed account of the principle of action, details of the gear, and description of trials, see Trans. I.N.A. 1904 and A further development of the gyroscopic method is the appa ratus devised by Sperry. This consists essentially of two main gyroscopes rotating and precessing in opposite directions ; the plane of precession being horizontal and the axes of rotation normally athwartships. The two gyroscopes thus combine against roll but they neutralize each other as far as yawing is concerned. An auxiliary gyroscope maintains true horizontal. The principle of action can be stated briefly as follows :—As soon as the ship has rolled through a very small arc, an electrical contact con nected to the auxiliary gyroscope is established making a current which actuates the precession motors of the main gyroscopes and these continue working until the ship is again horizontal. The gear has been fitted in several ships, one being the U.S. Navy transport "Henderson" of 1 o,000 tons and it has been generally satisfactory. One of the later gyroscopic installations is that in the U.S. liner "Hawkeye State" of about 20,000 tons displace ment. The total weight of the equipment is of the order of 200 tons.