Occasionally it is demonstrated to the public, when ships make less than their designed speed on the originally calculated power, what engineers and constructors know so well—that theory and empirical formulae are weak supports; more often such proof is given by the necessity of developing far more than the designed power in order to make the required speed.
There must be something fundamentally wrong with the present system of steamship design when from carefully designed warships such varying results are obtained, and of which a few are noted below:
(a) The Good Hope class of armored cruisers in the English Navy failed to make their maximum contract speeds as economically as was afterwards done when their propellers were changed.
(b) In order to make her contract speed, it was necessary for the protected cruiser Charleston of our navy to develop about 29,000 horsepower instead of the 22-000 horsepower which her designers expected would be all that was necessary to make 22 knots. Her propellers have been changed twice, so the Charleston is now wearing her third set, and with little or no reduction of the enormous excess horsepower to make the maximum speed.
(c) Unaccountable excesses in speed and in horsepower are sometimes developed on contract trials, as was done by the sister battleships Kearsarge and Kentucky.
(d) The sister battleships Alabama and Illinois developed more speed and power than designed, though to a lesser extent than the Kearsarge and the Kentucky.
(e) The sister battleships Minnesota, Louisiana, Vermont and Kansas, to make 18 knots, used respectively 15,000, 15,850, 16,45c and 18,950 indicated horsepower.
(f) The sister armored cruisers North Carolina and Montana had to develop from 26,500 to about 28,000 horsepower to make 22 knots, though the original designs called for 23,000 indicated horsepower to make this speed.
Many more such examples could be cited.
Why are these differences?
Instead of being a matter of course, it is, in fact, always especially remarked and advertised by the contractors if a ship makes the required speed on her contract trial; and especially so when the power then developed closely approximates the designed power.
In spite of all efforts to avoid previous errors, and regardless of every endeavor to benefit by past experience, the successful designers of ships and their machinery must accomplish many of their results by a set of rough thumb rules.
There have been several important series of experiments carried on by men of international reputation in ship building and engineering circles in reference to underwater bodies of ships and propellers and the power necessary to make certain speeds. If we accept the statements of facts in the preceding paragraphs marked (a), (b), (c), (d), (e) and (f), then the only conclusion to be drawn is that the results of such experiments were not binding on the designers or else the latter gave little weight to such data because of its incompleteness.
There is a crying need of data for all classes of ships, and designers of marine machinery and constructors of hulls are seriously handicapped in their endeavors. Guesswork, hobbies and suppositions have a weight in determining the shapes of hulls, of propellers and of power to make the speeds wanted that would make the tax-payer and the average ship-owner gasp, did they but realize the meager system of data used as the basis of designs. As yet, nearly a century since steamships were first used, there is no accumulation of data which, when used, will determine absolutely the exact power required to drive any given under-water form through the water; nor is there any compiled data which allows the correct and exact determination of the pitch, the diameter and the surface of propellers in order to apply this exact power economically.
Of course, there are certain empirical formulae which are used in default of something better; and it is generally known among ship-builders that, for certain purposes, a particular shape of under-water body is more economical or faster. To accomplish the same purposes, similar hulls are copied, because there is not positive information to show that another form would give better or same results at less cost.
There are many thousands of ships built and building, and it seems incredible that tax-payers and business men do not compel ship-building to become mathematically exact. Certainly some efforts should be made to acquire enough data so that predictions can accurately be made of the power required to drive at all speeds up to the maximum a given under-water form through the water. This is especially true of our navy, where succeeding ships in the same classes have similar shapes.
Certain men appear to have discovered for themselves the best or most efficient form of under-water body for the fast racing yachts, but such information is closely guarded as a trade secret.
Propellers for driving ships through the water are almost as much a mystery now as when first introduced, and to-day various theories prevail as to the best type to be used on ships having similar under-water bodies.
Most propellers at present fitted to the ships of our navy were designed from the results of the performances of sister ships combined with empirical formulae derived from various experiments. Within limits the guess of any one engineer seems to have had as much weight as another, and in many cases there are wide discrepancies. It is a matter of record that experienced and successful battleship builders proposed, and did attempt, to install on a battleship of 12,500 tons displacement, intended to make 18 knots speed, the same propellers as were fitted to a battleship of 10,288 tons displacement intended to make about 16 knots speed.
In the earlier days of our modern steam navy, reliable knowledge and dependable skill were rarer, but as ships after ships of the modern type have been commissioned, many men have gained experience and acquired knowledge, and they demand greater efficiency on a given displacement. But the maximum possible efficiency is unobtainable when absolute data of horsepower for various displacements and different trims at progressive speeds is not in existence, even for one class of ships having similar underwater bodies.
Somebody or some organization, sometime and somewhere, will begin the systematic collection and compilation of this most important data in order that ship-owners' moneys may be expended economically. "It is up" to the American Navy to make the start so that in the future it cannot be said that our men-of-war are mere copies and that we have not done our share in the promotion of progress. American minds can contribute to the science and art of ship-building as of old and to the economical powering and steaming of ships. Let us determine that the American-built steamers of the future shall be famous as were the clipper-sailing ships of the past.
To do this there must be the best sort of mutual co-operation— strong team work—of all the corps of the navy and of all officers doing line and engineering duties.
Believing this to not be impossible, it is therefore suggested that the following six points be investigated and accurate results ascertained for all the different kinds of under-water bodies, using in the beginning the material now built and in use.
1st. Trim: Discovering the "Economical Trim," thereby saving the continual but enormous loss of "Wasted Horsepower."
2d. Pitch of Propellers: Discovering the "Economical Pitch" by increasing or decreasing the pitch of propellers now in use.
3d. Area of Propellers: Discovering the "Economical Area" by increasing or decreasing the area of propellers now in use.
4th. Diameter of Propellers: Discovering the "Economical Diameter" by increasing or decreasing the diameter of propellers now in use.
5th. Having secured the 2d, 3d and 4th points immediately above, obtain the limits of revolutions of the "Economical Propellers" of various diameters, pitches and areas beyond which it is unwise, to go on account of cavitation.
6th. "Errors of Models" used in the experimental basin, so that from them an average "Error of Model" may be obtained, by which to judge the performances of any shape of model sought to be tried.
It is realized that the first cost of doing all this work immediately and at one time would be prohibitive, but there ought to be a system and plan of work laid out to cover a long course of years, which will be followed by our ships and their officers. Then time and money will not be wasted in the future as it has been in the past.
Determine for all time each fact in order, and then proceed to the next, so that the coming generations will gratefully feel the effect and not be hostilely critical of our expenditures of time, money and energy, as surely will be the case if present methods are not improved.
Formerly, in the days of sailing ships, men were very careful to ascertain the trim at which a sailing ship would make the greatest speed--whether it was better to have their ships trimmed down by the stern or by the head, and if so, how much. Of course, they were assisted in arriving at what they termed the correct result by the experience of hundreds of years; but even so, the proper trim was in many cases a matter of individual judgment, because exact data were unobtainable.
There is much time devoted to developing the shooting efficiency of all our ships, but where is the time that should be set aside to ascertain what the economical steaming conditions are? There has been too much of the haphazard in this, though we have methods of obtaining, at given speeds, the exact power required to drive a ship of a certain under-water form, having any reasonably predetermined trim, over several standard mile courses where the depth of the water is known.
To-day one hears very little about the best trim of a modern man-of-war, in order economically to make any given speed up to the maximum.
Contractors apparently have been allowed on the contract trials to run the new ships at any trim. Should they not be required to run them at the same trim then as the ships will have when in commission and full of stores, coal, water and ammunition? When our modern men-of-war are delivered by the contractors, the trim, as it happens to be, seems to be accepted as a matter of course.
Ships in commission and displacing moo or 1500 tons more than when running their contract trials often equal or exceed the contract speed. And no reasons are usually given; but certainly there are causes for such astonishing results! Why is it that these increased displacements, in such a large number of cases, make such small differences in speed on full-power runs?
In comparing sister ships it should be noted:
(a) That the Missouri trims by the stern and the Maine by the head, and the Missouri made more speed on her contract trial; and since then has always made more speed on full-power runs than the Maine.
(b) The Illinois trims more by the stern than does the Alabama and the Illinois habitually steamed more economically and also faster on full-power trials.
(c) Taking the Louisiana class of battleships, the one down most by the stern required least horsepower to make 18 knots and more power was required as the trim by the stern decreased.
(d) The following table shows for the U. S. S. Montana the difference in data:
1st. "When that ship trims 7 3/8 inches by the stern.
2d. When that ship trims 5 3/4 inches by the head; the screws being the same both times and the displacement differing but 345 tons. The Montana was docked in December, 1908, four months after this data was taken, and the bottom was then found to be almost clean, having but little grass and but few barnacles, and they small.
It will be noted that, although the ship's force developed 411 horsepower more than the contractors, yet because of bad trim, 3.77 less revolutions per minute were made with consequently slower speed: the coal consumption was about the same. Consider the loss in power and speed caused by improper trim!
(e) The North Carolina shows similar data to that in paragraph (d) above. Consider her loss in power and speed caused by improper trim!
Since November, 1901, the U. S. S. Kearsarge has in various years standardized many times. The data of six of these on three different standard mile courses are given below in Table II:
From the data in the above table the curves in Plate I are drawn.
Of course, in any sets of observations and tables taken by different sets of men at different times many errors must enter, but the general conclusions to be drawn are as follows:
(a) "When the Kearsarge is at, what for lack of a better name may be called the "Economical Trim," changes in displacement, even as much as goo or moo tons, do not apparently seem to require a noticeable increase either of horsepower or of revolutions.
All four curves above seem to confirm the preceding statement (a), also those immediately following in (b) and (c). Imagine what great advantages could have been taken of this, if it had been known in time, so as to have allowed a heavier battery, more armor, more speed, or a greater cruising radius to the Kearsarge and to the Kentucky!
(b) Changes in foulness of the bottom necessitate a small increase in revolutions of the main engines and a large increase in horsepower, hut in the latter case it does not seem to make so large a difference as has heretofore been generally believed.
(c) Changes in trim make great differences in the necessary revolutions and power and the coal consumption at the same actual speeds over the ground.
From Table III the coal-speed (knot) curve on Plate VI is plotted; and also the coal-speed (revolution) curve on Plate VII.
When ships habitually cruise, at sea in squadron, at any given speed, a standard number of revolutions is required in order to maintain position. The major part of coal on the larger men-ofwar is carried forward the midships section, and as coal is used for steaming, the trim changes, thus necessitating a changing of revolutions, which is usually to a decreased standard revolutions. Below in Plate VIII is a set of revolution-trim-speed curves of the Kearsarge obtained from three trims:
1st. 12 inches by the stern.
2d. 20 inches by the stern.
3d. 36 inches by the stern.
(d) When the Kearsarge is at her "Economical Trim"—between 20 and 28 inches by the stern-the following facts obtained from Plate VIII should be noted.
1st. Fewer revolutions are required to maintain the same speed;
2d. Consequently less horsepower is needed;
3d. And therefore from 5 to 6 per cent less coal is burned than when trimmed 12 inches by the stern.
The gain, or saving of coal, to be made in the steaming of the Kearsarge, by considering the neglected factor of trim is certainly a strong argument to continue to its legitimate conclusion, this investigation and the procurement of data at all speeds for all the probable trims that might arise during service.
Tables and curves similar to all these of the Kearsarge here shown could probably be made for all other ships of the American Navy; and all our ships should proceed at once to obtain the necessary data.
Plate IX taken from the data of the standardization trials of the Louisiana class of battleships shows that enormous increase of horsepower is necessary when the ships of this class are trimmed not enough by the stern, though they have very nearly the same displacements and all make the same speed-18 knots.
There is apparently a relation between the horsepower for the Minnesota down by the stern and the horsepower of the Kansas down by the head. On the Kansas 4950 horsepower were apparently wasted. Certainly it seems as though this power must havebeen uselessly expended: but was it, when we take into consideration the neglected factor of trim?
The theory of the writer is, that that amount of excess power developed by the Kansas was used up in changing her trim and therefore her under-water body shape to that assumed by the Minnesota when the latter was making 18 knots. Lacking a better name, let us call such excess power "Wasted Horsepower" and assume the following formula for the work that is done.
Wasted horsepower — T x D 2,240/33,000
where T is the change of trim in feet, and D is the displacement in tons.
Then we have an expression for the work that must constantly be done to keep the Kansas trimmed so that she may make the required contract speed—that made by the Minnesota.
This formula may possibly be found to be, by the experiments proposed on pages 14 to 17, a general law in the application of power to ships in order to make the higher speeds.
Using the above formula and by it reducing the Minnesota, the Louisiana and the Vermont to a trim by the head of inches— and that of the Kansas—they would require respectively 3060, 1730 and 1000 more horsepower to make 18 knots than was developed on the trims they had on their contract standardization trials.
The propellers of the Kansas have been changed since the above data were taken and less power is now needed in order to make 18 knots: the horsepower necessary to make this speed more nearly approaches a mean of her sisters.
The results of the last column do not conform exactly to the formula assumed for "Wasted Horsepower," but many different errors, personal, sea, wind, etc., may enter into these sets of observations taken at different times, under different conditions of weather and sea and by different sets of observers. There may also be other factors as yet unknown which require the use of some of this "wasted horsepower."
Although the Minnesota gave better results than any of her sister ships when she was 31 inches by the stern, is that her proper trim? The curve shows that she could have made 18 knots on less horsepower if she had been trimmed more by the stern.
If the "Economical Trim" of the Minnesota is 31 inches, she being roughly 450 feet long; and if the most "Economical Trim" of the Kearsarge is between 20 inches and 28 inches, the latter being roughly 370 feet long; what is the most" Economical Trim" of the North Carolina class, they being roughly 500 feet long?
Why is there no second general law that tells us the "Best Angle of the Keel with the Surface of the Water" in order that we may at once determine the "Economical Trim" of the North Carolina class?
The underwater bodies of all three classes of ships may possibly be considered similar; then the illustration below might obey this as yet unknown general law of the "Angle of Trim," which surely must exist. If it does cover the case, then approximately the most "Economical Trim" of the North Carolina class would be 34.7 inches by the stern if 31 inches is the best trim of the Minnesota, and full of stores, coal, water and ammunition, to 36 inches by the stern will necessitate 53 inches x 1600 (approximate moment to change trim 1 inch), or 84,000 foot tons. Could not this have been taken into account and allowed for when the designs of this class of ships were first made?
Consider the "Wasted Horsepower" of the North Carolina and of the Montana the first year of their commission when they cruised about 30,000 miles and burned 20,000 tons of coal! Think of the "Wasted Horsepower" of the Louisiana, Vermont, and Kansas in the past! and then again in the future, unless this subject is investigated and true results ascertained! Think of the worse than wasted coal and the useless hard work of the human beings who made steam for the "Wasted Horsepower" of any and all ships of the merchant marine and of all navies, past, present and future when under way and not at their most economical trims! Think of the useless wear and tear on the boilers and on the machinery when developing this "Wasted Horsepower!"
What is the effect of varying the angle of the bilge keels to the stream lines? Ought not there to be calculated tables supplied which will show the increase of head resistance and its cost in "wasted horsepower" when the trim is wrong?
We hear much about the necessary horsepower required in order to make a certain speed, while the essential element of trim is apparently neglected.
All merchantmen know in a general way and from intimate every-day experience that a ship steams faster under the same daily coal consumption when trimmed down by the stern. But where is the exact data that is to inform them what the most "Economical Trim" is? What have they to guide them in their endeavors except a thumb rule to stow the cargo so as to make the stern as deep as possible?
And are not men-of-war in the same relative condition of ignorance and neglect?
Plates X and XI clearly show the tendency on the part of smaller craft to squat and for the bow to come out of the water when making a high rate of speed. Probably if large ships made the same speed the forward portion of their keels would be proportionately as much out of the water.
When the larger ships are traveling at the higher speeds we know that they draw less water forward and probably more aft than if floating at rest.
Some lucid explanation should be sought of the causes of the constantly increasing ratio of horsepower to speed as the speed increases above the "Economical Speed." On every ship the following should be ascertained and kept on file for constant use:
1st. The "Economical Speed."
2d. The power used at the various speeds for propulsive effect alone.
3d. The amount of power used at each speed for changing the trim—"Wasted Horsepower."
4th. The amount of power used in securing the proper "lifting effect" necessary before each higher speed can be made.
5th. The skin resistance clean, and in different stages of foulness.
6th. Head resistance with bilge keels having a head resistance of zero or at its minimum. Also the head resistance of the bilge keels at all the various improper trims possible.
At the present time there are in our navy several classes of ships, and in most cases sister ships have similar under-water bodies. Why not have ships of the same class having the same or similar under-water bodies standardize at the same time? The average result would be accurate because of the elimination of personal and other errors.
Therefore, it is suggested that battleships and armored cruisers standardize their propellers as follows:
1st. Either singly,
2d. Or else in squadrons; in case it is determined that the wake of a ship 400 or 5oo yards in advance has no effect on the ship or ships following.
A (for trim) : At speeds of 8, 10, 12, 13, 14, 15, 16, etc., up to maximum speed at increments of one knot, and that indicator cards be taken during each run.
1. 12 inches by the head.
2. 6 inches by the head.
3. On an even keel.
4. 6 inches by the stern.
5. 12 inches by the stern.
6. 18 inches by the stern.
7. 24 inches by the stern.
8. 30 inches by the stern.
9. 36 inches by the stern.
B (for varying pitch) : Having ascertained the trim to be known as the "Economical Trim" at which each ship will make the most speed under the least revolutions with consequent least expenditure of power and smallest coal and water consumptions then standardize at that "Economical Trim."
10. With a decrease in pitch of 6 inches.
11. With a decrease in pitch of 12 inches.
12. With a decrease in pitch of 18 inches.
13. With an increase in pitch of 6 inches.
14. With an increase in pitch of 12 inches.
15. With an increase in pitch of 18 inches.
Possibly the information discovered as result of experiments numbers io-15 inclusive, will show that there is a more "Economical Pitch " than that at which the propellers of the vessel under trial are set. If such is found to be the case, then further experiments should be made at the increased or decreased pitch, whichever is the more economical to determine the maximum efficiency to be gained with the propellers of that diameter and area.
C (Testing of varying propeller areas) : Trials should be made with increments of i square foot in area of the surfaces of propellers and then with decrements of i square foot in order to arrive at the "Economical Area" of propellers. Tests would thereby be made of the prevalent theories concerning the efficiency
supposed to be gained by increasing the surface of propellers.
D (Testing varying propeller diameters) : Trials should be made with different diameters of propellers, varying the diameters by 6 inches each way in order to arrive at the "Economical Diameter" of propellers.
E (Finding when cavitation begins) : Once the "Economical Pitch," the "Economical Area" and the "Economical Diameter" of a propeller is discovered that becomes the "Economical Propeller." Then construct a suitable tank where the propeller or its small model may be revolved sufficiently fast to set up cavitation, determining carefully with suitable instruments the power required at the various revolutions. The value of this data would be of scientific and practical value for turbine-driven ships. Besides which more would be discovered concerning this curious phenomenon of cavitation.
Experiments called for under paragraphs B, C and D immediately above will no doubt be enormously expensive; the writer hopes that more economical methods can be found which will furnish the results so essential to ship designing and ship powering.
F (Error of model) : Many efforts have been made to make accurate comparisons of the steaming performances of different types of vessels and the writer believes it to be impossible of accomplishment, except by comparing the results obtained under same conditions and from the comparisons of small models of ships of the different classes. In recent years an experimental basin has been built and splendidly equipped at one of our navy yards and sea-going officers and the service at large (as well as engineers and ship-builders in civil life) hope for useful results and consequent improvements in the designs of succeeding ships.
Experiments with models are useful and instructive, but they could be made invaluable in the future designs of hulls, and in determining the horsepower necessary to drive ships economically at the predetermined speeds, if the following method is adopted:
After obtaining the data as proposed in the preceding pages, then construct for each ship, whose data are so obtained, a model, and from it obtain similar data. The difference between the performances of the ship and the performances of the model in the experimental basin would give what may be called the "Error of the Model." The "Errors of the Models" of many ships would give an average "Error of Model" which would be of vast benefit not only in naval construction and in the design of naval machinery, but would be of incalculable benefit to the shipping world at large.
Any proposed shapes of hulls within the bounds of reason could easily and quickly be compared with models of ships whose exact data had been already determined. And hull designers could cheaply and in very short time try out all their pet theories as to proper forms of under-water bodies for ships of all types, whether racing yachts or colliers—torpedo-boats or battleships.
More experience might also be obtained so that the direction of stream lines would not still be shrouded in darkness.
It seems to be the established policy of the Navy Department to have the Atlantic Fleet cruise near Provincetown during the summer and near Guantanamo during the winter, and off both these localities there are good standard mile courses, though with different depths of water. The retardation caused by shallow water might be more than guessed at if comparisons of similar performances of many different ships were made, for at Guantanamo the water is much deeper than at Provincetown, Barren Island or Rockland.
An immense amount of time during each year is devoted to working out squadron maneuvers, to preparation for target practice, to giving liberty and for repairs. While the first two of these are being done, ships could easily obtain the data that hull and engine designers need so urgently. Even if two weeks, twice a year, were required on this most important duty, such a small amount of time should not be withheld where efficiency and economy are the returns?
Our navy is constantly growing and the time will soon come when economy of operation will be demanded by tax-payers, and this is to be obtained only by discovering the maximum possibilities, and designing ships and their machinery therefrom.