Naval History Homepage - and Site Search

Period of "Pax Britannica", 1815-1914, Technical Development from Contemporary Sources


From the United States Navy Manual, "Naval Ordnance and Gunnery" 1937

 First battle of ironclads - CSS Virginia v USS Monitor, 1862 (Photo Ships, click to enlarge)

return to "Pax Britannica", 1815-1914


Most of is about the warships, men and battles of the Royal Navy in the 20th century. Moving back into the pre-World War 1-era and before that the Victorian period, I thought it time to explore the naval technology of the age, not only in its own right, but as a step towards understanding the ships of 1914-1918 and 1939-1945.

One of these areas is the development of warship armour from the 1850's on, but finding a clear and non-technical summary was not easy until I discovered "Gene Slovers US Navy Pages", which includes, amongst its wealth of information, the USN "Naval Ordnance and Gunnery Manual" of 1937. I am grateful to Jeff J who now runs the site for agreeing to me downloading this resource. My apologies to Jeff for occasionally slipping into the spelling of "armour".

This excerpt is not very technical and there is no mathematics, but there are links to the rest of the manual if you want more. If you vaguely understand that the hardness and toughness of steel depends on the carbon content and the addition of various other elements, such as nickel and chromium, you might find this as interesting as I did.

The technology of those past decades is impressive!

Gordon Smith, Naval-History.Net




Wrought Iron - Steel - Compound - Nickel-steel - Harvey - Krupp - Manufacture of Krupp cemented

The first publicly recorded proposal to sheath the hulls of naval vessels with a metal shield was made by Sir William Congreve in The Times in London on February 20, 1805. A similar proposal was later made by John Stevens, of Hoboken, New Jersey, in 1812. Nothing came of these proposals for years, but in spite of such lack of encouragement, John Stevens and his sons undertook a series of experiments by which they determined the laws of penetration of iron plates by cannon balls and the maximum thickness of iron plate necessary to defeat any known gun.

In 1842, Robert L. Stevens, one of John Stevens’ sons, presented the results of these experiments, and a new design of a floating battery, to a committee of Congress. These experiments of Stevens aroused great interest both in America and Europe. In 1814 a Frenchman, General Paixhan, had also pointed out the necessity of armoring ships, and in 1845 Dupuy de Lome designed an armored frigate for the French government. These proposals resulted in the laying of the keel of the armored Stevens Battery in Hoboken in the spring of 1854, followed in a few months by four keels in Toulon*, and but a few months later by three more in England*. One of the French floating batteries was fittingly named the Congreve. During the following year occurred the first engagement in which armored vessels participated - the bombardment of the Kinburn forts in the Crimean War by three of the French batteries.

*Note: believed to refer to the three French broadside ironclads of the Gloire-class - Gloire, Invincible and Normandie laid down in 1858, and four British ironclads of two classes in 1859 - Warrior and Black Prince, Defence and Resistance. Warrior has now been restored and can be seen in Portsmouth, England

French Gloire (Photo Ships)

British Warrior (Photo Ships)

Wrought Iron Armor

The only metal, practicable and available in table quantity, at this period, was iron, wrought or cast, and all experiments showed that the wrought iron was far superior, pound for pound, to the cast iron, in defeating projectiles. Wrought iron was therefore adopted for marine use, and these first ironclads were protected by solid plates between 4 and 5 inches thick backed by 36 inches of solid wood timbers.

The most costly experiments were carried out, especially in Europe, where the iron industry was most highly developed, to improve the resisting power of wrought iron armor. Tests were made of laminated plates, with the laminations in contact, and with wooden timbers between, but in all cases, single plates gave greater resistance per pound of protection.

During the Civil War most of the armor employed on American vessels was laminated wrought iron, but this was necessitated by lack of facilities for manufacturing single plates of proper thickness, rather than by any superiority for that type of armor.

Proposals for the production of face-hardened armor were early made. At first these proposals were to face wrought plates with cast or chilled iron plates, but as these schemes involved the same loss of efficiency as was exhibited by laminated plates, and because of the insecurity of the bond between the plates, the hard plate failed to secure the full support of the tough back; they all failed to compare favorably with single iron plates. But as far back as 1863, a Mr. Cotchette proposed, in England, the welding of a 1-inch blister steel plate to a 3-inch wrought iron plate; and later, in 1867, Jacob Reese of Pittsburgh, Pa., patented a cementation compound which he stated could be used for cementing and hardening armor plate.

Efforts to carry out these proposals failed for many reasons, primarily because the general development of metallurgy was not equal to the task. It should be remembered that the Bessemer process of making steel in a converter was developed between 1855 and 1860 and that the Siemens-Martin process of making steel in an acid open hearth was developed in France and England a few years later, each process being brought to the United States several years after its European development.

Cast iron has been used for armor in land fortifications, where weight is of slight importance, but as stated above, it has never been applied to naval vessels. The most prominent example of cast iron armor is the famous Gruson turret. These turrets were made of large iron castings, the exterior surfaces of which were chilled by heavy iron chills in the moulds, and were of an oblate-spheroidal shape. Due partly to their shape, but also to the fine quality of the iron and their great thickness, these turrets were considered of great value and were used extensively in protecting European frontiers. The first Gruson turret was tested in 1868 by the Prussian government.

Steel Armor.

In 1876 gun power and projectile quality had so increased that about 22 inches of iron was necessary to accomplish the defeat of a projectile from the heaviest cannon, but in that year occurred, at Spezia, a trial which revolutionized armor manufacture and permitted a reduction in thickness. In these trials a 22-inch mild steel, oil-quenched plate manufactured by the great French firm of Schneider et Cie. completely outclassed all its iron competitors. This plate is reputed to have contained about .45 per cent carbon and to have been hammered down to the required thickness from an ingot about seven feet high. The process of manufacture was kept secret.

This steel plate, while possessing superior ballistic resistance, was more prone to breaking up and this difficulty led to the next real development, which logically resulted from efforts to combine the hardness of steel in the face of a plate with the toughness of iron in its back. The steel used in these plates was made in Siemens-Martin open-hearth Furnaces.

Compound Armor.

Thus resulted a new type of armor - the compound type - the two principal examples of which were the Wilson Cammel compound plate in which an open-hearth steel face was cast on top of a hot wrought iron back plate, and the Ellis-Brown compound plate in which a steel face plate was cemented to an iron back plate by pouring molten Bessemer steel between them.

In both these processes, which were English, the plates were rolled after compounding. For the next ten years there was no especial development in armor manufacture other than minor improvements in the technique of manufacture, and great competition and controversy existed as to the relative quality of all-steel and compound armor. The all-steel armor was a simple steel of about .30 per cent to .40 per cent carbon, while the steel face of the compound armor contained between .50 per cent and .60 per cent carbon. These two classes of armor, their comparative value depending largely on the skill with which they were made, were approximately 25 per cent superior to their wrought iron predecessor, that is to say a 10-inch all-steel or compound plate would resist the same striking energy that a 12.6-inch iron plate would withstand.

Nickel-Steel Armor.

The next step in advance occurred about 1889 when Schneider introduced nickel into all-steel armor, and with the advent of nickel-steel armor began the complete elimination of compound armor. The nickel greatly increased the strength and toughness of steel. The amount of nickel in the first few examples of nickel-steel armor varied between 2 per cent and 5 per cent but finally settled down to about 4 per cent. At about this same time oil and water quenching were successfully applied to armor plates by Schneider.  After forging under a hammer, and annealing, the plate was heated to a tempering heat and its face was then dipped for a short depth in oil, this tempering being followed by a low temperature anneal.

These improvements resulted in a further increase of about 5 per cent in the resistance of armor; that is to say, a 10-inch nickel-steel treated plate equaled about 13 inches of iron.

It was at this stage of development that the manufacture of armor was undertaken in America by the Bethlehem Iron Company, under the supervision of Mr. John Fritz, and shortly afterward by the Carnegie Steel Company, under Schneider patents. The first deliveries of armor for the old Texas, Maine, Oregon, and other ships of that period consisted of heat-treated nickel steel, containing about .20 per cent carbon, .75 per cent manganese, .025 per cent phosphorus and sulphur, and 3.25 per cent nickel.

Harvey Armor.

In 1890 the next great improvement was begun by the introduction of the Harvey process which was first applied to armor when a Creusot 10.5-inch steel plate was Harveyized at the Washington Navy Yard. This process, the invention of H. A. Harvey of Newark, N. J., consisted in carburizing (cementing) the face of a steel plate by heating it and holding it to a very high temperature (about that of molten cast iron) for from two to three weeks, with the face to be hardened in intimate contact with a bone charcoal or other carbonaceous compound.

The result of this treatment was to raise the carbon content of the face to between 1 per cent and 1.10 per cent, with a gradual reduction in carbon content beneath the surface until the effect of the carburization vanished at a depth of about 1 inch. Later the plate was oil quenched and then water quenched, both operations at a uniform temperature throughout the plate, the result being that the super-carburized face assumed a very hard condition, while the back of the plate was toughened. In other words, the face of the plate was super-hardened because of its higher carbon content.

In 1887 Tressider patented, in England, a method of improving the chilling of the heated surface of a plate by forcing against it, under considerable pressure, a dense spray of fine streams of water. This scheme improved on the previous dipping because it kept fresh cool water against the heated surface, thus facilitating the extraction of heat by eliminating the retarding influence of the layer of steam which would otherwise have been formed. This water spraying was now combined with the Harvey process and we had the nickel steel, cemented, oil-tempered, water-sprayed, face-hardened armor known as Harveyized armor and sometimes simply Harvey armor.

A typical chemical analysis of the Harvey armor of this period shows the carbon content to have been about .20 per cent, manganese about .60 per cent and nickel about 3.25 per cent to 3.50 per cent.

Shortly after the adoption of the Harvey process it was shown that the ballistic quality of a plate could be improved by reforging after cementation. This forging, giving a reduction of from 10 to 15 per cent in thickness, was conducted at a low temperature. It was first adopted because it gave more precise regulation of thickness, improvement of surface finish, and some refining of the structure in advance of heat treatment. This process was patented by Mr. Corey of the Carnegie Steel Company, under the name “double forging.”

Harveyized armor immediately established its superiority over all other types. The improvement amounted to another 15 per cent to 20 per cent increase in resistance, 13 inches of Harvey armor equaling about 15.5 inches of nickel-steel armor.

Krupp Armor.

During the eighties, another alloying element, chromium, had been introduced into small crucible heats of steel, and the resulting alloy was found, when properly heat treated, to possess great hardness. Steel makers, in spite of persistent efforts, failed to produce large nickel-chrome steel ingots, or to properly forge and treat them when produced, until the great German works, Krupp, solved the problem about 1893.

Krupp also adopted the cementation process for armor, but instead of using solid hydrocarbon as in the Harvey process, used a gaseous hydrocarbon, illuminating gas being passed while hot across the face of the heated plate. This gaseous cementation has been frequently used, but has been gradually superseded by the use of solid hydrocarbon. Gaseous cementation was used at Bethlehem in 1898 but has since been abandoned and is not now used on American armor plates.

At about the same time Krupp developed a process of deepening the hardening on one side of a cemented steel plate. To do this, the plate was imbedded in clay or loam, with the cemented side exposed, and then the exposed face was subjected to a very hot and quick heat. As the heat penetrated gradually, the exposed face became much hotter than the back, thus permitting “decremental hardening” by water spraying. A piece of steel heated above a certain temperature will become very hard if quenched in water, while its physical properties will be little affected if it is quenched when below that temperature. For the sake of convenience, let us call this certain temperature a critical temperature.  Now as the face of the plate is heated above this critical temperature there will always be a plane in the plate at the critical temperature, and as the heating is continued this plane of critical temperature will gradually travel inward toward the back, eventually reaching the back if the heating is continued long enough.

However, the plane of critical temperature was only allowed to sink between 30 per cent and 40 per cent of the thickness, and when that position was reached, the plate was hurriedly withdrawn from the furnace, put in a spraying pit and subjected to a powerful spray of water, at first on the superheated side and a moment later on both sides, the double spraying being done to prevent, as much as possible, the warping which a spray on but one side would produce.

This process, called decremental face hardening, produces a very hard face, between 30 per cent to 40 per cent of the plate’s thickness, and at the same time leaves the other 60 per cent to 70 per cent of the thickness in its original tough condition. It should be specifically noted that this method of hardening depends on the decremental heating and does not necessarily involve any variations in carbon content. In other words, in this type of face hardening, the front portion of the plate is super-hardened because of its higher temperature, the depth of the hardening being subject to regulation, and greater than the depth of cementation, if desired.

The process of face hardening, being the final treatment, was, of course, applied after the plate had been heat treated. The latter refined the existing coarseness of the grain and produced a fiber in the steel increasing its strength and ductility.

The success of the Krupp process was immediate, and all armor manufacturers soon adopted it. In all plates thicker than about 5 inches, the Krupp armor was about 15 per cent more efficient than its immediate predecessor, Harveyized armor, 11.9 inches of Krupp being about equal to 13 inches of Harveyized armor. The Krupp process was applied to armor for American vessels in 1900. Most of the armor made for the past 25 years has been Krupp cemented armor. (as of 1937)

During the past 15 years, various slight improvements have been made in the technique of manufacture; and it is, as now made, possibly 10 per cent better, ballistically, than it was during its early use.

Manufacture of Krupp Cemented Armor (K.C.)

Carbon being the principal hardening element, the natural tendency is to carry the carbon content as high as possible. The higher the carbon, however, the more difficult becomes manufacture, tears appear in the forging, fibering the plate becomes more difficult, and the plate becomes brittle, making it liable to cracking and excessive spalling (detaching of surface fragments) on ballistic test. The addition of nickel increases the toughness of the plate, and permits it, when properly treated, to be fibered; while the chromium adds hardness without the extreme brittleness which would accompany this hardness if produced solely by carbon. The chromium also renders the steel particularly sensitive to heat treatment and thus facilitates the final decremental water hardening.

A typical chemical analysis of a modern Krupp cemented plate is as follows:

Carbon .35
Nickel 3.90
Chrome 2.00
Manganese .35
Silicon .07
Phosphorus .025
Sulphur .020

In this connection it is interesting to note that when K. C. armor was first introduced into America, the plates ran about .27 per cent carbon, 3.75 per cent nickel, and 1.75 per cent chromium. The increase in carbon and chromium and the increased resistance of modern K. C. armor is indicative of the improvement in metallurgical skill.

A modern process of manufacture may be briefly summarized as follows:

(1) The charge of pig iron and ore or pig iron and scrap is melted in a basic open-hearth furnace and is then poured into an iron or sand mould.

The dimensions of an ingot are varied to suit conditions, for instance, an ingot for a three-gun turret port plate is about 42” X 150” X 250” and weighs about 425,000 pounds, while one for a belt plate is about 26” X 132” X 200” and weighs about 200,000 pounds.

(2) The ingot, while still hot, is stripped from the mould, cleaned, and prepared for forging (Plate I)

Plate I - Armor ingot, weight about 400,000 pounds being carried on two cranes to be forged
Plate II - Forging a "belt plate" under a 10,000 tons press

(3) The ingot is then reheated and forged under a hydraulic press to within about 15 per cent of the final thickness. The forging reduces the ingot to about one-third of its previous thickness. The segregation of impurities in the central upper portion is discarded by cutting off (Plate II).

(4) The forging is annealed to produce a partially fibrous condition of the microstructure to prevent cracking in cooling and to eliminate the strains due to forging.

(5) It is then super-carburized (Plate III). The time required for this step varies with the size of the forging; large forgings take from 10 to 14 days.

(6) Reheating, reforging nearly to the final thickness, and annealing follow.

(7) Several heat treatments may follow to develop the proper physical properties in the fiber (Plate IV).

Plate III - Three plates loaded on furnace bottom ready to run in for "carburizing"

Plate IV - Coupon (or sample) from 10-inch plate. Broken cold under press
(note 10 inch scale on right)

(8) The forging is next machined to the rough dimensions.

(9) It is then reheated and formed to shape.

(10) The front face is then heated above the critical temperature, depending upon the depth of chill desired, and hardened by oil or water spraying (Plate V).

Plate V - A side armor, or "belt" plate,
on the spray after final water hardening

(11) After low heating, the curvature of the plate is rectified.

(12) The plate is machined to the finished dimensions.

Krupp Non-Cemented Armor.

From the discussion of decremental hardening as applied to K. C. armor, it is quite apparent that a plate may be hardened without prior cementation. It should be mentioned here that the carburized face is much more liable to tearing and cracking during forging and bending to shape than is the other portion of the plate, a condition which renders the fabrication of thin plates more difficult than thick ones.

These facts led the Bethlehem Steel Company to undertake the manufacture of armor by the general Krupp process without cementation. Later the Midvale Steel Company adopted the same idea. Such armor is generally referred to as Krupp non-cemented or K. N. C. armor. In structure this armor differs appreciably from K. C. armor. For instance, there is no super-carburized face, and the chill itself is generally harder and somewhat deeper; and there is a further difference in chemical composition, in that, while the carbon and chrome may be somewhat higher, the nickel may be equal or lower, as compared to K. C. armor.

Non-cemented armor is fully equal, ballistically, to K. C. armor, and can, in fact, be made of superior ballistic resistance, but it has an unfortunate tendency to spalling, both on projectile impact and from internal strain. It was this tendency to spalling which led to the abandonment of the process after a few years of use. A typical analysis shows carbon as high as .50 per cent, with nickel at 3.5 per cent and chromium at 2.30 per cent to 2.50 per cent.

Summary of Armor Development

It will be seen from the preceding review that each change in armor has added something, and that modern armor contains all the essentials of each successive product. First, for marine use, we had the simple wrought-iron armor, which was later developed into compound iron-steel armor. Then all-steel armor displaced the compound armor, and was, in turn, improved by the addition of nickel. Next we have a return to the hard face principle, but with homogeneous structure, in the application of Harveyizing. Finally we have the introduction of chromium and the development of decremental hardening as applied to both cemented and non-cemented plates.

The manufacture of efficient armor requires not only a high quality of metallurgical skill, but also expensive tools and equipment. The various improvements in its quality and manufacture have, therefore, been as closely related to the inventions and discoveries of metallurgy as to the commercial growth of the general steel industry. A knowledge of the nature and dates of the various stages in its development, therefore, gives the ordnance student an excellent picture of the evolution of the use of iron and steel. Frequently the demands of the ordnance engineer have caused developments in metallurgy which have had far reaching commercial application, but more frequently the general improvements made by the metallurgical engineer and chemist have opened new vistas to the designers of ordnance.


The general term Class A armor is applied, in the American Navy, to all face-hardened armor, whether Krupp cemented or Krupp non-cemented.

Ballistic Test.

During manufacture the processes are watched to insure uniformity in procedure and product. Chemical analyses are made to insure homogeneity in the metal, and test specimens are taken out of the ends of the plates to determine the uniformity of the physical properties. Finally, after water hardening, fragments called coupons are broken off diagonally opposite corners, in order that a microscopical examination may be made of the internal structure. (See Plate IV.)

Plates are arranged in “groups,” each containing from 600 to 1,200 tons, and after water hardening a plate is selected from the group, sent to the Proving Ground, and there tested ballistically. If successful, the group is passed to completion; if unsuccessful, the group is rejected for retreatment. The ballistic test consists, at present, in subjecting the plate to the normal impact of two major caliber A. P. projectiles at specified velocities, the velocity depending on the gauge of the plate and size of the projectile. This relation will be discussed under the heading of “penetration.”

In testing and experimenting with armor, great energies must be absorbed. For instance, a 16-inch projectile striking normally at 2,000 f. s. delivers about 60,000 foot-tons. Assuming that the plate is 10 ft. X 20 ft. X 15 in., that gives about 300 foot-tons per square foot of plate area, or about 1,100 foot-tons per ton of plate. The structure to which the plate is attached must, therefore, be of great strength. Such structures are called plate-butts. Similar structures used for testing the action of projectiles against plates are called projectile-butts.

Particular attention must be given in conducting all impact tests to eliminate, so far as is possible, all movement of the attacked plate, or, in other words, to preserve the rigidity of the structure as a whole. This is particularly true when plates are attacked at other angles than normal, for in such cases there may be most powerful end thrusts, which, if uncontrolled, will permit end movement and therefore subject the projectile to uncontrolled and unknown “whip,” thereby vitiating the value of the test. Plate VI shows an armor plate after undergoing test.

Plate VI - A Ballistic Test Plate

The two shots required for the routine ballistic test are for the purpose of determining whether the manufacturer’s current manufacturing methods, in general, and the group represented, in particular, are up to the required standard. That condition having been established, and the group’s acceptability settled, it is customary to attack the plate with one or two additional shots to determine its precise ballistic limit. To determine this ballistic limit the velocity of impact is increased until penetration is just or nearly secured. Experienced personnel can, with ordinary armor, generally secure this result (after two ballistic test shots) in two more shots, and sometimes even with one shot.

In experimenting with and testing armor it is the ballistic limit which is the significant figure. How it is expressed will be shown later. But it is readily seen that by always working for it, a record can easily be compiled, in time, which will show not only the general average of performance, but also the peak of performance. And by always carefully investigating the plates which give the greatest resistance, the general standards may be raised.


Considerable discussion took place in early days of ship armor as to whether armor could best be defeated by “racking” or “punching.” Racking was produced by very large solid ball shot at low velocity and resulted in knocking plates off the ship’s side, thus exposing her vitals; and punching was produced by elongated projectiles, at high velocity, and resulted in perforation and the immediate attack of the ship’s vitals. This discussion was settled during our Civil War very emphatically, and later in the Huascar-Cochrane and Blanco Encalado engagement; for the shots which secured decisive effects, practically without exception, were piercing shots. This discussion finds its echo in more recent controversies concerning the relative merits of armor-piercing and high-capacity projectiles. Here, again, war decided emphatically, for in the World War it was conclusively shown that the decisive hits were those made by armor-piercing projectiles.

The principal function of armor is, therefore, to prevent, so far as possible, the penetration of a projectile into a ship. But the “racking” effect cannot be neglected. This is at present, primarily a matter of securing; but the use of guns with projectiles of great weight, the demand for increased speed and consequent sacrifice in protection, and the probable obliquity of impact, are gradually forcing attention back to the “racking” or “smashing” attack.

The rest of this Section and the remainder of the Chapter, which covers US Navy Class B and Light Armour,
but is mathematical and more technical, can be found here,

together with

The complete manual - Naval Ordnance 1937 - an excellent introduction to all aspects of naval gunnery at this time

return to "Pax Britannica" 1815-1914
or to Naval-History.Net


revised 27/9/15