APPENDIX V.

TEST OF EXPERIMENTAL SHRAPNEL FOR 3-INCH GUN.

FRANKFORD ARSENAL,

Philadelphia, Pa., November 24, 1902.

SIR: In compliance with instructions contained in O.O. letter No. 20210, dated November 17, 1902, I have the honor to submit the following report upon the experiments and investigations made at this arsenal to determine the best form of construction for shrapnel. These experiments have been of a brilliant character, and have resulted in the production of a shrapnel embodying many original features and far superior, it is thought, to any in use in this or any foreign service. The scope and results of these investigations are clearly described in the following extracts from a report of Capt. B. W. Dunn, Ordnance Department, to whom belongs the credit of inaugurating these experiments and of carrying them to a satisfactory conclusion:

Excepting the Hotchkiss Company, in France, which still manufactures a headcharge shrapnel with cast-iron separators, the universal preference abroad is for the base-charge shrapnel with a resin matrix to protect the lead balls from deformation by the shock of discharge. In England the bursting charge is small, 1.6 ounces, and the balls and matrix are assembled in a perforated tin cylinder placed in the drawnsteel case. In Germany the latest models contain about 3 ounces of powder, no lining for the case is used, and with the resin matrix is mixed a smoke-producing compound. All steel cases are drawn with increased thickness of walls around the powder chamber, which terminates in a shoulder on which rests an arched steel diaphragm to support the balls and protect the powder charge.

Our present service 3-inch shrapnel contains two features thought to be in advance of the latest European model-steel-jacketed balls and a steel case with hexagonal interior cross section.

The principal objects in view in the tests herein reported were the relative advantages of (1) head and base charge, and (2) lead balls with resin matrix in a circular case or steel-jacketed balls without matrix in a hexagonal case.

HEAD AND BASE CHARGE.

From reliable information of the results obtained abroad the preference for a base charge was pronounced at this arsenal before undertaking these tests.

The principal disadvantage of the head charge is that the shrapnel case is required to meet two conflicting conditions; it must be strong while in the gun to prevent prematures, and it must be weak during flight to insure the complete opening of the case and the liberation of all the balls at the point of burst.

All head-charge shrapnel on the separator plan manufactured in this country have failed, more or less, to meet one or the other of these conditions. The difficulty is very much increased when the shrapnel is required for fixed ammunition. In this case a support for the band other than that given by the walls of the case is necessary, and the solid separators used for the purpose are so firmly locked in the case by band pressure that they are liable to imprison all balls and separators below them.

Our service 3-inch shrapnel has met the two conditions in the field gun, but it fails to meet them in the 3-inch R. F. gun. The tangential stress on the case, due to rotation of the shrapnel, is for the latter gun more than twice that of the former.

Another important disadvantage of the head-charge shrapnel is that the terminal velocity of the bullets is decreased by the bursting charge.

The natural angle of the cone of dispersion of the bullets, as fixed by the velocities of rotation and translation, is considerably increased by the head charge.

This was thought at one time to be an advantage, since the shrapnel is generally used against troops in open order. The general increase in the velocity of shrapnel and the unavoidable variations in the burning of time fuses, however, make it impossible to accurately control the point of burst; and, assuming that for a given range the shrapnel is liable to burst anywhere within a space of 150 yards in length, it is found that the best average results are obtained with small cones of dispersion.

The principal advantages of the base-charge form are: (1) The case is required to be strong both in the gun and at the point of burst, where it is desired that the case remain intact in order to furnish a path for the powder gases to work over and increase the terminal velocity of the bullets; (2) the diaphragm separating the bullets from the powder gives the band support required for fixed ammunition; (3) the cone of dispersion is decreased; (4) the velocity of the bullets is increased about 200 feet per second, the difference in this velocity for head and base charge being in some cases more than 250 feet per second.

The disadvantage has been in securing a light diaphragm that will support the balls on discharge. Diaphragms of the thickness used in English shrapnel, and made of the best special grades of steel that the Midvale and Bethlehem steel companies could furnish, were fired, and on recovery from the sand bank were found to be dished downward.

This has been overcome here by the novel arrangement to be described later.

RESIN MATRIX AND STEEL JACKETS FOR BULLETS.

The resin matrix and the steel jackets perform the same functions on dischargethey protect the lead balls from deformation. The matrix, however, requires the assistance of the tangential strength of the case to prevent upsetting, while the steel jackets do not.

The resin is much cheaper, its cost being comparatively nothing, while that of the jackets is about 25 cents per shrapnel.

The cone of dispersion is slightly less-i. e., better-with the matrix the resin tending to hold the bullets together and preventing the free entrance of powder gases into the spaces between the bullets to spread them laterally.

From the fact that the difference between the number of bullets counted on the pattern target and the total number in the shrapnel is greater with the matrix than with the jackets, it is concluded that the matrix prevents a complete separation of

the bullets.

The data presented by the pattern tests submitted herewith show that the penetration of the lead bullets used with matrix is less than that of the steel-jacketed bullets.

An important disadvantage of the resin matrix is its strong adhesion to the steel case, which offers a resistance of 6,000 to 8,000 pounds to longitudinal movement, an nitial resistance to the powder charge which, as the experiments here have shown, will cause rupture of the case. Under the same conditions the case loaded with steel-jacketed balls without matrix does not rupture. When the case ruptures, the increase in velocity of balls is comparatively small, as should be expected.

on.

The initial resistance to motion of the mass of bullets has to be carefully guarded. It is for this reason that the head of the English shrapnel is pinned instead of screwed The increased tamping effect of the atmosphere even may cause a case with full velocity to break when it does not break with a lower velocity. An aluminum head can be screwed on where a cast-iron or steel head of the same dimensions would have to be pinned to prevent rupture of case. The larger the powder charge the more tendency to rupture. A small charge can thus produce a better effect than a larger one. This fact is assumed to have influenced the English, who use about one-half the charge used in German shrapnel.

It is possible that the Germans may have overcome the liability of the case to break when a large bursting charge is used in connection with a resin matrix. If so, the advantage of the steel jackets will be reduced to the one item of increased penetration, which is not considered of sufficient value to justify the extra cost. Without steel-jacketed balls, the hexagonal steel case, which is slightly more expensive than the round one, will not be needed, the adhesion of resin to the balls and to the case being sufficient to prevent any relative rotation between the two.

A recommendation, then, as to the continued use of steel jackets and the hexagonal case is withheld for a final report after tests of the Ehrhardt shrapnel.

We have on hand, or contracted for, steel-jacketed balls and hexagon cases sufficient for 25,000 3-inch shrapnel, and recommendations for the model of shrapnel to be made from this material will be made after briefly describing the experimental models fired and the results obtained from them.

DESCRIPTION OF MODELS.

(See accompanying drawings.)

Experimental model No. 1.-This illustrates the present English model in the size of bursting charge and the method of attaching head. They use a heavier ball than ours (35 instead of 41 to the pound) and a heavier case.

Experimental model No. 2a.-In this the bursting charge is increased to 34 ounces, according to the German practice. In all our models using a cast-iron head, the head is pinned instead of screwed on, since experiments in the bombproof showed that even at rest, without any of the tamping effect of atmospheric resistance, the cases with screwed-on heads break up.

No matrix was used in this model.

Experimental models Nos. 2b and 2c.—A resin matrix was added to model No. 2 to determine its influence. For model 2c a resin matrix was used with lead balls.

Experimental model No. 3.-Recognizing the advantage of the base charge, and the impossibility of securing it in the usual way in the cases on hand, which were drawn without a base powder chamber or shoulder for support of a diaphragm, a spider diaphragm steel casting was devised, and to avoid danger of prematures from friction of powder grains over the irregular surfaces of the spider legs the powder was compressed into the six triangular prism spaces, leaving a cylindrical space in the center to hold fine-grain powder to act as an igniting charge.

In compressing the powder tapered pins were placed at the bottoms of the cavities and afterwards drawn out, thus securing an increased igniting surface to overcome what was feared would be an objectionable slowness of burning of the compressed powder. Experimental model No. 4.-The success attending the compression of powder in the spider diaphragm led to the next step.

In model No. 4 the entire charge of 3 ounces is compressed, using the shrapnel case as a mold, and applying a total pressure of 60 tons, or about twice as much as the charge receives in imparting acceleration to the mass of bullets resting on it.

If no relative motion takes place among the particles of the charge there will be no friction or heat developed, and there will be no danger of a premature explosion. No relative motion can take place, because the powder has already been subjected to a pressure twice as great as it receives in service.

After compression all particles of loose powder are removed, and the upper surface of the compressed charge, as well as the adjacent steel surfaces, are covered with a coat of shellac or other protecting paint. A thin-steel disk to distribute the pressure from the balls is all that is now needed to replace the comparatively heavy diaphragm needed to protect the loose-powder charge. The cylindrical space in the center contains the igniting charge of loose powder.

Experimental model No. 5.-This differs from the preceding model only in the head, a light aluminum alloy casting being substituted for the cast iron in order to add another layer of balls. This head can be screwed on, which is a preferable mode of attachment to the pins, without increasing the initial resistance enough to cause the case to break up.

MODELS OF MOUNTAIN SHRAPNEL.

These differ from the models already referred to in the matter of weight.

The drawing of model No. 8 shows the supporting hexagonal ring that is necessary to make our present head-charge shrapnel available for use with fixed ammunition. TABLE NO. 1.-Results of firing tests of 15-pound shrapnel.

When the firing recorded in Table No. 1 was done there were no facilities for measuring the velocity of the bullets after bursting of the shrapnel. This measurement was made in subsequent firings. The average penetration in white pine was obtained by placing in front of the pattern target a number of dressed pine boards, tacked together without spaces between them. The presence of knots in the boards interfered somewhat with the uniformity of the data. The steel jackets were broken and stripped when they struck these knots. All pattern tests are numbered, and the reference numbers in Table No. 1 will permit ready examination of the pattern made by any of the models.

The pattern made when the case does not break up is very regular and characteristic. A clear space in the center contains the three main fragments-head, case, and diaphragm.

The circles on patterns were drawn by hand to cover the space over which the spread of the bullets is practically uniform, and their diameters are taken to measure the effective dispersion. It will be noticed that the two shrapnel containing lead balls in a resin matrix give lower penetration and smaller pattern circles. Of the three compressed charges, one broke up the case, and for this shot the penetration

falls off.

This breaking shows that the initial resistance to motion of bullets, plus atmospheric tamping for full velocity, is almost sufficient to cause even the slow-burning compressed charge to develop a destructive pressure.

There was no tendency of these shrapnel to break when the velocity was reduced to 1,360 feet per second.

Table 2 shows up very plainly the advantage of the base over the head charge, and of the case that does not break up over that which does.

In some cases the average penetration does not check well with the measured velocity of bullets. As in Table 1, the lead balls and resin matrix combination gives the least penetration and dispersion.

All cases containing the resin matrix broke up.

A number of tests were made of 12-pound shrapnel in 3-inch Hotchkiss and 75-mm. Vickers-Maxim mountain guns. They show, as in the preceding tables, the advantage of the base charge in the increased velocity of the bullets.

UTILIZATION OF SHRAPNEL MATERIAL ON HAND.

For some time doubt existed as to the strength of the hexagonal steel cases obtained to execute an order for 25,000 15-pound shrapnel for the 3-inch field gun.