A description is given of the experimental technique devised to apply the method outlined theoretically in part I to the measurement of the dynamic compressive yield strength of various steels, duralumin, copper, lead, iron and silver. A polished piece of armour steel was employed as a target, and cylindrical specimens were fired at it at various measured velocities from Service weapons. The distance between the weapon and target was made short to ensure normal impact, and apparatus was devised for the precise measurement of striking velocity over this short range. The dynamic compressive yield strength was computed from the density of the specimen, the striking velocity, and from measurements of the dimensions of the test piece before and after test. Details are given of the accuracy of the various measurements, and of their effect on the values of yield strength. The method was found to be inaccurate at low and high velocities. For instance, with mild steel, satisfactory results were only obtainable within the range 400 to 2500 ft. /sec. The range of velocities within which satisfactory results could be obtained varied with the quality of the material tested, soft metals giving results within a much lower range than that necessary for harder materials. Because of its failure at low velocities, the method could not be employed to bridge the gap between static and dynamic tests. The rate of strain employed in the dynamic tests could not be measured, but was estimated to be of the order of 10,000 in. /in. /sec. With the materials tested little change of dynamic strength occurred within the range of striking velocities employed, probably because the rate of strain did not vary to any great extent with the striking velocity. Within the range of weapons available, that is, from a 0·303 in. rifle up to a 13 pdr. gun (calibre 3·12 in.), little change of dynamic strength occurred with alteration of the initial dimensions of the specimens, probably because the corresponding change of rate of strain was not large. In general, the dynamic compressive yield strength
S
was greater than the static strength
Y
represented by the compressive stress giving 0·2% permanent strain. For steels of various types, regardless of chemical composition and heat treatment, there was a relation between
S
/
Y
and the static strength
Y
, the ratio decreasing from approximately 3 when
Y
was 20 tons/sq. in. to 1 when
Y
was 120 tons/sq. in. A similar relation occurred with duralumin,
S
/
Y
varying from 2·5 at
Y
= 8 tons/sq. in. to 1·4 at
Y
= 25 tons/sq. in. Dynamic compressive yield values were obtained for soft materials such as pure lead, copper and Armco iron, which, under static conditions, gave no definite yield values. A plot of the unstrained length of the specimen
X
, expressed as
X
/
L
(where
L
= initial overall length), versus the final overall length
L
1
, expressed as
L
1
/
L
, was made for the various materials. Any specified value of
X
/
L
was associated with greater values of
L
1
/
L
for the more ductile materials, such as copper and lead, than for the brittle materials, such as armour plate and duralumin.
The use of flat-ended projectiles for determining dynamic yield stress - II. Tests on various metallic materials