3. 1. NOSE SHAPE EFFECT

In this section, we select only the data of the zero attack-angle from Table 1 in order to see the effect of the nose shape on penetrator dynamics. Figures 8a and 8b show the penetration path-length of three kinds of the cone-nose penetrators with respect to the impact velocity. Although the mass of the models is known to affect the penetration path-1ength, we neglect its effect here because of small variation of the mass in the present experiments. Figures 8a and 8b indicate that the penetration path length increases with the impact velocity, though its effect in the case of the oblique impact (Figure 8b) is less obvious. We do not know at present why the effect of impact velocity on the path-1ength is so weak in the case of oblique impacts. Figures 8a and 8b also show that the penetration path-1ength is not significantly affected by of the cone-nose both for normal and oblique impact. Considering the cross-sectional area of nose-tip for the = 0.3 being three times as large as that for = 0.1, this result is rather remarkable. The weak dependence of on the penetration path-length may indicate that the streamlines of the target soil around the penetrator are not affected by the truncation of the cone-tip of the penetrator. A relevant observation on this fact is that we always find a hard conic soil in front of the truncated nose of the penetrator when we recover the penetrator after the experiment.

Comparison of Figures 8a with 8b shows the effect of impact angle on the penetration path-length. Figure 8b includes the data of impact angles of = 50±3^°. As for the cone-nose penetrators, no significant difference in path length can be seen between normal and oblique impacts; in both cases, the penetration path lengths are ~40 cm at v = 100 m/sec and ~60 cm at v = 1 50 m/sec.

Figures 9a and 9b show the penetration path length of three kinds of the ogive-nose penetrators with respect to the impact velocity. It indicates that the penetration path-length increases with the impact velocity. The dependence of the path-length on the impact velocity is less clear in the case of the normal impact (Figure 9a). This is probably due to the limitation of depth of the sand target. In the present experiments, we used the sand target to a depth of about 70 cm. Although the target sand was made as homogeneous as possible in terms of the hardness, the bottom layer might have been very hard because of the weight of overlying sand and compaction by vibration of the rotation wheel. This inference is also confirmed by the measurement of hardness distribution in the target. As shown in Figure 4, the cone penetrometer data indicates the existence of hard layer as a sudden increase of resistant force at the depth of 55 cm to 65 cm. Therefore the hard layer at the bottom of the target box may have prevented the projectile from penetrating much deeper and it will explain that the path-length in the case of the normal impact at a high-velocity range seems to be saturated at a depth of ~65 cm

On the other hand, the penetrator could not reach the bottom layer even at a high velocity impact ranging from 120 to 170 m/s, in the case of oblique impacts of 6-50^°. Therefore the data shown in Figure 9b should be used to infer the effect of impact velocity on the path-length, rather than Figure 9a.

As is mentioned earlier, the course of the penetrator movement in the target sand deviates from a straight line and the attitude of the penetrator at the rest position differs from the original one. In Figures 10a and 10b, we show the histogram of the degree of inflection in terms of defmed in equation (5). The data in Figure 10 include the data on the cone-nose and the ogive nose. We did not observe any significant effects of the nose shape on this inflection angle, though we observed a truncated-nose penetrator shows smaller inflection than that with non-truncated cone-nose penetrators [8] . The data for the oblique impact evidently have a peak between 5^° and 10^° , while the data for normal incidence are distributed around zero. This slight difference may reflect the effect of a torque applied on the nose tip just after initial contact at the oblique impact. But the absolute values of inflection angle are almost within 10^° for both normal incidence and oblique impact. This suggests that the oblique impact does not have a significant effect on penetration trajectory.