- Email: [email protected]
The initiation of solid explosives by shock waves
842
FORMATION, COMBUSTION, EXPLOSION AND DETONATION OF SOLIDS
free surface area decreases and the rate of decomposition falls. Further work on silver cyanamide is summarized elsewhere. 14 REFERENCES 1. GARNER, W. E. (ed.): Chemistry of the Solid State, Butterworths Scientific Publications, London, 1955. 2. BARTLETT, B. E., TOMPKINS, F. C., AND YOUNG, D. A.: Proe. Roy. Soe. (London), A2~6, 206 (1958). 3. GRAY, P., AND WADDINGTON, T. C.: Proc. Roy. Soc. (London), A241, 110 (1957). 4. McLAREN, A. C., AND RODGERS, G. T.: Proc. Roy. Soc. (London), A246, 250 (1958). 5. DEB, S. K., EVANS, B. L., AND YOFFE, A. D.: Nature, (London), 180, 294 (1957). 6. Discussion on the initiation and growth of explosion in solids, Proc. Roy. Soc. (London), A2.~6, 145 (1958).
7. COOK, G. B.: Sixth Symposium (International) on Combustion, Reinhold Publishing Corporation, New York, 1956. 8. DES, S. K., AND YOFFE, A. D.: Trans. Faraday Sou., 55, 106 (1959). 9. CAMP, M., DEB, S. K., EVANS, B. L., MONTAGu-POLLOCK, H. M., AND YOFFE, A. D.: Fourth International Symposium on the Reactivity of Solids, Elsevier, Delft, 1961. 1O. BOWDEN, F. P., AND YOFFE, A. D.: Fast Reactions in Solids, Butterworths Scientific Publications, London, 1958. 11. McAuSLAN, J. H.: Ph.D. Thesis, Cambridge, England, 1957. 12. PASHLEY, D. W.: Advances in Physics, 5, 173 (1956). 13. MITCHELL,J. W., ANDMOTT, N. F.: Phil. Mag., 2, 1149 (1957). 14. BOWDENF. P., ANn MONTAGe-PoLLOCK,H. M.: Nature (London), 191,556 (1961).
92 THE INITIATION OF SOLID EXPLOSIVES BY SHOCK WAVES By C. H. JOHANSSON, N. LUNDBORG AND T. SJOLIN Introduction When investigating the propagation of detonation among solid explosives through plates of inert material, the process has, in most cases, been recorded at the surface of the charges. 1-6 It is, however, of great interest to extend the investigation to the interior of the charges 7-1°and observe step by step how detonation begins and spreads through the secondary charge. For this purpose we have recorded the light emission from cylindrical charges of compressed TNT, 1.55 g/cm ~ density with radial drill holed and semicylindrical acceptor charges coaxial with the cylindrical donor charges. The results indicate that the process in axial planes of undivided cylindrical acceptors is only slightly affected if the charge is cut along an axial plane and one of the halves is removed, and the cut surface is covered with a Plexiglass plate or the charges are submerged in water.
Cylindrieal Charges w i t h Radial Drill Holes The donor and acceptor charges were cylinders 21 mm in diam, and 80 mm in length with 2 mm
radial drill holes with a 10 mm distance between the centers. The plate between the charges was 50 x 50 mm plate of hard-rolled aluminum (>99 per cent A1). Detonation propagation was obtained in all instances with a plate thickness of 20 mm, but in no instance with a thickness of 22 mm. Figure 1 shows the arrangement (left) and the result (right) of streak camera photos in two tests with a 20 mm plate. The drill holes drawn with solid lines and the plain circles of the diagram refer to the one test (No. 1); the holes drawn with dashed lines and the circles with crosses refer to the other. All drill holes except the two nearest the plate in the acceptor gave intense light spots characteristic of a passing detonation front. The hole adjacent to the A1 plate on the acceptor side gave a faint luminous trace compatible with the assumption that it was caused by the shock wave. The trace of light from the hole 5 mm from the A1 plate in the acceptor is slightly narrower than the other, but has the same intensity and shows that the detonation has commenced. In a control test substituting a similar cylinder of sulphur for the acceptor and with
INITIATION OF SOLID EXPLOSIVES BY SHOCK WAVES
843
eptor cm
EEC.
-MO
V__:_-
120 E~_-cu.-
Dekonofionfronh
I00
/
surfacy ShockwovefrontJ (
80
AI-plofe
/
I
/ / /
J,~ [ •
-
-
50 40 /
/
ox/$ surfoce
i--4__
l _ _
•
i--'---"
Donor
22/1111 0
I I I I I I I /0
20/us
FIG. 1. Arrangement of charges with bore holes and diagram according to two streak camera photographs. the holes placed as in test No. 1, no traces of light were obtained from the holes. According to the diagram in Figure 1, the shock wave in the acceptor becomes a detonation wave ca. 5 mm from the end face at the axis and thereafter moves forward with a normal detonation velocity. This is in agreement with the results which Marlow and Skidmore 7 obtaiued when they measured the front velocities at the axis with "single wire" and "twin wire" probes set in radial drill holes. The light emission at the envelope surface shows that the detonation does not move backward but leaves the end of the charge behind.
Semicylindrical Acceptors Figure 2 shows the arrangement of the charges and the screen with parallel slits in front of the axial plane of the semicylindrical acceptor. The
charges were submerged in an aquarium with water and placed in front of a streak camera with a rotating mirror• Figure 3 reproduces two recomings with a cylindrical donor 21 mm in diam and 80 mm in length, a semicylindrical aeceptor 63 mm in diam and 70 mm in length, and an intermediate cylinder of AI alloy 21 mm in diam and 18 mm in length. In Figure 3a, the axial plane of the acceptor had a coating of black varnish in which 11 slits were scratched at distances of 3 ram. In Figure 3b, the screen consisted of a brass plate which was placed 50 mm in front of the charge surface. The width of the slits was 0.5 mm and the distance between the slits 3 mm. In this case the first slit happened to be 3 mm to the right of the center and s l i t N o . 11 was outside the charge and received no light. By inserting a still picture of the slits in the pictures of Figure 3 it is easy to construct the
844
FORMATION, COMBUSTION, EXPLOSION AND DETONATION OF SOLIDS
t---
-~- Screen with s/its ------~
Camera
Streak camera pictures of the envelope surface of undivided cylindrical acceptors with diameters of 21, 40 and 63 mm respectively, show satisfactory agreement with the pictures of the front in Figure 4 (cf. Fig. 1). Streak camera pictures of the detonation front surface at the upper end of the semicylindrieal acceptor taken with 7 slits perpendicular to the axial plane show a small downward bend in the detonation front at the axial plane. I t is, however, very slight and probably too small to be of practical significance for the spread of the detonation front.
o
Incipience
---z.-•
Wo/er ~
Camera
--=----- , ~ ¢ / ~ e n
--
Accepfor
-- -_---A / - c q / i n d e r
_~-~D o n o / "
and
Spread of the Detonation
When the donor detonates, the upper surface of the A1 cylinder is suddenly given a velocity and generates a shock wave in the aceeptor, just as a projectile would. Our measurements show that the velocity necessary for initiation is about 650 m/sec in both cases. In a semieylindrical acceptor with a Plexiglass cover, the shock wave produces a faintly visible luminous front line in the axial plane. When the shock wave has proceeded a certain distance it changes over to detonation which spreads from an approximately plane, circular area upwards and sideways with ordinary velocity. The detonation front is approximately in the shape of an ellipsoid with the equation r2
(Dr + ro) 2 :FIG. 2. Semicylindrical acceptor arranged for streak camera photographing with 11 slits. luminous detonation front at various times. In Figure 4 is reproduced a pile of instantaneous pictures of the front with 0.39 t~sec interval obtained from Figure 3a. They are in accordance with the instantaneous photographs of the detonation front in Figure 5. In this case the charges were placed in air and to retard the expansion of detonation gases perpendicular to the axial plane and eliminate disturbing light from the air, a 10-ram thick Plexiglass plate was glued on to the surface. In addition the acceptor was not semicylindrical but a 15-mm thick slab of compressed TNT. These differences in the experimental conditions in comparison with Figure 2, are in all probability not essential to the propagation of the detonation front.
+
(z -- z0) 2
(Dt) 2
-- 1.
(1)
The z-axis coincides with the axis of the acceptor charge, the origin lies in the end surface of the acceptor, z0 is the initiation depth, r is the distance from the axis, r0 the radius of the initial circular surface, D the detonation velocity and t the time from the start of the initiation. In Figure 4, D = 6.65 mm/t~see, z0 = 8 ram, r0 = 5 mm and when the detonation starts the z-value of the upper surface of the A1 cylinder is about 1.7 mm. The detonation front is not visible below the line al ba~. We have raised the question if the chemical reaction started by the shock wave has caused a light-absorption similar to the phenomena observed by Johansson and Sternhoff 8 when a detonating charge of T N T in water is exposed to high pressure. However, experiments with undivided cylindrical charges according to Figure 1 show that the detonation does not spread in this area, probably because of the changes in density and properties of the ex-
845
INITIATION OF SOLID EXPLOSIVES BY SHOCK WAVES
.
o
FIG. 3. Streak camera photographs taken with the arrangement in Figure 2 plosives caused by the shock wave 9, 10. The line alba~ corresponds rather closely to the boundary where the front of shock and detonation waves meet. It is interesting to note that Cook, Pack and Gey ~ have found that the detonation in composition B in certain cases spreads symmetrically, whereas in others, it does not. The time from the pressure rise at the end surface until the detonation front first reaches the
envelope surface according to Figure 4 and Equation (1) is r ' = zo +
w
__R -
ro
(2)
D
in which w is the speed of the shock wave and R the radius of the acceptor. The quantities z0 and r0 are functions of the radius of the acceptor. For r'/zo to be constant, as Cook 4 has found, the v~ri-
. eptor
f"
FIG. 4. The detonation front drawn with 0.39 ~sec interval according to Figure 3a.
: ~:::
i~::i/ ~ ii~:ii:~
FIo. 5. Instantaneous photgraphs of the detonation front 846
FLAME FRONTS AND COMPRESSION WAVES
ations in (R - ro)/Zo must be small compared to D/w. Acknowledgments
The authors acknowledge their indebtedness to the Foundation Swedish Research of Detonics, Stockholm, and to the Nitroglycerin Aktiebolaget, Gyttorp, for granting facilities for this research work. REFERENCES
847
Seventh Symposium (International) on Combustion, p. 820. Butterworth and Company, Ltd., London, 1959. 4. COOK, M. A.: The Sciences of High Explosives, Reinhold Publishing Corporation, New York, 1958. 5. JAcoBs, S. J., AND MAJOWICS,J.: Seventh Symposium (International) on Combustion, p. 879. Butterworth and Company, Ltd., London, 1959. 6. BERGER, J., FAVIER, J., ET FAUQUIGNON, C.:
Compt. Rend. l'Acad. Sci., 247, 1305 (1958). 7. PERSSON, A.: Appl. Sci. Res., A6, 365 (1956). 8. MARLOW, W. R., AND SKIDMORE, 1. C. : Proc.
1. CACHIA,G. P., AND WHITBREAD, E. G.: Proc. Roy. Soe. (London), A246, 268 (1958). 2. EICHELBERGER, R. J., AND SULTANOFF, M. : Proc. Roy. Soc. (London), A246, 274 (1958). 3. CooK, M. A., PACK, D. H., AND GEY, W. A.
Roy. Soc. (London), A246, 284 (1958). 9. AHRENS, H., AND EITZ, E.: Nobel Hefte, 24, 244
(1958). 10. SCHALL,R.: Nobel Hefte, 21, 1 (1955). 11. JOHANSSON, C. n., AND STERNHOFF, L. : N a t u r e ,
183, p. 247 (1959).
93 FLAME FRONTS AND COMPRESSION WAVES DURING TRANSITION FROM DEFLAGRATION TO DETONATION IN SOLIDS By ROYCE W. GIPSON AND A N D R E J M A ~ E K Introduction
Studies of the problem of spontaneous transition from deflagration (slow, pressure-dependent burning) to detonation in gases go back as far as the turn of the century. 1 Since that time it has often been shown, experimentally, by means such as schlieren photography, that in an accelerated deflagration a shock front runs ahead of the flame. The interest in the analogous phenomenon in solids is of a much more recent date. The problem has been investigated at this laboratory both experimentally and theoretically; the leading idea has been the hypothesis, probably first suggested as applying to solids by Kistiakowsky/ that transition is caused by a precursor shock which arises because of rapid pressure increase behind the deflagration front and then propagates into unburned explosive. Previously reported experimental results 3, 4 with two fairly sensitive cast explosives, diethylnitramine dinitrate (DINA) and 50/50 pentolite, showed that thermally initiated slow burning
will under confinement go over into detonation even in charges of moderate dimensions. Furthermore, it was shown that the experimental run between the point of thermal ignition and onset of detonation (6 to 18 cm) coincides quite well with the theoretically computed run (10 to 15 cm) necessary for development of a shock wave under the same conditions of confinement. Such results tend to confirm the original postulate that the course of transition proceeds essentially through two steps, shock formation and shockinitiation of detonation. The theoretical mechanism of shock formation during buildup of detonation has been discussed/whereas the subject of shock initiation of detonation has been dealt with rather extensively in recent years.~. 6.7 Thus, it appears promising to seek an explanation of the largely unknown transition phenomenon in terms of better known processes. The present paper is concerned mainly with experiments designed to verify in a somewhat more detailed fashion the applicability of the
FORMATION, COMBUSTION, EXPLOSION AND DETONATION OF SOLIDS
free surface area decreases and the rate of decomposition falls. Further work on silver cyanamide is summarized elsewhere. 14 REFERENCES 1. GARNER, W. E. (ed.): Chemistry of the Solid State, Butterworths Scientific Publications, London, 1955. 2. BARTLETT, B. E., TOMPKINS, F. C., AND YOUNG, D. A.: Proe. Roy. Soe. (London), A2~6, 206 (1958). 3. GRAY, P., AND WADDINGTON, T. C.: Proc. Roy. Soc. (London), A241, 110 (1957). 4. McLAREN, A. C., AND RODGERS, G. T.: Proc. Roy. Soc. (London), A246, 250 (1958). 5. DEB, S. K., EVANS, B. L., AND YOFFE, A. D.: Nature, (London), 180, 294 (1957). 6. Discussion on the initiation and growth of explosion in solids, Proc. Roy. Soc. (London), A2.~6, 145 (1958).
7. COOK, G. B.: Sixth Symposium (International) on Combustion, Reinhold Publishing Corporation, New York, 1956. 8. DES, S. K., AND YOFFE, A. D.: Trans. Faraday Sou., 55, 106 (1959). 9. CAMP, M., DEB, S. K., EVANS, B. L., MONTAGu-POLLOCK, H. M., AND YOFFE, A. D.: Fourth International Symposium on the Reactivity of Solids, Elsevier, Delft, 1961. 1O. BOWDEN, F. P., AND YOFFE, A. D.: Fast Reactions in Solids, Butterworths Scientific Publications, London, 1958. 11. McAuSLAN, J. H.: Ph.D. Thesis, Cambridge, England, 1957. 12. PASHLEY, D. W.: Advances in Physics, 5, 173 (1956). 13. MITCHELL,J. W., ANDMOTT, N. F.: Phil. Mag., 2, 1149 (1957). 14. BOWDENF. P., ANn MONTAGe-PoLLOCK,H. M.: Nature (London), 191,556 (1961).
92 THE INITIATION OF SOLID EXPLOSIVES BY SHOCK WAVES By C. H. JOHANSSON, N. LUNDBORG AND T. SJOLIN Introduction When investigating the propagation of detonation among solid explosives through plates of inert material, the process has, in most cases, been recorded at the surface of the charges. 1-6 It is, however, of great interest to extend the investigation to the interior of the charges 7-1°and observe step by step how detonation begins and spreads through the secondary charge. For this purpose we have recorded the light emission from cylindrical charges of compressed TNT, 1.55 g/cm ~ density with radial drill holed and semicylindrical acceptor charges coaxial with the cylindrical donor charges. The results indicate that the process in axial planes of undivided cylindrical acceptors is only slightly affected if the charge is cut along an axial plane and one of the halves is removed, and the cut surface is covered with a Plexiglass plate or the charges are submerged in water.
Cylindrieal Charges w i t h Radial Drill Holes The donor and acceptor charges were cylinders 21 mm in diam, and 80 mm in length with 2 mm
radial drill holes with a 10 mm distance between the centers. The plate between the charges was 50 x 50 mm plate of hard-rolled aluminum (>99 per cent A1). Detonation propagation was obtained in all instances with a plate thickness of 20 mm, but in no instance with a thickness of 22 mm. Figure 1 shows the arrangement (left) and the result (right) of streak camera photos in two tests with a 20 mm plate. The drill holes drawn with solid lines and the plain circles of the diagram refer to the one test (No. 1); the holes drawn with dashed lines and the circles with crosses refer to the other. All drill holes except the two nearest the plate in the acceptor gave intense light spots characteristic of a passing detonation front. The hole adjacent to the A1 plate on the acceptor side gave a faint luminous trace compatible with the assumption that it was caused by the shock wave. The trace of light from the hole 5 mm from the A1 plate in the acceptor is slightly narrower than the other, but has the same intensity and shows that the detonation has commenced. In a control test substituting a similar cylinder of sulphur for the acceptor and with
INITIATION OF SOLID EXPLOSIVES BY SHOCK WAVES
843
eptor cm
EEC.
-MO
V__:_-
120 E~_-cu.-
Dekonofionfronh
I00
/
surfacy ShockwovefrontJ (
80
AI-plofe
/
I
/ / /
J,~ [ •
-
-
50 40 /
/
ox/$ surfoce
i--4__
l _ _
•
i--'---"
Donor
22/1111 0
I I I I I I I /0
20/us
FIG. 1. Arrangement of charges with bore holes and diagram according to two streak camera photographs. the holes placed as in test No. 1, no traces of light were obtained from the holes. According to the diagram in Figure 1, the shock wave in the acceptor becomes a detonation wave ca. 5 mm from the end face at the axis and thereafter moves forward with a normal detonation velocity. This is in agreement with the results which Marlow and Skidmore 7 obtaiued when they measured the front velocities at the axis with "single wire" and "twin wire" probes set in radial drill holes. The light emission at the envelope surface shows that the detonation does not move backward but leaves the end of the charge behind.
Semicylindrical Acceptors Figure 2 shows the arrangement of the charges and the screen with parallel slits in front of the axial plane of the semicylindrical acceptor. The
charges were submerged in an aquarium with water and placed in front of a streak camera with a rotating mirror• Figure 3 reproduces two recomings with a cylindrical donor 21 mm in diam and 80 mm in length, a semicylindrical aeceptor 63 mm in diam and 70 mm in length, and an intermediate cylinder of AI alloy 21 mm in diam and 18 mm in length. In Figure 3a, the axial plane of the acceptor had a coating of black varnish in which 11 slits were scratched at distances of 3 ram. In Figure 3b, the screen consisted of a brass plate which was placed 50 mm in front of the charge surface. The width of the slits was 0.5 mm and the distance between the slits 3 mm. In this case the first slit happened to be 3 mm to the right of the center and s l i t N o . 11 was outside the charge and received no light. By inserting a still picture of the slits in the pictures of Figure 3 it is easy to construct the
844
FORMATION, COMBUSTION, EXPLOSION AND DETONATION OF SOLIDS
t---
-~- Screen with s/its ------~
Camera
Streak camera pictures of the envelope surface of undivided cylindrical acceptors with diameters of 21, 40 and 63 mm respectively, show satisfactory agreement with the pictures of the front in Figure 4 (cf. Fig. 1). Streak camera pictures of the detonation front surface at the upper end of the semicylindrieal acceptor taken with 7 slits perpendicular to the axial plane show a small downward bend in the detonation front at the axial plane. I t is, however, very slight and probably too small to be of practical significance for the spread of the detonation front.
o
Incipience
---z.-•
Wo/er ~
Camera
--=----- , ~ ¢ / ~ e n
--
Accepfor
-- -_---A / - c q / i n d e r
_~-~D o n o / "
and
Spread of the Detonation
When the donor detonates, the upper surface of the A1 cylinder is suddenly given a velocity and generates a shock wave in the aceeptor, just as a projectile would. Our measurements show that the velocity necessary for initiation is about 650 m/sec in both cases. In a semieylindrical acceptor with a Plexiglass cover, the shock wave produces a faintly visible luminous front line in the axial plane. When the shock wave has proceeded a certain distance it changes over to detonation which spreads from an approximately plane, circular area upwards and sideways with ordinary velocity. The detonation front is approximately in the shape of an ellipsoid with the equation r2
(Dr + ro) 2 :FIG. 2. Semicylindrical acceptor arranged for streak camera photographing with 11 slits. luminous detonation front at various times. In Figure 4 is reproduced a pile of instantaneous pictures of the front with 0.39 t~sec interval obtained from Figure 3a. They are in accordance with the instantaneous photographs of the detonation front in Figure 5. In this case the charges were placed in air and to retard the expansion of detonation gases perpendicular to the axial plane and eliminate disturbing light from the air, a 10-ram thick Plexiglass plate was glued on to the surface. In addition the acceptor was not semicylindrical but a 15-mm thick slab of compressed TNT. These differences in the experimental conditions in comparison with Figure 2, are in all probability not essential to the propagation of the detonation front.
+
(z -- z0) 2
(Dt) 2
-- 1.
(1)
The z-axis coincides with the axis of the acceptor charge, the origin lies in the end surface of the acceptor, z0 is the initiation depth, r is the distance from the axis, r0 the radius of the initial circular surface, D the detonation velocity and t the time from the start of the initiation. In Figure 4, D = 6.65 mm/t~see, z0 = 8 ram, r0 = 5 mm and when the detonation starts the z-value of the upper surface of the A1 cylinder is about 1.7 mm. The detonation front is not visible below the line al ba~. We have raised the question if the chemical reaction started by the shock wave has caused a light-absorption similar to the phenomena observed by Johansson and Sternhoff 8 when a detonating charge of T N T in water is exposed to high pressure. However, experiments with undivided cylindrical charges according to Figure 1 show that the detonation does not spread in this area, probably because of the changes in density and properties of the ex-
845
INITIATION OF SOLID EXPLOSIVES BY SHOCK WAVES
.
o
FIG. 3. Streak camera photographs taken with the arrangement in Figure 2 plosives caused by the shock wave 9, 10. The line alba~ corresponds rather closely to the boundary where the front of shock and detonation waves meet. It is interesting to note that Cook, Pack and Gey ~ have found that the detonation in composition B in certain cases spreads symmetrically, whereas in others, it does not. The time from the pressure rise at the end surface until the detonation front first reaches the
envelope surface according to Figure 4 and Equation (1) is r ' = zo +
w
__R -
ro
(2)
D
in which w is the speed of the shock wave and R the radius of the acceptor. The quantities z0 and r0 are functions of the radius of the acceptor. For r'/zo to be constant, as Cook 4 has found, the v~ri-
. eptor
f"
FIG. 4. The detonation front drawn with 0.39 ~sec interval according to Figure 3a.
: ~:::
i~::i/ ~ ii~:ii:~
FIo. 5. Instantaneous photgraphs of the detonation front 846
FLAME FRONTS AND COMPRESSION WAVES
ations in (R - ro)/Zo must be small compared to D/w. Acknowledgments
The authors acknowledge their indebtedness to the Foundation Swedish Research of Detonics, Stockholm, and to the Nitroglycerin Aktiebolaget, Gyttorp, for granting facilities for this research work. REFERENCES
847
Seventh Symposium (International) on Combustion, p. 820. Butterworth and Company, Ltd., London, 1959. 4. COOK, M. A.: The Sciences of High Explosives, Reinhold Publishing Corporation, New York, 1958. 5. JAcoBs, S. J., AND MAJOWICS,J.: Seventh Symposium (International) on Combustion, p. 879. Butterworth and Company, Ltd., London, 1959. 6. BERGER, J., FAVIER, J., ET FAUQUIGNON, C.:
Compt. Rend. l'Acad. Sci., 247, 1305 (1958). 7. PERSSON, A.: Appl. Sci. Res., A6, 365 (1956). 8. MARLOW, W. R., AND SKIDMORE, 1. C. : Proc.
1. CACHIA,G. P., AND WHITBREAD, E. G.: Proc. Roy. Soe. (London), A246, 268 (1958). 2. EICHELBERGER, R. J., AND SULTANOFF, M. : Proc. Roy. Soc. (London), A246, 274 (1958). 3. CooK, M. A., PACK, D. H., AND GEY, W. A.
Roy. Soc. (London), A246, 284 (1958). 9. AHRENS, H., AND EITZ, E.: Nobel Hefte, 24, 244
(1958). 10. SCHALL,R.: Nobel Hefte, 21, 1 (1955). 11. JOHANSSON, C. n., AND STERNHOFF, L. : N a t u r e ,
183, p. 247 (1959).
93 FLAME FRONTS AND COMPRESSION WAVES DURING TRANSITION FROM DEFLAGRATION TO DETONATION IN SOLIDS By ROYCE W. GIPSON AND A N D R E J M A ~ E K Introduction
Studies of the problem of spontaneous transition from deflagration (slow, pressure-dependent burning) to detonation in gases go back as far as the turn of the century. 1 Since that time it has often been shown, experimentally, by means such as schlieren photography, that in an accelerated deflagration a shock front runs ahead of the flame. The interest in the analogous phenomenon in solids is of a much more recent date. The problem has been investigated at this laboratory both experimentally and theoretically; the leading idea has been the hypothesis, probably first suggested as applying to solids by Kistiakowsky/ that transition is caused by a precursor shock which arises because of rapid pressure increase behind the deflagration front and then propagates into unburned explosive. Previously reported experimental results 3, 4 with two fairly sensitive cast explosives, diethylnitramine dinitrate (DINA) and 50/50 pentolite, showed that thermally initiated slow burning
will under confinement go over into detonation even in charges of moderate dimensions. Furthermore, it was shown that the experimental run between the point of thermal ignition and onset of detonation (6 to 18 cm) coincides quite well with the theoretically computed run (10 to 15 cm) necessary for development of a shock wave under the same conditions of confinement. Such results tend to confirm the original postulate that the course of transition proceeds essentially through two steps, shock formation and shockinitiation of detonation. The theoretical mechanism of shock formation during buildup of detonation has been discussed/whereas the subject of shock initiation of detonation has been dealt with rather extensively in recent years.~. 6.7 Thus, it appears promising to seek an explanation of the largely unknown transition phenomenon in terms of better known processes. The present paper is concerned mainly with experiments designed to verify in a somewhat more detailed fashion the applicability of the