Combustion, Explosion, and Shock Waves, Vol. 46, No. 2, pp.

216–224, 2010

Generalized Analysis of Explosion Propagation

in Condensed Systems

G. T. Volodin1 UDC 544.454:662.215.4

Translated from Fizika Goreniya i Vzryva, Vol. 46, No. 2, pp. 111–120, March–April, 2010.
Original article submitted May 7, 2009.

The complex problem of the propagation of explosion of condensed explosives in gas

media (air) is solved analytically. A generalized model is constructed which describes
the following processes: explosion propagation in an explosive charge from the time of
initiation to the establishment of the normal detonation regime (explosion transient);
proper normal detonation; expansion of explosion products into the medium (air) with
the formation of a blast wave and motion of the perturbed medium in the blast wave.
Unlike in the well-known formulations of such problems using the classical hypothesis
about adiabatic explosive processes, the present study is based on the more natural
hypothesis about an exponential density profile of the perturbed medium and explo-
sion products. The method of Salamakhin is extended to real explosion processes.
Key words: propagation of explosion, shock wave, detonation, contact surface,
explosion transient.

INTRODUCTION the SW front, the distribution of the parameters in the

perturbed region at any time was solved in [7]. The
Investigation of explosions of condensed explosive obtained solution corresponds to the model of explo-
(HE) charges involves the necessity of formulating and sion propagation in a HE charge which is the closest to
solving several classes of interrelated problems (of the real explosive process because, unlike in traditional ap-
physics of explosion) related to explosion propagation proaches based on the adiabaticity hypothesis, it uses
in a HE charge, detonation as the limiting regime of ex- the more natural assumption of an exponential density
plosion propagation, and interaction of expanding ex- profile of the perturbed region.
plosion products with the medium. The explosion propagation in the initial region is
Explosion Propagation in a HE charge. It is followed by a limiting stationary regime — normal
known [1–5] that the chemical transformation of a HE detonation propagating in the main HE charge with
can start with combustion or explosion, depending on a constant velocity D0 .
the properties and size of the charge and the energy Normal Detonation Regime. Among the ba-
of the initiator. In practical calculations, the complex sic detonation problems are: the structure of the det-
process of explosion with a variable velocity is usually onation wave (DW); the stability of the detonation
replaced by detonation, for which the shock wave (SW) regime; the distribution of the parameters of detona-
velocity is constant if the initial region in which explo- tion products (DP) in the DW.
sion is converted to detonation is neglected [1]. How- Difficulties in creating a mathematical model for
ever, experiments have shown [6] that the initial region detonation are related, first, to the necessity of using
has a substantial extent, and, hence, it should be con- an equation of state for DPs that provides the most
sidered in calculations of explosive loads. adequate description of their real behavior in the DW.
The problem of determining the sizes of the initial A fairly full review and analysis of existing equa-
region, the law of motion of the SW front, the parame- tions of state for explosion products of condensed HEs
ters in it, the heat release law in a narrow zone behind is given in a paper [8–10], which focuses on two ap-
proaches: deriving the equation of state based on exper-
1
Tula State University, Tula 300600; [email protected].
216 0010-5082/10/4602-0216 
c 2010 Springer Science + Business Media, Inc.
Generalized Analysis of Explosion Propagation in Condensed Systems 217

imental data on detonation for the solution of various produced by the SW reaches a minimum only for an
practical problems; deriving the equations of state of exponential form of the density profile in the DW.
explosion products for the purpose of predicting detona- Expansion of DPs and Motion of the Per-
tion parameters for the cases where direct experimental turbed Medium in the SW. After SW arrival at
data on detonation are absent. the surface of the charge, the perturbed region consists
Another difficulty is related to the choice of a clos- of two subregions: subregion I, occupied by expanding
ing equation [11] in the system of the governing equa- explosion products, and subregion II of the perturbed
tions of the problem. In many studies, this equation rep- medium in the SW. Subregion I is separated from II
resents a particular mathematical formulation of the hy- by a thin (according to numerous experiments) contact
pothesis about adiabatic flow of the perturbed medium surface layer (CS) in which particles of the perturbed
by virtue of its rapidity. However, it is questionable medium are mixed with explosion products. The outer
that heat conduction in the perturbed region is absent boundary of subregion II is the SW front.
because of the rapidity of the process. The motion of the perturbed subregions I and II
There is another hypothesis [10, 12] opposite to is accompanied by repeated propagation of rarefaction
the hypothesis about an adiabatic explosive process, waves, each time reducing the amplitudes of the state
namely: it is assumed that the process in DP is ho- variables until they become equal to the parameters of
mothermic: the temperature T depends only on time t the ambient medium.
and does not depend on the coordinate r (for one- To describe the motion of the perturbed subre-
dimensional motion), i.e., ∂T /∂r = 0. This assumption gions I and II, it is necessary to simultaneously in-
implies that, in the region encompassed by the SW, the tegrate two systems of nonlinear partial differential
heat transfer rate tends to infinity at the earliest stage equations, each of which describes the behavior of the
of explosion. Considering the extremely high tempera- medium in a corresponding subregion. The above-
tures due to the SW and other explosion mechanisms mentioned systems are related by common boundary
in the vicinity of the center of explosion, the assump- conditions on the CS.
tion of a zero temperature gradient in the radial direc-
tion seems justified. However, under this condition, the
study is limited to the early stage of explosion.
To avoid studying complex microprocesses in the GENERAL FORMULATION
DP and introducing doubtful constraints in the physi- OF THE PROBLEM
cal model of the problem, Salamakhin [13] proposed an
original approach: to approximate the density distribu- We assume that expansion of DPs begins at the
tion in the perturbed region by an exponential function time when the DW front arrives at the surface of
with a time-dependent exponent which is found from in- the charge. This time is taken as the reference time.
tegral mass and energy conservation laws written for the We seek a solution of the problem for a time interval
entire perturbed region [13]. A comparative analysis of 0 < t < t∗ , where t∗ for the time when the SW degener-
the closing equations corresponding to the proposed hy- ates to a sound wave. For any t > 0, we assume that the
potheses about the behavior of the perturbed medium perturbed region consists of the inner subregion I occu-
during explosion is given in [14], where advantages of pied by DPs and the outer subregion II, occupied by the
this method over other approximate methods is shown. moving ambient medium. The medium surrounding the
The density profile method was developed in [15], charge is in rest before the SW arrival: the velocity of
where a solution in quadratures to the problem of det- its particles u1 = 0, pressure p1 , and density ρ1 are con-
onation of a condensed HE was obtained using one of stant. The interface between the subregions considered
the known equations of state of DPs [16] neglecting the is a discontinuity at which the pressure and particle ve-
insignificant variation in the specific heat capacity in locities on both sides are identical, and the density and
the temperature interval studied. temperatures are different. In this case, the velocity of
The further development of the density profile motion of the CS u∗ is equal to the velocity of motion
method is related to the analysis of the form of the dr∗
of particles of the medium on this surface: u∗ = ,
closing equation defining the density distribution in the dt
where r∗ is the coordinate of the CS.
DW. Such an analysis was performed in [17] using a
The last statement is based on numerous experi-
direct variational method: the chosen test extremals mental data, according to which the layer between sub-
used to define the density profiles (exponent, hyper- region I and II in which DP are mixed with particles of
bolic sine, hyperbolic cosine, and exponential function)
the medium has a negligibly small thickness compared
showed that the functional defining the total impulse to the extent of the subregions considered. The outer
218 Volodin

boundary of the perturbed region is the SW front, at

which the parameters of the medium change suddenly R̄ρ̄T̄ R̄0
p̄ = , R̄ = ,
from the initial values p1 , ρ1 , and u1 = 0 to the param- 1 − bρ̄ M
eters at the SW front p2 , ρ2 , and u2 > 0. p̄(1 − bρ̄)
It should be especially noted that the expansion Ē = ,
(χ1 − 1)ρ̄
of DPs and the motion of the medium in the SW are
substantially affected by rarefaction waves. The effect where ρ̄, ū, p̄, T̄ , and Ē|r, t are the required functions
of rarefaction waves repeatedly propagating in the per- which describe the distributions of the density, parti-
turbed region can be taken into account from the wave cle velocity, pressure, temperature, and specific internal
nature of its density variation. energy on the coordinate r at any fixed time t. The first
Assuming that the density of explosion products in two equations of this system are the laws of conservation
the DW is distributed as of mass and momentum, respectively, the third equation
formulates the density distribution of explosion prod-
ρ̄ = ρ3 exp(β x̄ν ), (1) ucts, and the function β(t) takes into account the wave
where ρ3 is the density at the center of symmetry, x̄ = nature of the density by virtue of the repeatedly prop-
r/r2 , r2 = D0 t (D0 is the detonation velocity), and agating rarefaction waves. The next to last equation
ν is the flow dimension parameter (ν = 1, 2, and 3 for of system (2) is the equation of state for the expand-
plane, cylindrical, and spherical waves, respectively), we ing DPs, which is a simplified van der Waals equation.
obtain a closed analytical solution for the detonation The last equation of the system expresses the specific
regime [18] (β = β0 = const and ρ3 = const). internal energy of the expanding DPs. The function
We assume that, with the beginning of expansion ρ3 (t) is the time-varying density at the center of sym-
of DPs, their density is also distributed under a similar metry, M is the molar mass, R̄0 is the universal gas
law (1), where now ρ3 = ρ3 (t), β = β(t), x̄ = r/r∗ , and constant, and b is the covolume.
r∗ is the coordinate of the CS. The boundary conditions of system (2) are
Taking into account the wave nature of the den- as follows:
sity variation in the perturbed region, we specify the at the center of symmetry,
function ū(0, t) ≡ 0, (3)
β(t) = β0 exp(−γτ ) cos ωτ, on the CS (at r = r∗ )
where τ = a1 t/r0 [a1 is the sound velocity in the unper- p̄(r∗ , t) = p∗ , ρ̄(r∗ , t) = ρ+
∗ , ū(r∗ , t) = u∗ , (4)
turbed medium and r0 is the radius (characteristic size)
of the HE charge], β0 is the initial value of the function where p∗ , ρ+
∗, and u∗ |t are the pressure, density, and
β, which is found from the solution of the detonation particle velocity, respectively, on the CS; ρ+ ∗ denotes
problem (for example, for TNT, β0 = 0.60586 [18]), and the density on the inner side of the CS (on the side of
the parameters γ and ω are determined from experimen- explosion products).
tal data. It is convenient to define the parameter ω as The initial conditions are specified distributions
of the density, particle velocity, pressure, temperature,
πδ
ω= , and internal energy in the perturbed region at the ini-
q0
tial time t = 0, which are known from the solution of
where q0 is the initial value of the measure of the SW the detonation problem [18].
strength, q0 = a1 /D0 and δ is a parameter found from In the following, we need the integral law of con-
experimental data. servation of mass for subregion I:
r∗
1
ρ̄rν−1 dr = ρ0 r0ν . (5)
Subregion I ν
0

The motion of the DPs occupying subregion I at

each time t > 0 is defined by the system of equations
∂ ρ̄ ∂ ρ̄ ∂ ū ρ̄ū Integration of the System of Equations
+ ū + ρ̄ + (ν − 1) = 0,
∂t ∂r ∂r r
1. The density distribution is described by the
∂ ū ∂ ū 1 ∂ p̄
+ ū + = 0, third equation of system (2), in which the functions of
∂t ∂r ρ̄ ∂r time ρ3 (t) and β(t), however, are not known. The func-
ρ̄ = ρ3 exp(βξ), ξ = x̄ν , x̄ = r/r∗ , (2) tion ρ3 (t) can be expressed in terms of the function β(t)
Generalized Analysis of Explosion Propagation in Condensed Systems 219

and the coordinate r∗ (t) of the CS, using the integral (r ∈ [r∗ , r2 ]) at the time t; ρ2 = ρ(r2 , t) is the density
of mass (5), into which we the substitute the expression at the SW front; the function α(t) is found during the
for the density from the third equation of system (2): solution of the problem.
β ρ0 The equation of state for the perturbed medium is
ρ3 (t) = , (6) written in calorific form
exp β − 1 R∗ν p
E= , (10)
where R∗ = r∗ /r0 . (χ − 1)ρ
Hence, the density distribution in subregion I is
where χ = cp /cv is the ratio of the heat capacities at
defined by the function ρ̄ in system (2), where ρ3 (t) is
constant pressure and constant volume.
expressed by Eq. (6). The function R∗ (t) is found by
The outer boundary of subregion II is the SW front
solving the problem of motion of the perturbed medium
(r = r2 ), and the inner boundary is the CS (r = r∗ ).
in subregion II.
Consequently, in addition to the boundary conditions
2. The function ū(r, t) defining the particle velocity
on the CS
distribution in subregion I is found by integrating the
first equation of system (2), in which the function ρ̄ is u(r∗ , t) = u∗ , ρ(r∗ , t) = ρ−
∗ , p(r∗ , t) = p∗ , (11)
replaced by the expression found for it using boundary
where ρ−
∗ is the density of the medium on the outer
condition (3):
side of the CS, i.e., on the side of the SW, we write the
  
r∗ β̇ 1 − exp(−βξ) conditions at the SW front:
ū(r, t) = u∗ x̄ 1 − 1− , (7)
u∗ νβ ξ(1 − exp(−β)) u(r2 , t) = u2 , ρ(r2 , t) = ρ2 , p(r2 , t) = p2 . (12)
where the dot above the function denotes the time dif- In this case, the dynamic compatibility condition [17]
ferentiation operator. should be satisfied:
3. The function p̄(r, t) describing the pressure dis- ρ1 2
tribution in subregion I is found by integrating the sec- 1− = (1 − q 2 ),
ρ2 χ+1
ond equation of system (2), into which we substitute (13)
the obtained functions ρ̄(r, t) and ū(r, t) using bound- p 2 − p1 2χ  1  u2 2q  1 
ary condition (4) for pressure: = 2
−1 , = 2
−1 ,
p1 χ+1 q a1 χ+1 q
p̄ ρ3 u2∗ where the quantity q = a1 /D (D = dr2 /dt) defines
h̄ = =1− [Π2 (1 − x̄2 )
p∗ p∗ the measure of the SW strength (q = 0 for an infinitely
+ Π12 (1 − x̄ν−2 ) + Π22 (1 − x̄2ν+2 ) strong SW and q = 1 for a SW degenerated into a sound
wave). Formulas (13) were obtained from the conserva-
+ Π32 (1 − x̄3ν+2 ) + Π42 (1 − x̄4ν+2 ) tion laws across the SW front [1].
+ Π52 (1 − x̄5ν+2 ) + Π62 (1 − x̄6ν+2 )], (8) The initial conditions are

where the functions Π2 , Π12 , . . . , Π62 are found in [17]. r2 (0) = r0 , r∗ (0) = r0 . (14)
Distributions (6)–(8) contain the yet unknown pa- The integral law of conservation of mass for subregion II
rameters of the medium u∗ and p∗ on the CS (at r = r∗ ), is written as
which we find from the solution of the problem of mo- r2
tion of subregion II in the SW and boundary conditions 1
ρrν−1 dr = ρ1 (r2ν − r0ν ). (15)
(4) on the CS. ν
r∗

The integral law of conservation of energy for the entire

Subregion II perturbed region (subregion I and II) has the form
 
δν p1 (r2ν − r0ν )
The governing equations for subregion II can be θ0 + ρ0 (Q0 + E0 )r0ν +
ν χ−1
written as
∂ρ ∂ρ ∂u ρu r∗  
+u +ρ + (ν − 1) = 0, ρ̄ū2 p̄(1 − bρ̄) ν−1
∂t ∂r ∂r r = δν + r dr
 α(t) (9) 2 χ1 − 1
∂u ∂u 1 ∂p ρ r 0
+u + = 0, = ,
∂t ∂r ρ ∂r ρ2 r2 r2  

ρu2 p
where ρ, u, and p|r, t are the density, particle velocity, + + rν−1 dr , (16)
and pressure, respectively, at the point r of subregion II 2 χ−1
r∗
220 Volodin

where E0 is the specific internal energy of the HE in On the other hand, the third equation of system (9) and
the unperturbed state, θ0 is the energy of the explosion relations (13) for a strong shock leads to
initiator released at the point (ν = 3), per unit length 2
(ν = 2), and per unit area (ν = 1), Q0 is the specific en- u2 = ṙ2 . (22)
χ+1
ergy of explosive transformation, χ1 is the ratio of heat
capacities for explosion products, χ is the ratio of heat Comparing Eqs. (21) and (22), we determine
capacities for the perturbed medium in subregion II, 2
δν = 2π(ν − 1) + (ν − 2)(ν − 3). α = α0 = . (23)
χ−1
Thus, in the case of a strong shock, the density distri-
bution in subregion II is described by the function
Integration of the System of Equations
  χ−12ν
r
Case of a Strong Shock. For the case of a strong ρ = ρ2 , (24)
r2
shock, the following relation holds:
and the particle velocity distribution by the function
q 2  1. (17)  
2a1 r
1. We introduce the dimensionless variables x = u= (25)
(χ + 1)q r2
r/r2 , R = r2 /r0 , and R∗ = r∗ /r0 .
The density distribution in subregion II is defined or
 
by the third equation of system (9), in which the func- r
u = u2 .
tion α(t) is not known. It can be found by using the in- r2
tegral law of conservation of mass (15) and relation (17)
3. We substitute the expression for the function u
for a strong shock. Indeed, substitution of the expres-
from (25) into the second equation of system (9) and in-
sion for the density from the third equation of system
tegrate the obtained equation with respect to the func-
(9) into the integral law (15) yields
  tion p(r, t), using boundary condition (12) for the pres-
ρ2 1 − xα+1
∗ sure, where, according to the second equation of system
α=ν −1 , (18)
ρ1 1 − R−ν (13), the function p2 for a strong shock is expressed as
where x∗ = r∗ /r2 . In view of (17), the first equation of 2χp1
p2 = . (26)
system (13) leads to (χ + 1)q 2
ρ2 χ+1 As a result,
= . (19)
ρ1 χ−1 p
h= = 1 − μ(1 − xα0 +2 ), (27)
From (18) and (19), we find the following expression for p2
the function α in the case of a strong shock: where
   
χ + 1 1 − xα+1
∗ 2ρ1 a21 R dq χ−1
α=ν −1 . (20) μ= + .
χ − 1 1 − R−ν (χ − 1)(α0 + 2)q 2 p2 q dR χ + 1
Relation (20) in implicit form defines the function α in Thus, the solution of the problem for the case of
terms of the law of motion of the CS r∗ (t) and the law a strong shock is defined by the functions ρ̄(r, t) from
of motion of the SW front r2 (t). However, as will be the third equation of system (2), ū(r, t) in the form
shown below, for a strong shock, the function α keeps (7), p̄(r, t) from relation (8) for subregion I, and the
a constant value and can be determined in seeking the functions ρ(r, t) in the form (24), u(r, t) from (25), and
particle velocity distribution in subregion II. p(r, t) in the form (27).
2. We substitute the expression for the density This solution includes the unknown parameters of
from the third equation of system (9) into the first equa- the medium ρ2 , u2 , and p2 at the SW front and the
tion of this system and integrate the obtained equation parameters of the medium ρ+ −
∗ , ρ∗ , u∗ , and p∗ on the
using boundary condition (12) for the velocity. As a CS. The laws of motion for the SW front and the CS
result, are yet unknown.
α ṙ2 The parameters at the SW front can be expressed
u= r.
α + 1 r2 in terms of the measure of the SW strength q and the pa-
From this for r = r2 , we obtain rameters of the unperturbed medium by formulas (13).
α The function ρ∗ on the CS has a discontinuity of
u2 = ṙ2 . (21) the first kind. On the CS from the side of the explosion
α+1
Generalized Analysis of Explosion Propagation in Condensed Systems 221

products, ρ∗ = ρ+ +
∗ , where ρ∗ is determined from the Shock of Arbitrary Strength. Generally (for a
third equation of system (2) for x̄ = 1 in the form SW of arbitrary strength), the statement of the problem
of expansion of explosion products remains the same
ρ+
∗ = ρ3 exp(β). (28)
as in the case of a strong shock. The solution of this
Here ρ3 is expressed by formula (6). problem is defined by formulas (6)–(8) which include the
On the side of subregion II, the density on the CS unknown parameters u∗ and p∗ on the CS (at r = r∗ ).
ρ−
∗ is defined by the third equation of system (9), in These parameters are found by solving the problem of
which, it is necessary to set r = r∗ , namely: motion of the perturbed medium in subregion II and
ρ− α the boundary conditions on the CS.
∗ = ρ2 x∗ , (29)
Unlike in the case of a strong shock, in the general
or according to (24), case of integration of the system of equations for sub-
2ν region II, the relation for the function α, according to
ρ− χ−1
∗ = ρ2 x∗ . (18) and (13), becomes
The functions u∗ and p∗ on the CS are continuous and,  
χ+1 1 − xα+ν
∗
hence, according to formulas (25) and (27), they can be α=ν − 1 . (35)
χ − 1 + 2q 2 1 − R−ν
written as
2a1 In this case, the density distribution in subregion II is
u∗ = x∗ , (30) described by the function
(χ + 1)q  α
p∗ r
h∗ = = 1 − μ(1 − x∗α0 +2 ). (31) ρ = ρ2 . (36)
p2 r2
These expressions imply that the obtained solution of The particle velocity distribution is obtained by inte-
the problem is expressed in terms of the yet unknown grating the first equation of system (9) using the bound-
laws of motion for the CS r∗ = r∗ (t) and the SW front ary condition for the velocity (12):
r2 = r2 (t). u
Let us find the relationship between the specified z= = x1−α−ν (1 − ε1 ) + ε1 x − ε2 x ln x. (37)
u2
laws. Substitution of the value of α from (23) into Here
Eq. (20) yields   
1 α̇ ρ̇2
2 ε1 = αṙ2 + r2 − , (38)
R∗ = R χ+1 . (32) (α + ν)u2 α+ν ρ2
Relation (32) links the law of motion of the CS and the α̇r2
ε2 = .
law of motion of the SW front for the case of a strong (α + ν)u2
shock (q 2  1).
The pressure distribution is obtained by integrating the
Thus, in view of relation (32), the solution of second equation of system (9) using the boundary con-
the problem is determined through one unknown func-
dition for the pressure (12):
tion q(R).
p ρ2 r2
To find the function q(R), we use the integral law h= =1− [(ε7 + ε13 )ϕ1 (x)
of conservation of energy (16). Substituting the ob- p2 p2
tained functions ρ̄, ū, p̄, ρ, u, and p into (16), calculat-
+ (ε8 + ε15 )ϕ2 (x) + (ε9 + ε12 )ϕ3 (x)
ing the quadratures, and performing elementary trans-
formations, we write the following system of ordinary + (ε10 + ε14 )ϕ4 (x) + ε11 ϕ5 (x) + ε15 ϕ6 (x)]. (39)
differential equations for the functions q(R) and τ (R):
The functions on the right side of (39) are not given here
dq dτ
= ϕ, = q. (33) because of space limitations; they can be found in [17].
dR dR Solution (36)–(39) includes the unknown functions
The function ϕ is not given here because of space limi- q, α, R∗ , and τ |R. The first equation for the specified
tations (its expression is given in [17]). functions is obtained by substituting the functions (ρ,
The initial conditions for system (33), according to u, h, ρ, u, and h) into the energy integral (16) and per-
(14), have the form forming elementary but rather lengthy transformations:
q(1) = q0 , τ (1) = 0, (34) d2 α d2 q d2 R∗
b51 2
+ b52 2 + b53 = b54 . (40)
where q0 is the measure of the SW strength at the initial dR dR dR2
time (at R = 1), determined from experimental curves The second equation is found by differentiating Eq. (37)
of D(r2 ) or by calculations [1, 12, 19]. with respect to R:
222 Volodin

dα dq dR∗ In the regime of a strong shock, α0 = 2ν/(χ −

ω1 − ω2 ω3 − ω2 ω41
dR dR dR 1) = const, the function q0 is found from experimental
− ω2 (ω42 + ω5 ) = 0 (41) 2
data, for example, [1, 6, 8, 19, 20], R∗0 = R0χ+1 , τ0 =
(ω1 , ω2 , . . . , ω42 are presented in [17]). The third equa- 0
R

tion is obtained using relation (38) and the boundary q0 dR = q0 (R0 − 1), α10 = 0, and the function q10 is
1
condition u(r∗ , t) = u∗ : found from the value of q0 [17], z10 = 2/(χ + 1).
dR∗ dα dq Numerical integration of system (44) with the
= ε85 + ε86 + ε87 , (42) asymptotics at the initial point for a strong shock is
dR dR dR
where ε85 –ε87 are given in [17]. performed by the Runge–Kutta–Fehlberg method with
Differentiating Eqs. (41) and (42) with respect to an automatic choice of the step [21] using as an example
R, we write, together with Eq. (40), the following sys- the data corresponding to explosion of a pressed TNT
tem of equations for the second derivatives of the func- charge [1] in air. Simultaneously, according to the so-
tions q, α, and R∗ |R: lution obtained, the distributions of the parameters in
the perturbed subregions I and II are found from the
d2 α d2 q d2 R∗ corresponding formulas, and the laws of motion of the
b51 + b 52 + b 53 = b54 ,
dR2 dR2 dR2 SW and CS are determined.
d2 α d2 q d2 R∗ The calculations were performed for explosion in
ω1 2 − ω2 ω9 2 − ω2 ω41 = b55 , (43)
dR dR dR2 air (with standard parameters ρ1 = 1.225 kg/m3 , u1 =
d2 α d2 q d2 R∗ 0, and p1 = 1.012 · 105 N/m2 ) for a TNT charge with
ε86 2 + ε87 2 − = b56 , characteristics ρ0 = 1.56·103 kg/m3 , D0 = 7·103 m/sec,
dR dR dR2
where Q0 = 4.2·106 J/kg, χ1 = 1.25, χ = 1.4, r0 = 5.1·10−3 m,
γ = 0.2, β0 = 0.60686, δ = 0.0075, and E0 = 0.228 ·
dα dq
b55 = −ω1 + (ω2 ω3 + ω2 ω3 ) 106 J/kg.
dR dR
As an example, Fig. 1 shows distributions of the
  dR∗  
+ (ω2 ω41 + ω2 ω41 ) + ω2 (ω42 + ω5 ) + ω2 (ω42 + ω5 ), dimensionless density y = ρ/ρ2 , particle velocity z =
dR u/u2 , and pressure h = p/p2 on the dimensionless co-
dα dq
b56 = −ε85 − ε86 − ε87 , ordinate x = r/r2 for cylindrical (ν = 2) and spherical
dR dR (ν = 3) symmetries. The subscripts 1 and 2 refer to
the prime denotes the operator of differentiation with the values of R equal to 50 and 60 (Fig. 1a), 20 and 24
respect to R. (Fig. 1b), respectively. From these figures, it is possible
Solving system (43) for the second derivatives of to determine the boundary between subregions I and
the functions q, α, and R∗ |R, introducing the notation II and their relative extent from the density jump for
dα dq dR∗ the time corresponding to the given values of R. For
= α1 , = q1 , = z1 example, for a cylindrical shock wave at the time cor-
dR dR dR
dτ responding to the position of the SW front R = 50,
and using the equation = q, we obtain the system the boundary between subregions I and II is observed
dR
of equations for the functions q, α, R∗ , τ , α1 , q1 , and in the section x = 0.6, and for R = 60, in the section
z1 |R: x = 0.53, i.e., with time, the relative extent of subre-
gion I decreases, which corresponds to the physics of
dα dq dR∗ dτ
= α1 , = q1 , = z1 , = q, the phenomenon considered.
dR dR dR dR The boundary between subregions I and II and
dα1 Δ1 dq1 Δ2 dz1 Δ3 (44)
= , = , = , their relative extent can also be determined from the
dR Δ dR Δ dR Δ curves of z(x), namely: the above-mentioned boundary
where Δ, Δ1 − Δ3 are presented in [17]. is located in the region in which the linearity of the
For system (44), the initial conditions are the con- curve of z(x) is disturbed during motion from subre-
ditions imposed on the sought functions at the time of gion II to the center of symmetry, since, according to
SW arrival at the charge surface, i.e., for R = 1. How- (25), the velocity in subregion II is distributed linearly.
ever, this point is not a singular point. As the asymp- For example, in the case of spherical symmetry (ν = 3),
totics of the solution near the singular point we use the this picture is observed in the section x = 0.7 at R = 20
values of the corresponding functions from the solution and in the section x = 0.6 at R = 24.
of the problem for a strong shock: α = α0 , q = q0 , A distinctive feature of the curves of h(x) is that
R∗ = R∗0 , τ = τ0 , α1 = α10 , q1 = q10 , z1 = z10 at the curves corresponding to later times are above the
R = R0 , where R0 = 1.001.
Generalized Analysis of Explosion Propagation in Condensed Systems 223

CONCLUSIONS

A unified mathematical model was proposed to de-

scribe the entire sequence of explosion propagation: ini-
tiation by an explosion initiator in an explosive charge
→ shock-wave propagation in the charge with transi-
tion to normal detonation → proper normal detonation
→ expansion of the explosion products in the ambient
medium (air) → and motion of the air blast wave formed
by the expanding explosion products.
The initial-boundary-value problems providing a
mathematical description of the specified sequence of
phenomena were solved using the effective analytical
method proposed by Salamakhin [1, 13] and developed
in the present paper. The main idea of this method
is that the governing equations is supplemented by a
closing equation formulating the assumption of an ex-
ponential density distribution of the perturbed medium
and explosion products. This approach eliminated the
constraint related to the assumption of adiabaticity of
the explosive processes and provided an analytical so-
lution of the problem that made it possible to trace the
time-varying, complex structure of the explosive pro-
cess.

REFERENCES

1. T. M. Salamakhin, Physical Foundations of the

Mechanical Action of Explosion and Methods of
Determining Explosive Loads [in Russian], VIA, Moscow
(1974).
2. K. K. Andreev and A. F. Belyaev, Theory of Explosives,
[in Russian], Oborongiz, Moscow 1960.
3. Yu. B. Khariton, Mechanism of Explosive Reactions [in
Fig. 1. Distributions of the dimensionless density Russian], Goskhimizdat, Leningrad (1932).
y(x), particle velocity z(x), and pressure h(x) in the
case of cylindrical (a) (ν = 2) and spherical (b) flow 4. K. K. Andreev and Yu. B. Khariton, “Some
symmetry (ν = 3). considerations on the mechanism of self-propagating
reactions,” Dokl. Akad. Nauk SSSR, 1, No. 7 ((1934)).
5. M. Cowperthwaite, “Model solutions for the initiation
curves corresponding to early times, i.e., over the entire of condensed explosives,” in: Proc. Symposium of High
range x, the relation h2 (x) > h1 (x) holds. This picture Dynamic Pressures, Paris (1979).
is consistent with the physics of the phenomenon: at 6. K. K. Shvedov and K. K. Dremin, “Nonideal detonation
later times of the range considered, the pressure at the regimes of condensed explosives,” in: Explosive
SW front p2 is lower than that at the early times (even- Engineering (collected scientific papers) [in Russian],
tually, the pressure p2 decreases); hence, the parameter No. 60/17, Nedra, Moscow (1967).
h = p/p2 increases over the entire range x at later time 7. G. T. Volodin, “The study of explosion evolution in
intervals. condensed systems,” Combust., Expl., Shock Waves, 30,
The variations in the parameters of the perturbed No. 2, 243–248 (1994).
medium at the SW front and on the CS found in the 8. V. N. Zubarev, “Motion of explosion products behind
obtained solution are given in [17], together with an the front of a detonation wave,” J. Appl. Mech. Tech.
analysis of the behavior of the corresponding curves. Phys., 2, 45–50 (1965).
224 Volodin

9. V. N. Zubarev and A. A. Evstigneev, “Equations of 16. C. H. Johansson and P. A. Persson, Detonics of High
state of the products of condensed explosive explosions,” Explosives, Academic Press, London–New York (1970).
Combust., Expl., Shock Waves, 20, No. 6, 699–710 17. G. T. Volodin, Action of Explosion of Condensed
(1984). Explosive Charges in Gas and Liquid Media. Part 1.
10. A. N. Dremin, “Discoveries in detonation of molecular Parameters of Detonation and Shock Waves [in
condensed explosives in the 20th century,” Combust., Russian], Levsha, Tula (2003).
Expl., Shock Waves, 6, No. 6, 704–715 (2000). 18. G. T. Volodin, “Stationary detonation of charges
11. R. von Mises, Mathematical Theory of Compressible of condensed systems,” in: Differential Equations
Fluid Flows, Academic Press (1958). and Applied Problems (collected scientific papers) [in
12. L. P. Orlenko, Physics of Explosion [in Russian], Vol. 1, Russian], Levsha, Tula (1999), pp. 37–44.
Fizmatlit, Moscow (2002). 19. F. A. Baum, L. P. Orlenko, K. P. Stanyukovich,
13. N. S. Mel’nikova and T. M. Salamakhin, “Calculation of V. P. Chelyshev, and B. I. Shekhter, Physics of
a point explosion in various gases,” Prikl. Mekh. Tekh. Explosion [in Russian], Nauka, Moscow (1975).
Fiz., No. 4, 155–160 (1964). 20. A. N. Dremin, S. D. Savrov, V. S. Trofimov, and
14. S. M. Guirao and G. G. Bach, “Non-self-similar blast K. K. Shvedov, Detonation Waves in Condensed Media
waves,” Combust., Expl., Shock Waves, 27, No. 3, 342– [in Russian], Nauka Moscow (1970).
351 (1991). 21. V. P. D’yakonov, Handbook on Algorithms and Programs
15. G. T. Volodin, “Distribution of detonation products in BASIC for personal computers [in Russian], Moscow
parameters of condensed explosives,” Combust., Expl., (1987).
Shock Waves, 27, No. 1, 117–121 (1991).