Reaction rates and reaction rate constant conception.

One-temperature
case
Evgeniy G. Kolesnichenko and Yuriy E. Gorbachev

Citation: AIP Conf. Proc. 1501, 107 (2012); doi: 10.1063/1.4769483

View online: http://dx.doi.org/10.1063/1.4769483
View Table of Contents: http://proceedings.aip.org/dbt/dbt.jsp?KEY=APCPCS&Volume=1501&Issue=1
Published by the American Institute of Physics.

Related Articles
On the r-matrix structure of the hyperbolic BCn Sutherland model
J. Math. Phys. 53, 123528 (2012)
Explicit system-bath correlation calculated using the hierarchical equations of motion method
J. Chem. Phys. 137, 194106 (2012)
Multi-continuum Fano resonance in coupled quantum point contacts: A manifestation of the “integral” Fano
formula
J. Appl. Phys. 112, 103704 (2012)
Liouville type theorems for nonlinear elliptic equations involving operator in divergence form
J. Math. Phys. 53, 103706 (2012)
An integrable system and associated integrable models as well as Hamiltonian structures
J. Math. Phys. 53, 103508 (2012)

Additional information on AIP Conf. Proc.

Journal Homepage: http://proceedings.aip.org/
Journal Information: http://proceedings.aip.org/about/about_the_proceedings
Top downloads: http://proceedings.aip.org/dbt/most_downloaded.jsp?KEY=APCPCS
Information for Authors: http://proceedings.aip.org/authors/information_for_authors

Downloaded 17 Dec 2012 to 81.23.105.75. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 Reaction Rates and Reaction Rate Constant Conception.
One-Temperature Case
Evgeniy G. Kolesnichenko∗ and Yuriy E. Gorbachev†
∗
Gas Kinetics Lab, Moscow State University, Institute for Mechanics, Moscow, 117192, Russia
†
Research Department, Geolink Technologies LLC, St. Petersburg, 197342, Russia

Abstract. The new method of getting a normal solution for the generalized Boltzmann equation for reacting gas mixtures
was proposed. It is based on the following items: (i) slow variables are introduced via approximate summational invariants,
defined within the method, (ii) kinetic equations are presented in the form of a singularly perturbed system for gas-dynamic
(slow) variables and for the "fast" part of the distribution function, (iii) collisional integral is not expanded into the series over
the Knudsen number (no assumption is made that the part of the collisional integral, responsible for the chemical reactions,
can be treated as the perturbation of its "elastic" part). While deriving the gas-dynamic equations it is shown that the role
of non-equilibrium effects is much more essential in our approach, than is generally accepted. By non-equilibrium effects
we mean all kind of effects caused by deviation of the distribution function from its quasi-equilibrium value. From several
examples it was shown that non-equilibrium corrections and the traditional equilibrium rate constants could be of the same
order of magnitude. In this paper we derive expressions for corrections to equilibrium rate constants for arbitrary mixtures
and corresponding integral equations for corrections to the quasi-equilibrium distributions. In situations where corrections to
the reaction rates are not small non-equilibrium effects dramatically impacts the chemical kinetics of the reacting gas mixture.
This leads to the necessity of the revision of the concept of getting information on the reaction rates from the experiments.
Keywords: Kinetic theory, reactive mixture, transport equations, non-equilibrium effects
PACS: 05.20.Dd, 05.60 +w, 51.10+y, 82.20. Mj

INTRODUCTION
A new method of deriving gas-dynamic equations from the kinetic equations for chemically reacting gas mixtures
suggested in [1] allows to overcome the problems arising while using the traditional method, such as inability to
describe the experimentally observed transition of the mechanism of dissociation of polyatomic molecules from the
monomolecular one at high pressure to the bimolecular one at low pressure. The resulting gas-dynamic equations differ
significantly from those obtained on the basis of hypothesis of local quasi-equilibrium. All differences we have named
as non-equilibrium effects. Physically they originate from the high sensitivity of the high-threshold physical-chemical
processes to the distribution function perturbations. These effects can be subdivided into three groups. The first group
contains the effects caused by the perturbation of quasi-equilibrium distribution function by the physical-chemical
processes. Only these effects remain in the spatially homogeneous case. The second group consists of the terms
proportional to the velocity divergence that arise in the expressions for the corresponding rates of physical-chemical
processes. These effects are caused by the impact of expansion-compression of the gas on physical-chemical processes.
The first and second groups are represented in the zero order approximation. In the first order approximation the third
group of non-equilibrium effects appear, which consists of terms in the expressions for the reaction rates, that are
proportional to the scalar bilinear combinations of gradients of gas-dynamic variables. Due to the additive contribution
of all these terms to the gas-dynamic equations, they can be studied separately. The effects from the first group are
estimated in [2]. The effects from the second and third group have been estimated in [3–5]. In all cases corresponding
additional terms appear to be of the same order of magnitude as terms calculated with a quasi-equilibrium distribution
function.
The paper is organized as follows. In the next section we describe the basic items of our method of solving
the generalized Boltsmann equation for reacting gases. The general expression for reaction rates are derived in the
third Section. More detailed analysis of the reaction rates for one-temperature case is done in the forth Section. In
Conclusions we summarize the analysis and trace further research.

28th International Symposium on Rarefied Gas Dynamics 2012

AIP Conf. Proc. 1501, 107-114 (2012); doi: 10.1063/1.4769483
© 2012 American Institute of Physics 978-0-7354-1115-9/$30.00

107

Downloaded 17 Dec 2012 to 81.23.105.75. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 STATE-OF-THE-ART

Method Description
Following paper [1] we consider the reacting gas mixture described by the generalized Boltzmann equation
∂ Fα 1
+ vα · ∇Fα = Iα (F). (1)
∂t ε
Here Fα , Iα and vα is the distribution function, the collisional integral and the velocity of the molecule of sort α
respectively; ε is the ratio of characteristic time of inelastic process to characteristic gas-dynamic time. For reacting
gas mixture the collisional integral can be represented as [2, 6]
 
 
|ν  |≤2∨|ν |≤1
 ν Fβ Fα Fβ
Iα = ∑ ∑ dvν dvν  Wα ,ν (vα , kα , vν , kν |vν  , kν  ) ∏ sβ (kβ ) − sα (kα ) ∏ sβ (kβ ) . (2)
ν ,ν  kν ,k  ν β ∈ν β ∈ν


Here W is the transition probability that satisfies the microscopic reversibility law Wνν (vν , kν |vν  , kν  ) =
Wνν (vν  , kν  |vν , kν ); ν characterizes the reaction channel: ν = {ν1 , ..., να , ...}, where να is a number of particles
of sort α in channel ν ; |ν | = ν1 + ... + να + ... is a total number of particles in channel ν ; kν = {kα : α ∈ ν } and
kα are quantum numbers of species α , the same meaning has the subindex at vν ; the summation in (2) is carried out
over all pairs of channels for which the number of particles in the initial or final channel does not exceed two (with
arbitrary number of particles in opposite channel); sβ (kβ ) is a statistical weight of state kβ . Primes over sums and
integrals mean that the corresponding summation and integration was done in a way to avoid multiple account of the
same states.
The method of introduction of slow (gas-dynamic) variables is non-trivial for the systems for which a number
of slow variables exceeds the number of summational invariants [7]. Therefore the description of gases with internal
degrees of freedom appears to be a challenge to the theory. This problem can be solved by generalizing the definition of
slow variables. They should be determined not only by summational invariants, but also by approximate summational
invariants (ASI) ψi [1], that are defined by the following relationship

ψi , I (F) ≤ O(ε ), ψ , ϕ  = ∑ ∑ dvα ψα (vα , kα ) ϕα (vα , kα ) . (3)
α kα

Here the angle brackets denote the scalar product. The corresponding set of slow variables is determined as
Γi = ψi , F . (4)
According to this definition variables Γi evolve over a long time scale. Equations for slow variables are derived by
multiplying the Eq. (1) by ASI and integration over velocities and summation over quantum numbers.
(qe)
Following our previous publication [1], we represent Fα = Fα + Φα , where the first item does not depend on
(qe) (qe)
time and coordinates explicitly, i.e. it depends only on slow variables Fα = Fα ({Γi }). For quasi-equilibrium
(qe)
distribution functions Fα we chose the function that maximizes the entropy density for a fixed set of slow variables:
(qe)   (qe) is substituted
Fα = exp ln sα (kα ) − ∑M i=1 γi ψiα (vα , kα ) . Here γi are determined by the relation (4), when F
instead of F. Φα should satisfy relations
ψi , Φ = 0. (5)
The advantage of such quasi-equilibrium distribution function is that it naturalize the procedure of introduction of
generalized thermodynamics (multi-temperature for instance).
Using the new variable Φ the Eq. (1) together with the corresponding equation for slow variables can be transformed
into the singularly perturbed set of equations
(qe)  

∂ Φα M
∂F
ε = Iα (F (qe) + Φ) − ∑ α ψi , I F (qe) + Φ
∂t i=1 ∂ Γi
(6)
M (qe)
∂ Fα  

−ε v · ∇Φα + ε ∑ ∇ · vψi , F (qe) + Φ − v · ∇Γi ,
i=1 ∂ Γi

108

Downloaded 17 Dec 2012 to 81.23.105.75. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 ∂ Γi 1 
 

= ψi , I F (qe) + Φ − ∇· vψi , F (qe) + Φ . (7)

∂t ε
For its solution we use the modified Chapman-Enskog method described in [1]. Equation (6) coincides with that
commonly used only where all ψi are exact summational invariants (ESI).

Gas-Dynamic Equations
Considering the one-velocity case, one assumes that the following ASI satisfy condition (3)
(T )
ψnβ (α ) = δα ,β , ψp (α , vα , kα ) = mα vα , ψE (α , vα , kα ) = eα (vα ) + eα (kα ), ψi = ψi (kα ), (8)

(T )
where δα ,β is the Kronecker symbol, eα = mα v2α /2 and eα are translational and internal energies of the molecules of
sort α respectively. Corresponding gas-dynamic variables defined according to (4) are number densities of species nα ,
mean mixture momentum Γp = p = ρ u, where ρ = ∑α mα nα and u is a mixture mean velocity, and mean total energy
ΓE = ΓE (t) + ΓE (k) + ΓE (i) , ΓE (t) = 3n/2γE . ΓE (k) = ρ u2 /2, ψi are some additional scalar ASI (vibrational energy, etc)
that generate the corresponding gas-dynamic variables Γi .
Further we consider so-called "weak" non-equilibrium situation when Φα function is assumed to be of the order of
(qe)
ε . Collisional integral Iα (F) calculated with Fα does not vanish since this distribution contains not only exact sum-
∞  
mational invariants. So, using the representation for Φ in a form of series expansion Φ = ∑ ε s Φ(s) Γ, ∇Γ, ..., ∇s+1 Γ ,
s=0
the equation for the lowest order correction for the distribution function, Φ(0) (see Eq. (22) in paper [1]), can be written
as
1  M ∂ F (qe)  

JF (qe) (F (qe) )Φ(0) = − JF (qe) F (qe) + ∑ v · ∇Γi − ∇ · vψi , F (qe) . (9)
ε i=1 ∂ Γi

Here JF (G) is the linearized modified collisional operator JF (G) and
M
∂F
JF (G) = I(G) − ∑ ψi , I(G). (10)
∂
i=1 Γi

All ASI are exact invariants of JF (G) and JF (G) so all orthogonality conditions can be satisfied. From the structure of
the Eq. (9) it follows that Φ(0) can be represented as a sum

(0) (0) (0) (0) 1 (0)
Φα = Φ1α + Φ2α ∇ · u + Φ3α cα cα − c2α : ∇u + Φ4α cα · ∇E
3
(11)
N M
+ ∑ Φni α cα · ∇ni + ∑ ΦΓi α cα · ∇Γi .
i=1 i=N+5

This allows derivation of zero-order (Euler) gas-dynamic equations [1]

∂ nα
+ ∇ · (unα ) = Rα (ΓE , u), α = 1, ...N,
∂t
(12)
 
  

(0) (0) (qe) (ne) (0)

Rα = ψnα , I F (qe) + ψnα , I  F (qe) Φ1 + Φ2 ∇ · u ≡ Rα + Rα + R2 α ∇ · u,

∂ ρu 1  2 (qe)

+ ∇ · (ρ uu) + ∇ p = 0, p= mc , F = n/γE , (13)

∂t 3
∂ ΓE
+ ∇ · ((ΓE + p) u) = 0, (14)
∂t

109

Downloaded 17 Dec 2012 to 81.23.105.75. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 ∂ Γi
+ ∇ · (uΓi ) = Ri , i = N + 5, ..., M,
∂t
(15)
 
  

(0) (0) (qe) (ne) (0)

Ri = ψΓi , I F (qe) + ψΓi , I  F (qe) Φ1 + Φ2 ∇ · u ≡ Ri + Ri + R2 i ∇ · u.
(sh) (qe) (ne)
Here Rα are the reaction rates. The sum of the first two terms in the RHS of Eq. (12), Rα = Rα + Rα , describe the
(sih) (0)
kinetics of chemical reactions in spatially homogeneous systems, the last term, Rα = R2 α ∇ · u, describes the impact
of spatial inhomogeneity. Ri are the relaxation rates for Γi variables. Similarly, the sum of the first two terms in the
(sh) (qe) (ne)
RHS of Eq. (15), Ri = Ri + Ri , describe the relaxation in spatially homogeneous systems, while the last term
(sih) (0)
Ri = R2 i ∇ · u arises in spatially inhomogeneous systems. It should be emphasized that in contrast to the traditional
method our approach does not use any assumptions additional to that the ratio of fast process characteristic time to
the gas-dynamic one is of small value. For example we don’t use the expansion of the collisional integral over the
Knudsen number as done since the publication [8]. This means that obtained gas-dynamic equations are of the most
general form [1].

GENERAL EXPRESSIONS FOR REACTION RATES AND RELAXATION RATES

Spatially Homogeneous Case

Equation for the "scalar" part of the perturbation function (11), that is the only one that contributes the reaction rates
in spatially homogeneous case can be written as [1]
  
(0)
JF (qe) (F (qe) )Φ1 = −JF (qe) α F (qe) . (16)
α

Let us introduce the generalized relaxation rates as

 
 

R = {RΓ j } = ψ j , I F (qe) + ψ j , I  F (qe) Φ(0) = {Rp , RE , Rα , Ri } = R(sh) + R(sih) . (17)

The spatially homogeneous/inhomogeneous parts of the reaction/relaxation rates R(sh) are defined in the text after Eq.
(15). Then, using the definitions of the modified collisional operator (10) and that Rp = 0 and RE = 0, Eq. (16) can be
written as
  M−4
∂ Fα (qe) (sh)
(0)
I  (F (qe) )Φ1 = −Iα (F (qe) ) + ∑ Rj . (18)
α j=1 ∂ Γ j

(0)
From Eq. (18) and conditions (5) Φ1 can be determined as a unique solution. Since this equation is linear, this solution
should have the form
M−4
(0) (0) (0) (sh)
Φ1 = Φ1(0) + ∑ Φ1 j R j , (19)
j=1

where all Φ’s in the RHS are determined while solving Eq. (18). After substituting this expression into Eq. (17) we
(sh)
obtain the equation for R j
(ne)
R(sh) = R(qe) + R(0) + AR(sh) , (20)
 

(ne) (0) (0)
where R(0) is determined by the term Φ1(0) from Eq. (19) and A = {ai j } = ψi , I  F (qe) Φ1 j . Solution of Eq.
(20) can be written as 
(ne)
R(sh) = (I − A)−1 R(qe) + R(0) . (21)

110

Downloaded 17 Dec 2012 to 81.23.105.75. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 Spatially Inhomogeneous Case
Equation for the "divergency" parts of function (11) that is responsible for the spatially inhomogeneous effects can
be derived by gathering terms proportional to ∇u in Eq. (9) after substituting of expression (11)
  M−4 (qe)
∂ Fα M
∂F ∂F
(qe)
γE mα c2α (qe)
(qe)
(0) (0)
I  (F (qe) )Φ2 = ∑ R 2 j − ∑ α Γi − p α + Fα . (22)
α j=1 ∂Γj i ∂ Γi ∂ ΓE 3

(0)
Due to its linearity its solution can be represented in a way, similar to that obtained for Φ1

M−4
(0) (0) (0) (0)
Φ2 = Φ2(0) + ∑ Φ2 j R2 j . (23)
j=1

(0)
Using corresponding part of Eq. (17) we get the equation for R2 j

(0) (0) (0)

R2 = R(0) + A2 R2 , (24)
 

(0) (0) (0)
where R(0) is determined by the term Φ2(0) from Eq. (23) and A2 = {a2 i j } = ψi , I  F (qe) Φ2 j . Thus

(0) (0)
R2 = (I − A2 )−1 R(0) , (25)

that finally leads to the following expression for the reaction rate

(ne) (0)
R = (I − A)−1 R(qe) + R(0) + (I − A2 )−1 R(0) ∇u. (26)

This expression shows that all relaxation characteristics (reaction rates and relaxation rates) are linear combinations
of each other that mean the interdependency of chemical and relaxation processes. In the traditional approach the
relaxation rate can be separated into two terms [10]. The first is responsible for the gas-dynamic variables evolution,
due to their relaxation without impact of chemical reactions. The second is responsible for its evolution due to chemical
reactions. In our approach chemical reaction rates contain terms that are determined by each relaxation process via
corresponding rates.

REACTION RATES FOR ONE-TEMPERATURE CASE

For one-temperature case the only gas-dynamic variables that correspond to ASI are species number densities. Thus
no relaxation rates arise in gas-dynamic equations. Such simplifications allow for some new steps in deriving general
expressions for the reaction rates. First of all it is convenient to consider chemical processes in terms of individual
reactions
∑ νβ Mβ ←→ ∑ νβ Mβ , (27)
β ∈ν  β ∈ν 

where νβ are stoichiometric coefficients, i.e. a number of particles of sort β in channel ν , and ν  is a set of colliding
particles that consists of particles of sort α and all their partners ν (see Eq. (2)).
Let us denote the pair of individual reversible reactions by index r. Then the reaction rates can be represented in
terms of individual reaction rates Rα r
Rα = ∑(να r − να r )Rr . (28)
r

where να describes reagents, while να describes products. Taking into account factor να r − να r we can make
summation in equation (28) over all reactions in the system.

111

Downloaded 17 Dec 2012 to 81.23.105.75. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 Spatially Homogeneous Case

Let us represent the expression for Iα (F (qe) ) from the Eq. (18) in a way

 |ν |≤1∨|ν  |≤2
Iα F (qe) = ∑ ξα ,ν ;ν  (vα , kα )ηα , ν ;ν  ≡ ∑ ξα r (vα , kα )ηr , (29)
ν  ,ν r

 
  exp (−γE eν )
ξα r (vα , kα ) = ∑ dvν dvν  Wαν,ν (vα , kα , vν , kν |vν  , kν  ) ,
kν ,kν  ∏ Qβ (γE ) (30)
β ∈ν 

(e) (e)
ηr = ∏ nβ − Kc r ∏ nβ , Kc r = ∏ Qβ (γE ) ∏ Qβ (γE ). (31)
β ∈ν  β ∈ν  β ∈ν  β ∈ν 
 
Here we use the energy conservation law eν  = eν  , where eν = ∑ mβ c2β /2 + eβ (kβ ) is the energy of molecules
β ∈ν
(e)
participating in chemical reaction in channel ν . Kc r is the traditional equilibrium constant expressed in terms of
number densities for reaction r. Thus for the equilibrium part of the reaction rate (12) we have
|ν |≤2∨|ν  |≤2 
(qe) (qe)
Rα = ∑
 
∑ dvα ξα ,ν ;ν  (vα , kα )ηα ,ν ;ν  = ∑(να r − να r )Rr . (32)
ν ,ν kα r

Then we can write

(qe) (qe)
Rr = Kr ηr ,
 
(qe)   exp (−γE eν ) (33)
Kr = ∑ dvν  dvν  Wνν (vν  , kν  |vν  , kν  ) .
kν  ,kν  ∏ Qβ (γE )
β ∈ν 
(qe)
Here Kr is the traditional equilibrium rate constant of the forward reaction (27).
(ne) (0)
Now let us consider Rα . Using expressions (29)-(31) we can transform the Eq. (16) for Φ1 into the expression
  
  M (qe)
∂ Fα
(0)
 (qe)
JF (qe) (F )Φ1 = −∑ ξα r (vα , kα ) − ∑ ψi , ξα r (vα , kα ) ηr .
α r i=1 ∂ Γi

(0)
Thus we can conclude that Φ1α can be represented in the form of the sum of terms, proportional to ηr
(0) (qe)
Φ1α = Fα ∑ ϕ1α r (vα , kα )ηr . (34)
r

Using that the momentum and total energy are the exact summational invariants, i.e. that corresponding scalar products
vanish, for ϕ1α r we get the equation
  N
∂ Fα
(qe)
(qe)
JF (qe) (F (qe) )F (qe) ϕ1r
α
= −ξα r (vα , kα ) + ∑ ∂ nβ
(νβ r − νβr )Kr , (35)
β =1

(qe) (qe)
where (νβ r − νβr )Kr = ψnβ , ξα r (vα , kα ) and thus ϕ1α r is a function of the whole set of Kα r . Then
 
  
(ne) (0) (qe)
Rα = ψnα , I  F (qe) Φ1 = ψnα , I  F (qe) Fβ ∑ ϕ1β r (vβ , kβ )ηr
r
(36)
 

(qe) (ne)
= ∑ ψnα , I  F (qe) Fβ ϕβ r (vβ , kβ ) ηr = ∑ K̃α r ηr ,
r r

112

Downloaded 17 Dec 2012 to 81.23.105.75. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
 (ne)
where after some algebra the following expression for K̃α r can be obtained
⎛ ⎞
(ne)
K̃α r =∑ (να r − να r ) ⎝krr
+
 ∏ −
nβ − krr  ∏ nβ ⎠ . (37)
r β ∈νr β ∈νr

Here  
 ν  exp (−γE eν  )
+
krr  = ∑ dvν  dvν  Wν r (vν  , kν  |vν  , kν  ) ∑ ϕ1β r ,
kν  ,kν  r ∏ Qβ β ∈ν  r
(38)
β ∈νr
 
 ν  exp (−γE eν  )
−
krr  = ∑ dvν  dvν  Wν r (vν  , kν  |vν  , kν  ) ∑ ϕ1β r .
kν  ,kν  r ∏ Qβ β ∈ν  r
(39)
β ∈νr

As a result, for the part of the reaction rate responsible for the spatially homogeneous situation, we obtain
 
(sh) (qe) (ne)
Rα = ∑ (να r − να r )Kr + K̃α r ηr . (40)
r

(ne)
Since K̃α r is a function of all equilibrium rate constants, all species affect this reaction rate, not only those which
(qe) (ne)
directly participate in the reaction. Let us call (να r − να r )Kr + K̃α r as a "non-equilibrium rate constant" of forward
reaction. It is important to note that the "scalar" non-equilibrium effects does not shift the chemical equilibrium, that
is described by equation ηr = 0 for both equilibrium and "scalar" non-equilibrium cases in spatially homogeneous
situation.

Spatially Inhomogeneous Case

(sih) (0)
Spatially inhomogeneous part of the reaction rate, Rα , is determined by Φ2 that is the solution of the Eq. (22).
(0) (0) (qe)
After representing Φ2α in a way Φ2α = Fα ϕ2α we obtain
 
(0) (0)
R2 α = ∑(να r − να r )R2 r = ∑(να r − να r ) K2+r ∏ nβ − K2−r ∏ nβ . (41)
r r β ∈ν  β ∈ν

Here    
  exp [−γE eν  ]
K2+r = ∑ dvν  dvν  Wνν (vν  , kν  |vν  , kν  ) ∑ ϕ2β ,
kν  ,kν  ∏ Qβ (γE ) β ∈ν 
β ∈ν 
   
  exp [−γE eν  ]
K2−r = ∑ dvν  dvν  Wνν (vν  , kν  |vν  , kν  ) ∑ ϕ2β .
kν  ,kν  ∏ Qβ (γE ) β ∈ν 
β ∈ν 

From these expressions we can conclude that inhomogeneous parts of reaction rates contains only those reactions, in
which component α participates directly.
From the expression (41) it is seen that the spatially inhomogeneous part of the reaction rate does not vanish, when
the species concentrations possess their equilibrium values, as defined traditionally. Let us introduce effective rate
constants 
(e f f )+ (qe)+ (ne)+
Kα r = (να r − να r ) Kr + K2+r ∇ · u + K̃α r , (42)
and 
(e f f )− (qe)− (ne)−
Kα r = (να r − να r ) Kr + K2−r ∇ · u + K̃α r , (43)
(qe)+ (qe)− (e) (qe) (ne)+ (ne) (ne)− (e) (ne)
where Kr = Krqe and Kr = Kcr Kr , while K̃α r = K̃α r and K̃α r = Kcr K̃α r . From here and from the
(ne)
expression (37) for K̃α r it is seen that K (e f f )± depends not only on the velocity divergence, but on concentrations

113

Downloaded 17 Dec 2012 to 81.23.105.75. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions
of all species, participating in the pair of forward/reverse reactions. We can also mention that due to a spatial non-
equilibrium, a chemical equilibrium in the system is shifted. It should be emphasized that we don’t use the expansion
of collisional integral over the Knudsen number. This means that all terms in expressions (42) and (43) are generally
of the same order. As a consequence K (e f f )± can vanish, i.e. reactions can be blocked not only via "scalar" non-
equilibrium effects, as it was shown in paper [2], but via spatial inhomogeneity effects as well.

CONCLUSION
Studies of this paper show that via kinetics of high threshold physical-chemical processes all generalized relaxation
processes described by Euler equations became entangled (affecting each other). By generalized relaxation processes
here we mean all processes that are described by gas-dynamic variables related to approximate summational invariants.
Evolution equations for these variables contain source terms that determine coupling of their behavior with the evo-
lution of other variables. Derived gas-dynamic equations have more symmetrical form in comparison with traditional
ones from the point of view of description of mutual impact of the above mentioned gas-dynamic variables.
At present on the basis of the assumption that different chemical reactions are independent of each other it is
assumed that under special conditions it would be possible to study kinetics of the one separated process. Then, using
those obtained results, it would be possible to study other processes. Results obtained in this paper show that such
an approach is not universal. In general for each mixture we should built its own model. Aiming this we propose
to calculate, not the rate constants, but all integral brackets arising in the solutions of derived integral equations for
corrections to the quasi-equilibrium distribution function. Expansion over the compleat set of orthogonal polynomials
is assumed to be used while getting the solutions of integral equations.
We propose the following steps for further work: (1) Choosing the suitable set of polynomials, (2) Integral brackets
calculation, (3) Tabulating of the generating functions for generalized thermodynamics.

REFERENCES
1. E. G. Kolesnichenko, and Yu. E. Gorbachev, Appl. Math. Model. 34, 3778–3790 (2010).
2. E. G. Kolesnichenko, and Yu. E. Gorbachev, Physical-chemical kinetics in gas dynamics 12, URL:http://www.chemphys.
edu.ru/pdf/2011-06-16-001.pdf (2011).
3. E. G. Kolesnichenko, S. A. Losev, Khimia plasmi 7, Atomizdat, Moscow 1979.
4. G. Ya. Gerasimov, E. G. Kolesnichenko, Pisma v jurnal tehnicheskoi physiki 7, 16, 965–967 (1981).
5. G. Ya. Gerasimov, E. G. Kolesnichenko, Fluid Dyn. 18, 794–800 (1983).
6. E. G. Kolesnichenko, Kinetic equations of the theory of chemically reactive gases, Moscow State University 1983.
7. N. G. van Kampen, J. Stat. Phys. 46, 709–727 (1987).
8. I. Prigogine, E. Xhrouet, Physica 15, 913–932 (1949).
9. E. V. Stupochenko, S. A. Losev, A. I. Osipov, Relaxation in Shock Waves, Springer-Verlag, New York 1967.
10. B. F. Gordietz, A. I. Osipov, L. P. Shelepin, Kinetic processes in gases and molecular lasers, Nauka, Moscow 1980.

114

Downloaded 17 Dec 2012 to 81.23.105.75. Redistribution subject to AIP license or copyright; see http://proceedings.aip.org/about/rights_permissions