Nonlinear transport phenomena:
            models, method of solving and unusual
                          features

                              Vsevolod Vladimirov

              AGH University of Science and technology, Faculty of Applied
                                     Mathematics

                                 ´
                             Krakow, August 10, 2010




Summer School: KPI, 2010        Nonlinear transport phenomena                1 / 38
THE AIM OF THIS LECTURE IS TO PRESENT THE
        SIMPLEST PATTERNS SUPPORTED BY TRANSPORT
        EQUATIONS


        WHAT ARE PATTERNS?




Summer School: KPI, 2010   Nonlinear transport phenomena   2 / 38
THE AIM OF THIS LECTURE IS TO PRESENT THE
        SIMPLEST PATTERNS SUPPORTED BY TRANSPORT
        EQUATIONS


        WHAT ARE PATTERNS?




Summer School: KPI, 2010   Nonlinear transport phenomena   2 / 38
Examples of patterns:
              Solution to the system of reaction-diffusion equation

                                          ut = Du ∆u + a(1 − u)
                                         vt = Dv ∆v + u v 2 − k v
                               u(0, x) = f (x),      v(0, x) = g(x),
                           2   ∂2
              ∆=           i=1 ∂x2
                                 i
              periodically initiated cordial impulse; formation of the
              train of solitary waves;
              tsunami;
              tornado
              etc., etc.



Summer School: KPI, 2010             Nonlinear transport phenomena       3 / 38
Examples of patterns:
              Solution to the system of reaction-diffusion equation

                                          ut = Du ∆u + a(1 − u)
                                         vt = Dv ∆v + u v 2 − k v
                               u(0, x) = f (x),      v(0, x) = g(x),
                           2   ∂2
              ∆=           i=1 ∂x2
                                 i
              periodically initiated cordial impulse; formation of the
              train of solitary waves;
              tsunami;
              tornado
              etc., etc.



Summer School: KPI, 2010             Nonlinear transport phenomena       3 / 38
Examples of patterns:
              Solution to the system of reaction-diffusion equation

                                          ut = Du ∆u + a(1 − u)
                                         vt = Dv ∆v + u v 2 − k v
                               u(0, x) = f (x),      v(0, x) = g(x),
                           2   ∂2
              ∆=           i=1 ∂x2
                                 i
              periodically initiated cordial impulse; formation of the
              train of solitary waves;
              tsunami;
              tornado
              etc., etc.



Summer School: KPI, 2010             Nonlinear transport phenomena       3 / 38
Examples of patterns:
              Solution to the system of reaction-diffusion equation

                                          ut = Du ∆u + a(1 − u)
                                         vt = Dv ∆v + u v 2 − k v
                               u(0, x) = f (x),      v(0, x) = g(x),
                           2   ∂2
              ∆=           i=1 ∂x2
                                 i
              periodically initiated cordial impulse; formation of the
              train of solitary waves;
              tsunami;
              tornado
              etc., etc.



Summer School: KPI, 2010             Nonlinear transport phenomena       3 / 38
Examples of patterns:
              Solution to the system of reaction-diffusion equation

                                          ut = Du ∆u + a(1 − u)
                                         vt = Dv ∆v + u v 2 − k v
                               u(0, x) = f (x),      v(0, x) = g(x),
                           2   ∂2
              ∆=           i=1 ∂x2
                                 i
              periodically initiated cordial impulse; formation of the
              train of solitary waves;
              tsunami;
              tornado
              etc., etc.



Summer School: KPI, 2010             Nonlinear transport phenomena       3 / 38
Examples of patterns:
              Solution to the system of reaction-diffusion equation

                                          ut = Du ∆u + a(1 − u)
                                         vt = Dv ∆v + u v 2 − k v
                               u(0, x) = f (x),      v(0, x) = g(x),
                           2   ∂2
              ∆=           i=1 ∂x2
                                 i
              periodically initiated cordial impulse; formation of the
              train of solitary waves;
              tsunami;
              tornado
              etc., etc.



Summer School: KPI, 2010             Nonlinear transport phenomena       3 / 38
Examples of patterns:
              Solution to the system of reaction-diffusion equation

                                          ut = Du ∆u + a(1 − u)
                                         vt = Dv ∆v + u v 2 − k v
                               u(0, x) = f (x),      v(0, x) = g(x),
                           2   ∂2
              ∆=           i=1 ∂x2
                                 i
              periodically initiated cordial impulse; formation of the
              train of solitary waves;
              tsunami;
              tornado
              etc., etc.



Summer School: KPI, 2010             Nonlinear transport phenomena       3 / 38
Our aim to discuss the patterns supported by the various
        transport equations in case of one spatial variable.



        Advantages:
           ability to present and analyze the analytical solutions to
           non-linear models
           the possibility to analyze the problem by means of
           qualitative theory methods.

        By patterns we temporarily mean:
            non-monotonic solution to the modelling system,
            maintaining their shape during the evolution;
            or the non-monotonic solutions evolving in a self-similar
            mode
            or the non-monotonic solutions tending to infinity in finite
            time (blowing-up patterns)
Summer School: KPI, 2010     Nonlinear transport phenomena               4 / 38
Our aim to discuss the patterns supported by the various
        transport equations in case of one spatial variable.



        Advantages:
           ability to present and analyze the analytical solutions to
           non-linear models
           the possibility to analyze the problem by means of
           qualitative theory methods.

        By patterns we temporarily mean:
            non-monotonic solution to the modelling system,
            maintaining their shape during the evolution;
            or the non-monotonic solutions evolving in a self-similar
            mode
            or the non-monotonic solutions tending to infinity in finite
            time (blowing-up patterns)
Summer School: KPI, 2010     Nonlinear transport phenomena               4 / 38
Our aim to discuss the patterns supported by the various
        transport equations in case of one spatial variable.



        Advantages:
           ability to present and analyze the analytical solutions to
           non-linear models
           the possibility to analyze the problem by means of
           qualitative theory methods.

        By patterns we temporarily mean:
            non-monotonic solution to the modelling system,
            maintaining their shape during the evolution;
            or the non-monotonic solutions evolving in a self-similar
            mode
            or the non-monotonic solutions tending to infinity in finite
            time (blowing-up patterns)
Summer School: KPI, 2010     Nonlinear transport phenomena               4 / 38
Our aim to discuss the patterns supported by the various
        transport equations in case of one spatial variable.



        Advantages:
           ability to present and analyze the analytical solutions to
           non-linear models
           the possibility to analyze the problem by means of
           qualitative theory methods.

        By patterns we temporarily mean:
            non-monotonic solution to the modelling system,
            maintaining their shape during the evolution;
            or the non-monotonic solutions evolving in a self-similar
            mode
            or the non-monotonic solutions tending to infinity in finite
            time (blowing-up patterns)
Summer School: KPI, 2010     Nonlinear transport phenomena               4 / 38
Our aim to discuss the patterns supported by the various
        transport equations in case of one spatial variable.



        Advantages:
           ability to present and analyze the analytical solutions to
           non-linear models
           the possibility to analyze the problem by means of
           qualitative theory methods.

        By patterns we temporarily mean:
            non-monotonic solution to the modelling system,
            maintaining their shape during the evolution;
            or the non-monotonic solutions evolving in a self-similar
            mode
            or the non-monotonic solutions tending to infinity in finite
            time (blowing-up patterns)
Summer School: KPI, 2010     Nonlinear transport phenomena               4 / 38
Our aim to discuss the patterns supported by the various
        transport equations in case of one spatial variable.



        Advantages:
           ability to present and analyze the analytical solutions to
           non-linear models
           the possibility to analyze the problem by means of
           qualitative theory methods.

        By patterns we temporarily mean:
            non-monotonic solution to the modelling system,
            maintaining their shape during the evolution;
            or the non-monotonic solutions evolving in a self-similar
            mode
            or the non-monotonic solutions tending to infinity in finite
            time (blowing-up patterns)
Summer School: KPI, 2010     Nonlinear transport phenomena               4 / 38
Heat transport: the balance equation


             d
                      ρ u(t, x)d x = −         q dσ +         f (u(t, x); t, x)d x,   (1)
             dt   Ω                       ∂Ω              Ω

        where q is the density of the heat flux on the boundary ∂ Ω.
        On virtue of the Fick law, is as follows:

                                      q = −κ        u(t, x);



            u(t, x)        is the energy per unit mass (the temperature);
        ρ      is the heat capacity per unit volume;
             κ(u; t, x)       is the heat transport coefficient;
            f (u(t, x); t, x)     is the voluminal heat source.

Summer School: KPI, 2010            Nonlinear transport phenomena                           5 / 38
Heat transport: the balance equation


             d
                      ρ u(t, x)d x = −         q dσ +         f (u(t, x); t, x)d x,   (1)
             dt   Ω                       ∂Ω              Ω

        where q is the density of the heat flux on the boundary ∂ Ω.
        On virtue of the Fick law, is as follows:

                                      q = −κ        u(t, x);



            u(t, x)        is the energy per unit mass (the temperature);
        ρ      is the heat capacity per unit volume;
             κ(u; t, x)       is the heat transport coefficient;
            f (u(t, x); t, x)     is the voluminal heat source.

Summer School: KPI, 2010            Nonlinear transport phenomena                           5 / 38
Heat transport: the balance equation


             d
                      ρ u(t, x)d x = −         q dσ +         f (u(t, x); t, x)d x,   (1)
             dt   Ω                       ∂Ω              Ω

        where q is the density of the heat flux on the boundary ∂ Ω.
        On virtue of the Fick law, is as follows:

                                      q = −κ        u(t, x);



            u(t, x)        is the energy per unit mass (the temperature);
        ρ      is the heat capacity per unit volume;
             κ(u; t, x)       is the heat transport coefficient;
            f (u(t, x); t, x)     is the voluminal heat source.

Summer School: KPI, 2010            Nonlinear transport phenomena                           5 / 38
Heat transport: the balance equation


             d
                      ρ u(t, x)d x = −         q dσ +         f (u(t, x); t, x)d x,   (1)
             dt   Ω                       ∂Ω              Ω

        where q is the density of the heat flux on the boundary ∂ Ω.
        On virtue of the Fick law, is as follows:

                                      q = −κ        u(t, x);



            u(t, x)        is the energy per unit mass (the temperature);
        ρ      is the heat capacity per unit volume;
             κ(u; t, x)       is the heat transport coefficient;
            f (u(t, x); t, x)     is the voluminal heat source.

Summer School: KPI, 2010            Nonlinear transport phenomena                           5 / 38
Using the Green-Gauss-Ostrogradsky theorem, we are able to
        write down the following identities:

              −        q dσ ≡        κ    u(t, x) dσ =              [κ   u(t, x)] d x.
                  ∂Ω            ∂Ω                           Ω
        So, the balance equation can be rewritten as
                      ∂ u(t, x)
                  ρ             −    [κ    u(t, x)] − f [u(t, x); t, x] d x = 0.
             Ω           ∂t
        Since the volume Ω is arbitrary, fulfillment of the balance
        equation is possible providing that the integrand is equal to zero
        In the case of one spatial variable, we get this way the equation

        ∂ u(t, x)    ∂                   ∂ u(t, x)   ˜
                  =    κ (u(t, x), t, x)
                        ˜                          + f (u(t, x), t, x) ,
           ∂t       ∂x                      ∂x
                                                                    (2)
                   ˜
        κ = κ/ρ, f = f /ρ.
        ˜


Summer School: KPI, 2010            Nonlinear transport phenomena                        6 / 38
Using the Green-Gauss-Ostrogradsky theorem, we are able to
        write down the following identities:

              −        q dσ ≡        κ    u(t, x) dσ =              [κ   u(t, x)] d x.
                  ∂Ω            ∂Ω                           Ω
        So, the balance equation can be rewritten as
                      ∂ u(t, x)
                  ρ             −    [κ    u(t, x)] − f [u(t, x); t, x] d x = 0.
             Ω           ∂t
        Since the volume Ω is arbitrary, fulfillment of the balance
        equation is possible providing that the integrand is equal to zero
        In the case of one spatial variable, we get this way the equation

        ∂ u(t, x)    ∂                   ∂ u(t, x)   ˜
                  =    κ (u(t, x), t, x)
                        ˜                          + f (u(t, x), t, x) ,
           ∂t       ∂x                      ∂x
                                                                    (2)
                   ˜
        κ = κ/ρ, f = f /ρ.
        ˜


Summer School: KPI, 2010            Nonlinear transport phenomena                        6 / 38
Using the Green-Gauss-Ostrogradsky theorem, we are able to
        write down the following identities:

              −        q dσ ≡        κ    u(t, x) dσ =              [κ   u(t, x)] d x.
                  ∂Ω            ∂Ω                           Ω
        So, the balance equation can be rewritten as
                      ∂ u(t, x)
                  ρ             −    [κ    u(t, x)] − f [u(t, x); t, x] d x = 0.
             Ω           ∂t
        Since the volume Ω is arbitrary, fulfillment of the balance
        equation is possible providing that the integrand is equal to zero
        In the case of one spatial variable, we get this way the equation

        ∂ u(t, x)    ∂                   ∂ u(t, x)   ˜
                  =    κ (u(t, x), t, x)
                        ˜                          + f (u(t, x), t, x) ,
           ∂t       ∂x                      ∂x
                                                                    (2)
                   ˜
        κ = κ/ρ, f = f /ρ.
        ˜


Summer School: KPI, 2010            Nonlinear transport phenomena                        6 / 38
Using the Green-Gauss-Ostrogradsky theorem, we are able to
        write down the following identities:

              −        q dσ ≡        κ    u(t, x) dσ =              [κ   u(t, x)] d x.
                  ∂Ω            ∂Ω                           Ω
        So, the balance equation can be rewritten as
                      ∂ u(t, x)
                  ρ             −    [κ    u(t, x)] − f [u(t, x); t, x] d x = 0.
             Ω           ∂t
        Since the volume Ω is arbitrary, fulfillment of the balance
        equation is possible providing that the integrand is equal to zero
        In the case of one spatial variable, we get this way the equation

        ∂ u(t, x)    ∂                   ∂ u(t, x)   ˜
                  =    κ (u(t, x), t, x)
                        ˜                          + f (u(t, x), t, x) ,
           ∂t       ∂x                      ∂x
                                                                    (2)
                   ˜
        κ = κ/ρ, f = f /ρ.
        ˜


Summer School: KPI, 2010            Nonlinear transport phenomena                        6 / 38
Self-similar solutions to the heat equation.
        Statement 1. Any physical law can be written down in the
        form
                a = ϕ (a1 , a2 , ..., an ; an+1 , ...an+m ) , where
                                       a1 , a2 , ..., an
        are the physical quantities expressed in the basic units,
                                    a, an+1 , ...an+m
        are the physical quantities expressed in derived units.
        Example of basic units:
                      the length[L], the time [T ], the mass [M ]
        Examples of physical quantities expressed in derived
        units:
                           dx L       d2 x M · L
                     V =        , F =m 2         , etc., etc.
                           dt T       dt    T2

Summer School: KPI, 2010         Nonlinear transport phenomena        7 / 38
Self-similar solutions to the heat equation.
        Statement 1. Any physical law can be written down in the
        form
                a = ϕ (a1 , a2 , ..., an ; an+1 , ...an+m ) , where
                                       a1 , a2 , ..., an
        are the physical quantities expressed in the basic units,
                                    a, an+1 , ...an+m
        are the physical quantities expressed in derived units.
        Example of basic units:
                      the length[L], the time [T ], the mass [M ]
        Examples of physical quantities expressed in derived
        units:
                           dx L       d2 x M · L
                     V =        , F =m 2         , etc., etc.
                           dt T       dt    T2

Summer School: KPI, 2010         Nonlinear transport phenomena        7 / 38
Self-similar solutions to the heat equation.
        Statement 1. Any physical law can be written down in the
        form
                a = ϕ (a1 , a2 , ..., an ; an+1 , ...an+m ) , where
                                       a1 , a2 , ..., an
        are the physical quantities expressed in the basic units,
                                    a, an+1 , ...an+m
        are the physical quantities expressed in derived units.
        Example of basic units:
                      the length[L], the time [T ], the mass [M ]
        Examples of physical quantities expressed in derived
        units:
                           dx L       d2 x M · L
                     V =        , F =m 2         , etc., etc.
                           dt T       dt    T2

Summer School: KPI, 2010         Nonlinear transport phenomena        7 / 38
Self-similar solutions to the heat equation.
        Statement 1. Any physical law can be written down in the
        form
                a = ϕ (a1 , a2 , ..., an ; an+1 , ...an+m ) , where
                                       a1 , a2 , ..., an
        are the physical quantities expressed in the basic units,
                                    a, an+1 , ...an+m
        are the physical quantities expressed in derived units.
        Example of basic units:
                      the length[L], the time [T ], the mass [M ]
        Examples of physical quantities expressed in derived
        units:
                           dx L       d2 x M · L
                     V =        , F =m 2         , etc., etc.
                           dt T       dt    T2

Summer School: KPI, 2010         Nonlinear transport phenomena        7 / 38
Self-similar solutions to the heat equation.
        Statement 1. Any physical law can be written down in the
        form
                a = ϕ (a1 , a2 , ..., an ; an+1 , ...an+m ) , where
                                       a1 , a2 , ..., an
        are the physical quantities expressed in the basic units,
                                    a, an+1 , ...an+m
        are the physical quantities expressed in derived units.
        Example of basic units:
                      the length[L], the time [T ], the mass [M ]
        Examples of physical quantities expressed in derived
        units:
                           dx L       d2 x M · L
                     V =        , F =m 2         , etc., etc.
                           dt T       dt    T2

Summer School: KPI, 2010         Nonlinear transport phenomena        7 / 38
Statement 2. Derived physical quantities a, an+1 , ...an+m can
        be expressed through the basic ones in the following form:

                                                       a = ar1 ar2 ...arn Π,
                                                            1 2        n
                                     rj    rj      j
                      an+j = a11 a22 ...arn Πj ,
                                         n                  j = 1, 2, ...., m,

        where Π, Π1 , ...., Πm are dimensionless parameters.
        So, the physical law

                                     a = F (a1 , ...an ; ...an+j , ....)

        is equivalent to

                             1                                     rj     rj  rj
             Π=                           F a1 , a2 , ..., an ;...a11 a22 ...ann Πj ... .   (3)
                    ar1
                     1     ar2
                            2    ...arn
                                     n




Summer School: KPI, 2010                  Nonlinear transport phenomena                           8 / 38
Statement 2. Derived physical quantities a, an+1 , ...an+m can
        be expressed through the basic ones in the following form:

                                                       a = ar1 ar2 ...arn Π,
                                                            1 2        n
                                     rj    rj      j
                      an+j = a11 a22 ...arn Πj ,
                                         n                  j = 1, 2, ...., m,

        where Π, Π1 , ...., Πm are dimensionless parameters.
        So, the physical law

                                     a = F (a1 , ...an ; ...an+j , ....)

        is equivalent to

                             1                                     rj     rj  rj
             Π=                           F a1 , a2 , ..., an ;...a11 a22 ...ann Πj ... .   (3)
                    ar1
                     1     ar2
                            2    ...arn
                                     n




Summer School: KPI, 2010                  Nonlinear transport phenomena                           8 / 38
The essence of Π theorem

         Theorem. The RHS of the formula (3) depends, at most,
        on the m dimensionless parameters Π1 , ...., Πm , and does
        not depend on the dimension quantities a1 , a2 , ..., an .
        1

        So, the physical law

                             a = F (a1 , ...an ; ...an+j , ....)

        in dimensionless variables takes the form to

                               Π = Ψ (Π1 , Π2 , ...Πm ) .                    (4)



          1
            A. Samarskij, A. Mikhailov, Mathematical Modelling: Ideas, Methods,
        Examples, Mosclw, 1997, Ch. V.
Summer School: KPI, 2010        Nonlinear transport phenomena                      9 / 38
The essence of Π theorem

         Theorem. The RHS of the formula (3) depends, at most,
        on the m dimensionless parameters Π1 , ...., Πm , and does
        not depend on the dimension quantities a1 , a2 , ..., an .
        1

        So, the physical law

                             a = F (a1 , ...an ; ...an+j , ....)

        in dimensionless variables takes the form to

                               Π = Ψ (Π1 , Π2 , ...Πm ) .                    (4)



          1
            A. Samarskij, A. Mikhailov, Mathematical Modelling: Ideas, Methods,
        Examples, Mosclw, 1997, Ch. V.
Summer School: KPI, 2010        Nonlinear transport phenomena                      9 / 38
The essence of Π theorem

         Theorem. The RHS of the formula (3) depends, at most,
        on the m dimensionless parameters Π1 , ...., Πm , and does
        not depend on the dimension quantities a1 , a2 , ..., an .
        1

        So, the physical law

                             a = F (a1 , ...an ; ...an+j , ....)

        in dimensionless variables takes the form to

                               Π = Ψ (Π1 , Π2 , ...Πm ) .                    (4)



          1
            A. Samarskij, A. Mikhailov, Mathematical Modelling: Ideas, Methods,
        Examples, Mosclw, 1997, Ch. V.
Summer School: KPI, 2010        Nonlinear transport phenomena                      9 / 38
Applications. Point explosion (linear case)

                           u t = κ ux x ,                                             (5)
                                                       ∞
        u(0, x) = 0,        x ∈ R,      x = 0,             u(t, x) d x = Q = const.
                                                      −∞

        The solution to (5) can be expressed as u = f (t, κ, Q; x). Basic
        units:
                                   E · L2
                         t[T ], κ          , Q[E · L].
                                     T
        Derived units:
                                Q            √
                           u = √ Π, x = κ t Π1 .
                                 κt
        Thus, passing to dimensionless variables we get
                         √
                           κt            √
                    Π=        f (t, κ, Q; κ t Π1 ) = ϕ (Π1 ) .
                          Q
Summer School: KPI, 2010          Nonlinear transport phenomena                       10 / 38
Applications. Point explosion (linear case)

                           u t = κ ux x ,                                             (5)
                                                       ∞
        u(0, x) = 0,        x ∈ R,      x = 0,             u(t, x) d x = Q = const.
                                                      −∞

        The solution to (5) can be expressed as u = f (t, κ, Q; x). Basic
        units:
                                   E · L2
                         t[T ], κ          , Q[E · L].
                                     T
        Derived units:
                                Q            √
                           u = √ Π, x = κ t Π1 .
                                 κt
        Thus, passing to dimensionless variables we get
                         √
                           κt            √
                    Π=        f (t, κ, Q; κ t Π1 ) = ϕ (Π1 ) .
                          Q
Summer School: KPI, 2010          Nonlinear transport phenomena                       10 / 38
Applications. Point explosion (linear case)

                           u t = κ ux x ,                                             (5)
                                                       ∞
        u(0, x) = 0,        x ∈ R,      x = 0,             u(t, x) d x = Q = const.
                                                      −∞

        The solution to (5) can be expressed as u = f (t, κ, Q; x). Basic
        units:
                                   E · L2
                         t[T ], κ          , Q[E · L].
                                     T
        Derived units:
                                Q            √
                           u = √ Π, x = κ t Π1 .
                                 κt
        Thus, passing to dimensionless variables we get
                         √
                           κt            √
                    Π=        f (t, κ, Q; κ t Π1 ) = ϕ (Π1 ) .
                          Q
Summer School: KPI, 2010          Nonlinear transport phenomena                       10 / 38
Applications. Point explosion (linear case)

                           u t = κ ux x ,                                             (5)
                                                       ∞
        u(0, x) = 0,        x ∈ R,      x = 0,             u(t, x) d x = Q = const.
                                                      −∞

        The solution to (5) can be expressed as u = f (t, κ, Q; x). Basic
        units:
                                   E · L2
                         t[T ], κ          , Q[E · L].
                                     T
        Derived units:
                                Q            √
                           u = √ Π, x = κ t Π1 .
                                 κt
        Thus, passing to dimensionless variables we get
                         √
                           κt            √
                    Π=        f (t, κ, Q; κ t Π1 ) = ϕ (Π1 ) .
                          Q
Summer School: KPI, 2010          Nonlinear transport phenomena                       10 / 38
Applications. Point explosion (linear case)

                           u t = κ ux x ,                                             (5)
                                                       ∞
        u(0, x) = 0,        x ∈ R,      x = 0,             u(t, x) d x = Q = const.
                                                      −∞

        The solution to (5) can be expressed as u = f (t, κ, Q; x). Basic
        units:
                                   E · L2
                         t[T ], κ          , Q[E · L].
                                     T
        Derived units:
                                Q            √
                           u = √ Π, x = κ t Π1 .
                                 κt
        Thus, passing to dimensionless variables we get
                         √
                           κt            √
                    Π=        f (t, κ, Q; κ t Π1 ) = ϕ (Π1 ) .
                          Q
Summer School: KPI, 2010          Nonlinear transport phenomena                       10 / 38
Corollary.Ansatz
                                   Q
                              u = √ ϕ(ξ),               ξ = Π1
                                   κt
        should strongly simplify the point explosion problem (5).
        In fact, inserting this ansatz into the heat transport equation,
        we get
                                               d
                    2 ϕ[ξ] + ξ ϕ[ξ] + ϕ[ξ] ≡
                      ¨        ˙                  (ξϕ[ξ] + 2 ϕ[ξ]) = 0,
                                                             ˙            (6)
                                               dξ

        with the additional condition (as well expressed in the
        dimensionless variables):
                                       ∞
                                           ϕ[ξ] dξ = 1.                   (7)
                                      −∞




Summer School: KPI, 2010         Nonlinear transport phenomena              11 / 38
Corollary.Ansatz
                                   Q
                              u = √ ϕ(ξ),               ξ = Π1
                                   κt
        should strongly simplify the point explosion problem (5).
        In fact, inserting this ansatz into the heat transport equation,
        we get
                                               d
                    2 ϕ[ξ] + ξ ϕ[ξ] + ϕ[ξ] ≡
                      ¨        ˙                  (ξϕ[ξ] + 2 ϕ[ξ]) = 0,
                                                             ˙            (6)
                                               dξ

        with the additional condition (as well expressed in the
        dimensionless variables):
                                       ∞
                                           ϕ[ξ] dξ = 1.                   (7)
                                      −∞




Summer School: KPI, 2010         Nonlinear transport phenomena              11 / 38
Corollary.Ansatz
                                   Q
                              u = √ ϕ(ξ),               ξ = Π1
                                   κt
        should strongly simplify the point explosion problem (5).
        In fact, inserting this ansatz into the heat transport equation,
        we get
                                               d
                    2 ϕ[ξ] + ξ ϕ[ξ] + ϕ[ξ] ≡
                      ¨        ˙                  (ξϕ[ξ] + 2 ϕ[ξ]) = 0,
                                                             ˙            (6)
                                               dξ

        with the additional condition (as well expressed in the
        dimensionless variables):
                                       ∞
                                           ϕ[ξ] dξ = 1.                   (7)
                                      −∞




Summer School: KPI, 2010         Nonlinear transport phenomena              11 / 38
Statement 3. Solution to the problem
                            d
                               (ξϕ[ξ] + 2 ϕ[ξ]) = 0,
                                          ˙
                            dξ
                                     ∞
                                         ϕ[ξ] dξ = 1
                                    −∞
        is given by the formula
                                          1     ξ2
                                  ϕ[ξ] = √    e− 4 .
                                           4π


        Corollary. The only solution to the linear point explosion
        problem (5) is given by the formula
                                          Q       x2
                             u= √             e− 4 κ t .       (8)
                                         4κπt


Summer School: KPI, 2010     Nonlinear transport phenomena           12 / 38
Statement 3. Solution to the problem
                            d
                               (ξϕ[ξ] + 2 ϕ[ξ]) = 0,
                                          ˙
                            dξ
                                     ∞
                                         ϕ[ξ] dξ = 1
                                    −∞
        is given by the formula
                                          1     ξ2
                                  ϕ[ξ] = √    e− 4 .
                                           4π


        Corollary. The only solution to the linear point explosion
        problem (5) is given by the formula
                                          Q       x2
                             u= √             e− 4 κ t .       (8)
                                         4κπt


Summer School: KPI, 2010     Nonlinear transport phenomena           12 / 38
Figure:




Summer School: KPI, 2010   Nonlinear transport phenomena   13 / 38
Figure:




Summer School: KPI, 2010   Nonlinear transport phenomena   14 / 38
Figure:




Summer School: KPI, 2010   Nonlinear transport phenomena   15 / 38
Remark 1. For Q = 1 the previous formula written as
                                   1        x2
                              E=√       e− 4 κ t
                                  4πκt
        defines so called fundamental solution of the operator

                                ˆ    ∂      ∂2
                               L=      − κ 2.
                                    ∂t     ∂x
        Knowledge of this solution enables to solve the initial value
        problem.
         Theorem Solution to the initial value problem

                           ut − κ u x x ,       u(0, x) = F (x)

        is given by the formula
                                                   +∞
                u(t, x) = E ∗ F (t, x) ≡                E (t, x − y) · F (y) d y
                                                 −∞

        provided that the integral in the RHS does exist.
Summer School: KPI, 2010            Nonlinear transport phenomena                  16 / 38
Remark 1. For Q = 1 the previous formula written as
                                   1        x2
                              E=√       e− 4 κ t
                                  4πκt
        defines so called fundamental solution of the operator

                                ˆ    ∂      ∂2
                               L=      − κ 2.
                                    ∂t     ∂x
        Knowledge of this solution enables to solve the initial value
        problem.
         Theorem Solution to the initial value problem

                           ut − κ u x x ,       u(0, x) = F (x)

        is given by the formula
                                                   +∞
                u(t, x) = E ∗ F (t, x) ≡                E (t, x − y) · F (y) d y
                                                 −∞

        provided that the integral in the RHS does exist.
Summer School: KPI, 2010            Nonlinear transport phenomena                  16 / 38
Peculiarity of the solution
                                        Q       x2
                              u= √          e− 4 κ t
                                       4κπt
        to the linear point explosion:

          At any moment of time all the space ”knows ” that the point
               explosion took place. This is physically incorrect!!!




Summer School: KPI, 2010     Nonlinear transport phenomena              17 / 38
Peculiarity of the solution
                                        Q       x2
                              u= √          e− 4 κ t
                                       4κπt
        to the linear point explosion:

          At any moment of time all the space ”knows ” that the point
               explosion took place. This is physically incorrect!!!




Summer School: KPI, 2010     Nonlinear transport phenomena              17 / 38
Point explosion (nonlinear case)

                           ut = κ [u ux ]x ,                                              (9)
                                                           ∞
        u(0, x) = 0,            x ∈ R,     x = 0,              u(t, x) d x = Q = const.
                                                          −∞

        The solution to (9) can be expressed as u = f (t, κ, Q; x). Basic
        units:
                                     L2
                         t[T ], κ         , Q[E · L].
                                    E·T
        Derived units:
                                     Q2/3
                               u=            Π,     x = [κ Q t]1/3 Π1 .
                                    [κ t]1/3
        Thus, passing to dimensionless variables we get
                            [κ t]1/3
                    Π=               f t, κ, Q; Π1 [κ Q t]1/3 = ϕ (Π1 ) .
                             Q2/3
Summer School: KPI, 2010              Nonlinear transport phenomena                       18 / 38
Point explosion (nonlinear case)

                           ut = κ [u ux ]x ,                                              (9)
                                                           ∞
        u(0, x) = 0,            x ∈ R,     x = 0,              u(t, x) d x = Q = const.
                                                          −∞

        The solution to (9) can be expressed as u = f (t, κ, Q; x). Basic
        units:
                                     L2
                         t[T ], κ         , Q[E · L].
                                    E·T
        Derived units:
                                     Q2/3
                               u=            Π,     x = [κ Q t]1/3 Π1 .
                                    [κ t]1/3
        Thus, passing to dimensionless variables we get
                            [κ t]1/3
                    Π=               f t, κ, Q; Π1 [κ Q t]1/3 = ϕ (Π1 ) .
                             Q2/3
Summer School: KPI, 2010              Nonlinear transport phenomena                       18 / 38
Point explosion (nonlinear case)

                           ut = κ [u ux ]x ,                                              (9)
                                                           ∞
        u(0, x) = 0,            x ∈ R,     x = 0,              u(t, x) d x = Q = const.
                                                          −∞

        The solution to (9) can be expressed as u = f (t, κ, Q; x). Basic
        units:
                                     L2
                         t[T ], κ         , Q[E · L].
                                    E·T
        Derived units:
                                     Q2/3
                               u=            Π,     x = [κ Q t]1/3 Π1 .
                                    [κ t]1/3
        Thus, passing to dimensionless variables we get
                            [κ t]1/3
                    Π=               f t, κ, Q; Π1 [κ Q t]1/3 = ϕ (Π1 ) .
                             Q2/3
Summer School: KPI, 2010              Nonlinear transport phenomena                       18 / 38
Point explosion (nonlinear case)

                           ut = κ [u ux ]x ,                                              (9)
                                                           ∞
        u(0, x) = 0,            x ∈ R,     x = 0,              u(t, x) d x = Q = const.
                                                          −∞

        The solution to (9) can be expressed as u = f (t, κ, Q; x). Basic
        units:
                                     L2
                         t[T ], κ         , Q[E · L].
                                    E·T
        Derived units:
                                     Q2/3
                               u=            Π,     x = [κ Q t]1/3 Π1 .
                                    [κ t]1/3
        Thus, passing to dimensionless variables we get
                            [κ t]1/3
                    Π=               f t, κ, Q; Π1 [κ Q t]1/3 = ϕ (Π1 ) .
                             Q2/3
Summer School: KPI, 2010              Nonlinear transport phenomena                       18 / 38
Point explosion (nonlinear case)

                           ut = κ [u ux ]x ,                                              (9)
                                                           ∞
        u(0, x) = 0,            x ∈ R,     x = 0,              u(t, x) d x = Q = const.
                                                          −∞

        The solution to (9) can be expressed as u = f (t, κ, Q; x). Basic
        units:
                                     L2
                         t[T ], κ         , Q[E · L].
                                    E·T
        Derived units:
                                     Q2/3
                               u=            Π,     x = [κ Q t]1/3 Π1 .
                                    [κ t]1/3
        Thus, passing to dimensionless variables we get
                            [κ t]1/3
                    Π=               f t, κ, Q; Π1 [κ Q t]1/3 = ϕ (Π1 ) .
                             Q2/3
Summer School: KPI, 2010              Nonlinear transport phenomena                       18 / 38
Corollary. Ansatz
                                Q                                     x
                       u=              ϕ(ξ),       ξ = Π1 =
                            [κ Q t]1/3                            [κ Q t]1/3

        should strongly simplify the point explosion problem (9).
        In fact, inserting this ansatz into the heat transport equation,
        we get
                             d
                                (ξϕ[ξ] + 3 ϕ[ξ] ϕ[ξ]) = 0,
                                                ˙                     (10)
                            dξ
        with the additional condition (as well expressed in the
        dimensionless variables):
                                        ∞
                                            ϕ[ξ] dξ = 1.                       (11)
                                       −∞




Summer School: KPI, 2010          Nonlinear transport phenomena                   19 / 38
Corollary. Ansatz
                                Q                                     x
                       u=              ϕ(ξ),       ξ = Π1 =
                            [κ Q t]1/3                            [κ Q t]1/3

        should strongly simplify the point explosion problem (9).
        In fact, inserting this ansatz into the heat transport equation,
        we get
                             d
                                (ξϕ[ξ] + 3 ϕ[ξ] ϕ[ξ]) = 0,
                                                ˙                     (10)
                            dξ
        with the additional condition (as well expressed in the
        dimensionless variables):
                                        ∞
                                            ϕ[ξ] dξ = 1.                       (11)
                                       −∞




Summer School: KPI, 2010          Nonlinear transport phenomena                   19 / 38
Corollary. Ansatz
                                Q                                     x
                       u=              ϕ(ξ),       ξ = Π1 =
                            [κ Q t]1/3                            [κ Q t]1/3

        should strongly simplify the point explosion problem (9).
        In fact, inserting this ansatz into the heat transport equation,
        we get
                             d
                                (ξϕ[ξ] + 3 ϕ[ξ] ϕ[ξ]) = 0,
                                                ˙                     (10)
                            dξ
        with the additional condition (as well expressed in the
        dimensionless variables):
                                        ∞
                                            ϕ[ξ] dξ = 1.                       (11)
                                       −∞




Summer School: KPI, 2010          Nonlinear transport phenomena                   19 / 38
Statement 4. Solution to the problem
                        d
                           (ξϕ[ξ] + 3 ϕ[ξ] ϕ[ξ]) = 0,
                                           ˙
                        dξ
                                            ∞
                                                ϕ[ξ] dξ = 1
                                         −∞

        is given by the formula
                                        1
                                        6   ξΦ − ξ 2 if |ξ| < ξΦ ,
                                             2
                           ϕ[ξ] =
                                               0 otherwise,
        where ξΦ = (9/2)1/3 .
        Corollary. The only solution to the nonlinear point
        explosion problem (9) is given by the formula
                                                                      1/6
                         
                          (9/2)1/3 − x2                       t 2
                 Q       
                                      [κ Q t]2/3
                                                 if |x| < 9 (κ 2 Q)         ,
        u=
            6 [κ Q t]1/3 
                                  0                   otherwise
                         
                                                                      (12)
Summer School: KPI, 2010            Nonlinear transport phenomena               20 / 38
Statement 4. Solution to the problem
                        d
                           (ξϕ[ξ] + 3 ϕ[ξ] ϕ[ξ]) = 0,
                                           ˙
                        dξ
                                            ∞
                                                ϕ[ξ] dξ = 1
                                         −∞

        is given by the formula
                                        1
                                        6   ξΦ − ξ 2 if |ξ| < ξΦ ,
                                             2
                           ϕ[ξ] =
                                               0 otherwise,
        where ξΦ = (9/2)1/3 .
        Corollary. The only solution to the nonlinear point
        explosion problem (9) is given by the formula
                                                                      1/6
                         
                          (9/2)1/3 − x2                       t 2
                 Q       
                                      [κ Q t]2/3
                                                 if |x| < 9 (κ 2 Q)         ,
        u=
            6 [κ Q t]1/3 
                                  0                   otherwise
                         
                                                                      (12)
Summer School: KPI, 2010            Nonlinear transport phenomena               20 / 38
Statement 4. Solution to the problem
                        d
                           (ξϕ[ξ] + 3 ϕ[ξ] ϕ[ξ]) = 0,
                                           ˙
                        dξ
                                            ∞
                                                ϕ[ξ] dξ = 1
                                         −∞

        is given by the formula
                                        1
                                        6   ξΦ − ξ 2 if |ξ| < ξΦ ,
                                             2
                           ϕ[ξ] =
                                               0 otherwise,
        where ξΦ = (9/2)1/3 .
        Corollary. The only solution to the nonlinear point
        explosion problem (9) is given by the formula
                                                                      1/6
                         
                          (9/2)1/3 − x2                       t 2
                 Q       
                                      [κ Q t]2/3
                                                 if |x| < 9 (κ 2 Q)         ,
        u=
            6 [κ Q t]1/3 
                                  0                   otherwise
                         
                                                                      (12)
Summer School: KPI, 2010            Nonlinear transport phenomena               20 / 38
Figure:




Summer School: KPI, 2010   Nonlinear transport phenomena   21 / 38
Figure:




Summer School: KPI, 2010   Nonlinear transport phenomena   22 / 38
Figure:




Summer School: KPI, 2010   Nonlinear transport phenomena   23 / 38
Figure:




Summer School: KPI, 2010   Nonlinear transport phenomena   24 / 38
Conclusion. Nonlinearity of the heat transport coefficient
        causes the localization of the heat wave within the finite
        domain
                              < −L(t), L(t) >,
        where
                                                    1/6
                                  9 (κ t Q)2
                           L(t) =                         = C t1/3 .
                                      2




Summer School: KPI, 2010        Nonlinear transport phenomena          25 / 38
Unlimited growth of temperature and the complete
  localization of the heat energy


                           ut = κ [uσ ux ]x ,                                                  (13)
                              −n
        u(t, 0) = A (−t)           , t < 0, n > 0,          x ∈ R+ ,            lim u(t, x) = 0(14)
                                                                           t→−∞

        In general, the problem is self-similar and the ansatz
                                  A                                  x
                           u=         ϕ(ξ),          ξ=√
                                (−t)n                         Aσ     κ t1−n σ
        reduces the initial problem to corresponding ODE.
        If we assume that n = σ = 1, than ξ will not depend on t:
                                      A                          x
                                u=        ϕ(ξ),          ξ=√        .
                                     (−t)                        Aκ

Summer School: KPI, 2010             Nonlinear transport phenomena                           26 / 38
Unlimited growth of temperature and the complete
  localization of the heat energy


                           ut = κ [uσ ux ]x ,                                                  (13)
                              −n
        u(t, 0) = A (−t)           , t < 0, n > 0,          x ∈ R+ ,            lim u(t, x) = 0(14)
                                                                           t→−∞

        In general, the problem is self-similar and the ansatz
                                  A                                  x
                           u=         ϕ(ξ),          ξ=√
                                (−t)n                         Aσ     κ t1−n σ
        reduces the initial problem to corresponding ODE.
        If we assume that n = σ = 1, than ξ will not depend on t:
                                      A                          x
                                u=        ϕ(ξ),          ξ=√        .
                                     (−t)                        Aκ

Summer School: KPI, 2010             Nonlinear transport phenomena                           26 / 38
Unlimited growth of temperature and the complete
  localization of the heat energy


                           ut = κ [uσ ux ]x ,                                                  (13)
                              −n
        u(t, 0) = A (−t)           , t < 0, n > 0,          x ∈ R+ ,            lim u(t, x) = 0(14)
                                                                           t→−∞

        In general, the problem is self-similar and the ansatz
                                  A                                  x
                           u=         ϕ(ξ),          ξ=√
                                (−t)n                         Aσ     κ t1−n σ
        reduces the initial problem to corresponding ODE.
        If we assume that n = σ = 1, than ξ will not depend on t:
                                      A                          x
                                u=        ϕ(ξ),          ξ=√        .
                                     (−t)                        Aκ

Summer School: KPI, 2010             Nonlinear transport phenomena                           26 / 38
Unlimited growth of temperature and the complete
  localization of the heat energy


                           ut = κ [uσ ux ]x ,                                                  (13)
                              −n
        u(t, 0) = A (−t)           , t < 0, n > 0,          x ∈ R+ ,            lim u(t, x) = 0(14)
                                                                           t→−∞

        In general, the problem is self-similar and the ansatz
                                  A                                  x
                           u=         ϕ(ξ),          ξ=√
                                (−t)n                         Aσ     κ t1−n σ
        reduces the initial problem to corresponding ODE.
        If we assume that n = σ = 1, than ξ will not depend on t:
                                      A                          x
                                u=        ϕ(ξ),          ξ=√        .
                                     (−t)                        Aκ

Summer School: KPI, 2010             Nonlinear transport phenomena                           26 / 38
The reduced equation
                                   ϕϕ + ϕ2 − ϕ = 0,
                                    ¨ ˙
        together with the initial condition ϕ(0) = 1 (arising from the
                             A
        condition u(t, 0) = (−t) ), has the solution
                                      √                   √
                             (1 − ξ/ 6)2 , if 0 < ξ < √     6,
                  ϕ(ξ) =
                           0                         if ξ ≥ 6,
        The corresponding solution to the BV problem (13) is as
        follows:

                            A
                                                2                  √
                                 1−   √ x           ,   if 0 < x < 6 κ A,
                u(t, x) =   −t         6κA                        √
                            0                               if x ≥ 6 κ A.

        Moral of the tale: [1]. Heat energy is completely localized in
                           √
        the segment 0 < x < 6 κ A0 ;
        [2]. The√temperature u(t, x) tends to +∞ in every point
        x ∈ [0, 6 κ A0 ) as t → 0−.
Summer School: KPI, 2010         Nonlinear transport phenomena              27 / 38
The reduced equation
                                   ϕϕ + ϕ2 − ϕ = 0,
                                    ¨ ˙
        together with the initial condition ϕ(0) = 1 (arising from the
                             A
        condition u(t, 0) = (−t) ), has the solution
                                      √                   √
                             (1 − ξ/ 6)2 , if 0 < ξ < √     6,
                  ϕ(ξ) =
                           0                         if ξ ≥ 6,
        The corresponding solution to the BV problem (13) is as
        follows:

                            A
                                                2                  √
                                 1−   √ x           ,   if 0 < x < 6 κ A,
                u(t, x) =   −t         6κA                        √
                            0                               if x ≥ 6 κ A.

        Moral of the tale: [1]. Heat energy is completely localized in
                           √
        the segment 0 < x < 6 κ A0 ;
        [2]. The√temperature u(t, x) tends to +∞ in every point
        x ∈ [0, 6 κ A0 ) as t → 0−.
Summer School: KPI, 2010         Nonlinear transport phenomena              27 / 38
The reduced equation
                                   ϕϕ + ϕ2 − ϕ = 0,
                                    ¨ ˙
        together with the initial condition ϕ(0) = 1 (arising from the
                             A
        condition u(t, 0) = (−t) ), has the solution
                                      √                   √
                             (1 − ξ/ 6)2 , if 0 < ξ < √     6,
                  ϕ(ξ) =
                           0                         if ξ ≥ 6,
        The corresponding solution to the BV problem (13) is as
        follows:

                            A
                                                2                  √
                                 1−   √ x           ,   if 0 < x < 6 κ A,
                u(t, x) =   −t         6κA                        √
                            0                               if x ≥ 6 κ A.

        Moral of the tale: [1]. Heat energy is completely localized in
                           √
        the segment 0 < x < 6 κ A0 ;
        [2]. The√temperature u(t, x) tends to +∞ in every point
        x ∈ [0, 6 κ A0 ) as t → 0−.
Summer School: KPI, 2010         Nonlinear transport phenomena              27 / 38
The reduced equation
                                   ϕϕ + ϕ2 − ϕ = 0,
                                    ¨ ˙
        together with the initial condition ϕ(0) = 1 (arising from the
                             A
        condition u(t, 0) = (−t) ), has the solution
                                      √                   √
                             (1 − ξ/ 6)2 , if 0 < ξ < √     6,
                  ϕ(ξ) =
                           0                         if ξ ≥ 6,
        The corresponding solution to the BV problem (13) is as
        follows:

                            A
                                                2                  √
                                 1−   √ x           ,   if 0 < x < 6 κ A,
                u(t, x) =   −t         6κA                        √
                            0                               if x ≥ 6 κ A.

        Moral of the tale: [1]. Heat energy is completely localized in
                           √
        the segment 0 < x < 6 κ A0 ;
        [2]. The√temperature u(t, x) tends to +∞ in every point
        x ∈ [0, 6 κ A0 ) as t → 0−.
Summer School: KPI, 2010         Nonlinear transport phenomena              27 / 38
Figure:




Summer School: KPI, 2010   Nonlinear transport phenomena   28 / 38
Figure:




Summer School: KPI, 2010   Nonlinear transport phenomena   29 / 38
Figure:




Summer School: KPI, 2010   Nonlinear transport phenomena   30 / 38
Blow-up regime caused by the source term incorporation
        Model equation:
                             ut = κ [u ux ]x + q0 u2 .                      (15)
        Ansatz u(t, x) = P (t) Q(x) leads, after some algebraic
        manipulation, to the equations
         P −2 (t)P (t) = κ(Q(x) Q(x) ) + q0 Q(x)2 Q−1 (x) = C1 . (16)
                  ˙


         Theorem. The equation (15) has the following solution:

                                             q0                              2
                        4         cos2            x ,        if |x| < 2 π0 κ ,
        u(t, x) =                            8κ                          q
                  3 q0 (tf − t)      0                          otherwise.
                                                                         (17)
        The above solution
                                                                      2 π2 κ
              is localized within the interval [−L, L], L =             q0
                                                                             ;
              tends to +∞ ∀ x ∈ (−L, L), as t → tf − 0.
Summer School: KPI, 2010     Nonlinear transport phenomena                       31 / 38
Blow-up regime caused by the source term incorporation
        Model equation:
                             ut = κ [u ux ]x + q0 u2 .                      (15)
        Ansatz u(t, x) = P (t) Q(x) leads, after some algebraic
        manipulation, to the equations
         P −2 (t)P (t) = κ(Q(x) Q(x) ) + q0 Q(x)2 Q−1 (x) = C1 . (16)
                  ˙


         Theorem. The equation (15) has the following solution:

                                             q0                              2
                        4         cos2            x ,        if |x| < 2 π0 κ ,
        u(t, x) =                            8κ                          q
                  3 q0 (tf − t)      0                          otherwise.
                                                                         (17)
        The above solution
                                                                      2 π2 κ
              is localized within the interval [−L, L], L =             q0
                                                                             ;
              tends to +∞ ∀ x ∈ (−L, L), as t → tf − 0.
Summer School: KPI, 2010     Nonlinear transport phenomena                       31 / 38
Blow-up regime caused by the source term incorporation
        Model equation:
                             ut = κ [u ux ]x + q0 u2 .                      (15)
        Ansatz u(t, x) = P (t) Q(x) leads, after some algebraic
        manipulation, to the equations
         P −2 (t)P (t) = κ(Q(x) Q(x) ) + q0 Q(x)2 Q−1 (x) = C1 . (16)
                  ˙


         Theorem. The equation (15) has the following solution:

                                             q0                              2
                        4         cos2            x ,        if |x| < 2 π0 κ ,
        u(t, x) =                            8κ                          q
                  3 q0 (tf − t)      0                          otherwise.
                                                                         (17)
        The above solution
                                                                      2 π2 κ
              is localized within the interval [−L, L], L =             q0
                                                                             ;
              tends to +∞ ∀ x ∈ (−L, L), as t → tf − 0.
Summer School: KPI, 2010     Nonlinear transport phenomena                       31 / 38
Blow-up regime caused by the source term incorporation
        Model equation:
                             ut = κ [u ux ]x + q0 u2 .                      (15)
        Ansatz u(t, x) = P (t) Q(x) leads, after some algebraic
        manipulation, to the equations
         P −2 (t)P (t) = κ(Q(x) Q(x) ) + q0 Q(x)2 Q−1 (x) = C1 . (16)
                  ˙


         Theorem. The equation (15) has the following solution:

                                             q0                              2
                        4         cos2            x ,        if |x| < 2 π0 κ ,
        u(t, x) =                            8κ                          q
                  3 q0 (tf − t)      0                          otherwise.
                                                                         (17)
        The above solution
                                                                      2 π2 κ
              is localized within the interval [−L, L], L =             q0
                                                                             ;
              tends to +∞ ∀ x ∈ (−L, L), as t → tf − 0.
Summer School: KPI, 2010     Nonlinear transport phenomena                       31 / 38
Blow-up regime caused by the source term incorporation
        Model equation:
                             ut = κ [u ux ]x + q0 u2 .                      (15)
        Ansatz u(t, x) = P (t) Q(x) leads, after some algebraic
        manipulation, to the equations
         P −2 (t)P (t) = κ(Q(x) Q(x) ) + q0 Q(x)2 Q−1 (x) = C1 . (16)
                  ˙


         Theorem. The equation (15) has the following solution:

                                             q0                              2
                        4         cos2            x ,        if |x| < 2 π0 κ ,
        u(t, x) =                            8κ                          q
                  3 q0 (tf − t)      0                          otherwise.
                                                                         (17)
        The above solution
                                                                      2 π2 κ
              is localized within the interval [−L, L], L =             q0
                                                                             ;
              tends to +∞ ∀ x ∈ (−L, L), as t → tf − 0.
Summer School: KPI, 2010     Nonlinear transport phenomena                       31 / 38
Blow-up regime caused by the source term incorporation
        Model equation:
                             ut = κ [u ux ]x + q0 u2 .                      (15)
        Ansatz u(t, x) = P (t) Q(x) leads, after some algebraic
        manipulation, to the equations
         P −2 (t)P (t) = κ(Q(x) Q(x) ) + q0 Q(x)2 Q−1 (x) = C1 . (16)
                  ˙


         Theorem. The equation (15) has the following solution:

                                             q0                              2
                        4         cos2            x ,        if |x| < 2 π0 κ ,
        u(t, x) =                            8κ                          q
                  3 q0 (tf − t)      0                          otherwise.
                                                                         (17)
        The above solution
                                                                      2 π2 κ
              is localized within the interval [−L, L], L =             q0
                                                                             ;
              tends to +∞ ∀ x ∈ (−L, L), as t → tf − 0.
Summer School: KPI, 2010     Nonlinear transport phenomena                       31 / 38
Figure:




Summer School: KPI, 2010   Nonlinear transport phenomena   32 / 38
Figure:




Summer School: KPI, 2010   Nonlinear transport phenomena   33 / 38
Figure:




Summer School: KPI, 2010   Nonlinear transport phenomena   34 / 38
Figure:




Summer School: KPI, 2010   Nonlinear transport phenomena   35 / 38
THE ROLE OF SELF-SIMILAR SOLUTIONS




Summer School: KPI, 2010   Nonlinear transport phenomena   36 / 38
Maximum principle. Comparison theorems ( initial
  value problem).
        Let us consider the Cauchy problem


                ut = [κ(u) ux ]x ,                                               (18)
                       u(0, x) = u0 (x) ≥ 0,            x ∈ R,       κ(u) > 0.


        Theorem 1. Maximum of the solution u(t, x) at any time
        t > 0 does not exceed the maximum of the initial data:

                    maxt>0, −∞<x<∞ u(t, x) ≤ max−∞<x<∞ u0 (x).

        2

            2
            A. Samarskij, A. Mikhailov, Mathematical Modelling: Ideas, Methods,
        Examples, Moscow, 1997, A. Samarskij, V. Galaktionov et al, Blow-up
        regimes in quasilinear parabolic equations, Berlin: Walter de Gruyter, 1995.
Summer School: KPI, 2010             Nonlinear transport phenomena                     37 / 38
Maximum principle. Comparison theorems ( initial
  value problem).
        Let us consider the Cauchy problem


                ut = [κ(u) ux ]x ,                                               (18)
                       u(0, x) = u0 (x) ≥ 0,            x ∈ R,       κ(u) > 0.


        Theorem 1. Maximum of the solution u(t, x) at any time
        t > 0 does not exceed the maximum of the initial data:

                    maxt>0, −∞<x<∞ u(t, x) ≤ max−∞<x<∞ u0 (x).

        2

            2
            A. Samarskij, A. Mikhailov, Mathematical Modelling: Ideas, Methods,
        Examples, Moscow, 1997, A. Samarskij, V. Galaktionov et al, Blow-up
        regimes in quasilinear parabolic equations, Berlin: Walter de Gruyter, 1995.
Summer School: KPI, 2010             Nonlinear transport phenomena                     37 / 38
Maximum principle. Comparison theorems ( initial
  value problem).
        Let us consider the Cauchy problem


                ut = [κ(u) ux ]x ,                                               (18)
                       u(0, x) = u0 (x) ≥ 0,            x ∈ R,       κ(u) > 0.


        Theorem 1. Maximum of the solution u(t, x) at any time
        t > 0 does not exceed the maximum of the initial data:

                    maxt>0, −∞<x<∞ u(t, x) ≤ max−∞<x<∞ u0 (x).

        2

            2
            A. Samarskij, A. Mikhailov, Mathematical Modelling: Ideas, Methods,
        Examples, Moscow, 1997, A. Samarskij, V. Galaktionov et al, Blow-up
        regimes in quasilinear parabolic equations, Berlin: Walter de Gruyter, 1995.
Summer School: KPI, 2010             Nonlinear transport phenomena                     37 / 38
Maximum principle. Comparison theorems ( initial
  value problem).
        Let us consider the Cauchy problem


                ut = [κ(u) ux ]x ,                                               (18)
                       u(0, x) = u0 (x) ≥ 0,            x ∈ R,       κ(u) > 0.


        Theorem 1. Maximum of the solution u(t, x) at any time
        t > 0 does not exceed the maximum of the initial data:

                    maxt>0, −∞<x<∞ u(t, x) ≤ max−∞<x<∞ u0 (x).

        2

            2
            A. Samarskij, A. Mikhailov, Mathematical Modelling: Ideas, Methods,
        Examples, Moscow, 1997, A. Samarskij, V. Galaktionov et al, Blow-up
        regimes in quasilinear parabolic equations, Berlin: Walter de Gruyter, 1995.
Summer School: KPI, 2010             Nonlinear transport phenomena                     37 / 38
Corollary (comparison theorem). Let u1 (t, x),
        u2 (t, x), u3 (t, x), are the solutions to the equation


          ut = [κ(u) ux ]x ,
                 u(0, x) = u0 (x) ≥ 0,              x ∈ R,      κ(u) > 0,

        corresponding to the initial data u1 (x), u2 (x), and u3 (x)
                                           0       0           0
        such that
                           u1 (x) ≤ u2 (x) ≤ u3 (x)
                            0         0        0

        for −∞ < x < ∞. Then

                           u1 (t, x) ≤ u2 (t, x) ≤ u3 (t, x)

        for all −∞ < x < ∞, and t > 0.


Summer School: KPI, 2010        Nonlinear transport phenomena               38 / 38
Corollary (comparison theorem). Let u1 (t, x),
        u2 (t, x), u3 (t, x), are the solutions to the equation


          ut = [κ(u) ux ]x ,
                 u(0, x) = u0 (x) ≥ 0,              x ∈ R,      κ(u) > 0,

        corresponding to the initial data u1 (x), u2 (x), and u3 (x)
                                           0       0           0
        such that
                           u1 (x) ≤ u2 (x) ≤ u3 (x)
                            0         0        0

        for −∞ < x < ∞. Then

                           u1 (t, x) ≤ u2 (t, x) ≤ u3 (t, x)

        for all −∞ < x < ∞, and t > 0.


Summer School: KPI, 2010        Nonlinear transport phenomena               38 / 38
Corollary (comparison theorem). Let u1 (t, x),
        u2 (t, x), u3 (t, x), are the solutions to the equation


          ut = [κ(u) ux ]x ,
                 u(0, x) = u0 (x) ≥ 0,              x ∈ R,      κ(u) > 0,

        corresponding to the initial data u1 (x), u2 (x), and u3 (x)
                                           0       0           0
        such that
                           u1 (x) ≤ u2 (x) ≤ u3 (x)
                            0         0        0

        for −∞ < x < ∞. Then

                           u1 (t, x) ≤ u2 (t, x) ≤ u3 (t, x)

        for all −∞ < x < ∞, and t > 0.


Summer School: KPI, 2010        Nonlinear transport phenomena               38 / 38
Corollary (comparison theorem). Let u1 (t, x),
        u2 (t, x), u3 (t, x), are the solutions to the equation


          ut = [κ(u) ux ]x ,
                 u(0, x) = u0 (x) ≥ 0,              x ∈ R,      κ(u) > 0,

        corresponding to the initial data u1 (x), u2 (x), and u3 (x)
                                           0       0           0
        such that
                           u1 (x) ≤ u2 (x) ≤ u3 (x)
                            0         0        0

        for −∞ < x < ∞. Then

                           u1 (t, x) ≤ u2 (t, x) ≤ u3 (t, x)

        for all −∞ < x < ∞, and t > 0.


Summer School: KPI, 2010        Nonlinear transport phenomena               38 / 38

More Related Content

PDF
Monte Carlo Statistical Methods
PDF
Monte Carlo Statistical Methods
PDF
Monte Carlo Statistical Methods
PDF
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
PDF
05 history of cv a machine learning (theory) perspective on computer vision
PDF
Lesage
PDF
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
PDF
Stability of adaptive random-walk Metropolis algorithms
Monte Carlo Statistical Methods
Monte Carlo Statistical Methods
Monte Carlo Statistical Methods
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
05 history of cv a machine learning (theory) perspective on computer vision
Lesage
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
Stability of adaptive random-walk Metropolis algorithms

What's hot (20)

PDF
Lesson 14: Derivatives of Logarithmic and Exponential Functions
PDF
NTU_paper
PDF
Introduction to MCMC methods
PDF
1 hofstad
PDF
Lesson 15: Exponential Growth and Decay
PDF
Recent developments on unbiased MCMC
PDF
04 structured support vector machine
PDF
Iskmaa
PDF
On estimating the integrated co volatility using
PDF
Nonlinear Manifolds in Computer Vision
PDF
Lesson 16: Inverse Trigonometric Functions
PDF
State estimate
PDF
Couplings of Markov chains and the Poisson equation
PDF
Unbiased Markov chain Monte Carlo methods
PPTX
PDF
Unbiased MCMC with couplings
PDF
Lesson 13: Related Rates Problems
PDF
Chapter 3 projection
PDF
Presentation cm2011
PDF
Rouviere
Lesson 14: Derivatives of Logarithmic and Exponential Functions
NTU_paper
Introduction to MCMC methods
1 hofstad
Lesson 15: Exponential Growth and Decay
Recent developments on unbiased MCMC
04 structured support vector machine
Iskmaa
On estimating the integrated co volatility using
Nonlinear Manifolds in Computer Vision
Lesson 16: Inverse Trigonometric Functions
State estimate
Couplings of Markov chains and the Poisson equation
Unbiased Markov chain Monte Carlo methods
Unbiased MCMC with couplings
Lesson 13: Related Rates Problems
Chapter 3 projection
Presentation cm2011
Rouviere
Ad

Viewers also liked (12)

PDF
Nonlinear transport phenomena: models, method of solving and unusual features...
PDF
Nonlinear transport phenomena: models, method of solving and unusual features...
PPTX
COMSOL Training Series (NNMDC Initiative)
PPTX
Comsol Multiphysics Presentation
PPTX
Pressure control valves
PPSX
Vapor liquid equilibrium using hysys
 
PDF
Hysys simulation
PDF
Simulation-Led Design Using SolidWorks® and COMSOL Multiphysics®
PDF
Transport phenomena-Mass Transfer 31-Jul-2016
PPTX
Pressure Relief Valve Sizing for Single Phase Flow
PPT
Reciprocating compressor
PDF
Basics of Compressor
Nonlinear transport phenomena: models, method of solving and unusual features...
Nonlinear transport phenomena: models, method of solving and unusual features...
COMSOL Training Series (NNMDC Initiative)
Comsol Multiphysics Presentation
Pressure control valves
Vapor liquid equilibrium using hysys
 
Hysys simulation
Simulation-Led Design Using SolidWorks® and COMSOL Multiphysics®
Transport phenomena-Mass Transfer 31-Jul-2016
Pressure Relief Valve Sizing for Single Phase Flow
Reciprocating compressor
Basics of Compressor
Ad

Similar to Nonlinear transport phenomena: models, method of solving and unusual features (1) (20)

PDF
Switkes01200543268
PDF
Causal Dynamical Triangulations
PDF
PPTX
RAMM.pptx
PDF
Qualitative measurement of Klauder coherent states using Bohmian machanics, C...
PDF
Lecture cochran
PDF
Fluctuations and rare events in stochastic aggregation
PDF
P diffusion_1
PDF
2_Kundur_DTSystemProperties_handouts.pdf
PDF
Pres110811
PDF
Math 497#w14
PDF
PDF
Bohmian trajectories from coherent states, Istanbul 3 July 2013
PDF
Mesoscopic multiclass traffic flow modeling on multi-lane sections
PDF
A current perspectives of corrected operator splitting (os) for systems
PDF
Numerical Solution of Stochastic Differential Equations in Finance
PDF
Numerical solution of spatiotemporal models from ecology
PDF
Jump-growth model for predator-prey dynamics
PDF
What happens when the Kolmogorov-Zakharov spectrum is nonlocal?
PDF
Nature-Inspired Optimization Algorithms
Switkes01200543268
Causal Dynamical Triangulations
RAMM.pptx
Qualitative measurement of Klauder coherent states using Bohmian machanics, C...
Lecture cochran
Fluctuations and rare events in stochastic aggregation
P diffusion_1
2_Kundur_DTSystemProperties_handouts.pdf
Pres110811
Math 497#w14
Bohmian trajectories from coherent states, Istanbul 3 July 2013
Mesoscopic multiclass traffic flow modeling on multi-lane sections
A current perspectives of corrected operator splitting (os) for systems
Numerical Solution of Stochastic Differential Equations in Finance
Numerical solution of spatiotemporal models from ecology
Jump-growth model for predator-prey dynamics
What happens when the Kolmogorov-Zakharov spectrum is nonlocal?
Nature-Inspired Optimization Algorithms

More from SSA KPI (20)

PDF
Germany presentation
PDF
Grand challenges in energy
PDF
Engineering role in sustainability
PDF
Consensus and interaction on a long term strategy for sustainable development
PDF
Competences in sustainability in engineering education
PDF
Introducatio SD for enginers
PPT
DAAD-10.11.2011
PDF
Talking with money
PDF
'Green' startup investment
PDF
From Huygens odd sympathy to the energy Huygens' extraction from the sea waves
PDF
Dynamics of dice games
PPT
Energy Security Costs
PPT
Naturally Occurring Radioactivity (NOR) in natural and anthropic environments
PDF
Advanced energy technology for sustainable development. Part 5
PDF
Advanced energy technology for sustainable development. Part 4
PDF
Advanced energy technology for sustainable development. Part 3
PDF
Advanced energy technology for sustainable development. Part 2
PDF
Advanced energy technology for sustainable development. Part 1
PPT
Fluorescent proteins in current biology
PPTX
Neurotransmitter systems of the brain and their functions
Germany presentation
Grand challenges in energy
Engineering role in sustainability
Consensus and interaction on a long term strategy for sustainable development
Competences in sustainability in engineering education
Introducatio SD for enginers
DAAD-10.11.2011
Talking with money
'Green' startup investment
From Huygens odd sympathy to the energy Huygens' extraction from the sea waves
Dynamics of dice games
Energy Security Costs
Naturally Occurring Radioactivity (NOR) in natural and anthropic environments
Advanced energy technology for sustainable development. Part 5
Advanced energy technology for sustainable development. Part 4
Advanced energy technology for sustainable development. Part 3
Advanced energy technology for sustainable development. Part 2
Advanced energy technology for sustainable development. Part 1
Fluorescent proteins in current biology
Neurotransmitter systems of the brain and their functions

Recently uploaded (20)

PDF
MBA _Common_ 2nd year Syllabus _2021-22_.pdf
PDF
BP 704 T. NOVEL DRUG DELIVERY SYSTEMS (UNIT 2).pdf
PDF
LIFE & LIVING TRILOGY - PART - (2) THE PURPOSE OF LIFE.pdf
PPTX
Core Concepts of Personalized Learning and Virtual Learning Environments
PPTX
What’s under the hood: Parsing standardized learning content for AI
PDF
LIFE & LIVING TRILOGY - PART (3) REALITY & MYSTERY.pdf
PDF
My India Quiz Book_20210205121199924.pdf
PDF
Race Reva University – Shaping Future Leaders in Artificial Intelligence
PDF
BP 505 T. PHARMACEUTICAL JURISPRUDENCE (UNIT 1).pdf
PPTX
B.Sc. DS Unit 2 Software Engineering.pptx
PDF
HVAC Specification 2024 according to central public works department
PDF
Paper A Mock Exam 9_ Attempt review.pdf.
PDF
medical_surgical_nursing_10th_edition_ignatavicius_TEST_BANK_pdf.pdf
PDF
FOISHS ANNUAL IMPLEMENTATION PLAN 2025.pdf
PDF
Vision Prelims GS PYQ Analysis 2011-2022 www.upscpdf.com.pdf
PDF
ChatGPT for Dummies - Pam Baker Ccesa007.pdf
PDF
LIFE & LIVING TRILOGY- PART (1) WHO ARE WE.pdf
PPTX
Share_Module_2_Power_conflict_and_negotiation.pptx
PPTX
ELIAS-SEZIURE AND EPilepsy semmioan session.pptx
PDF
Empowerment Technology for Senior High School Guide
MBA _Common_ 2nd year Syllabus _2021-22_.pdf
BP 704 T. NOVEL DRUG DELIVERY SYSTEMS (UNIT 2).pdf
LIFE & LIVING TRILOGY - PART - (2) THE PURPOSE OF LIFE.pdf
Core Concepts of Personalized Learning and Virtual Learning Environments
What’s under the hood: Parsing standardized learning content for AI
LIFE & LIVING TRILOGY - PART (3) REALITY & MYSTERY.pdf
My India Quiz Book_20210205121199924.pdf
Race Reva University – Shaping Future Leaders in Artificial Intelligence
BP 505 T. PHARMACEUTICAL JURISPRUDENCE (UNIT 1).pdf
B.Sc. DS Unit 2 Software Engineering.pptx
HVAC Specification 2024 according to central public works department
Paper A Mock Exam 9_ Attempt review.pdf.
medical_surgical_nursing_10th_edition_ignatavicius_TEST_BANK_pdf.pdf
FOISHS ANNUAL IMPLEMENTATION PLAN 2025.pdf
Vision Prelims GS PYQ Analysis 2011-2022 www.upscpdf.com.pdf
ChatGPT for Dummies - Pam Baker Ccesa007.pdf
LIFE & LIVING TRILOGY- PART (1) WHO ARE WE.pdf
Share_Module_2_Power_conflict_and_negotiation.pptx
ELIAS-SEZIURE AND EPilepsy semmioan session.pptx
Empowerment Technology for Senior High School Guide

Nonlinear transport phenomena: models, method of solving and unusual features (1)

  • 1. Nonlinear transport phenomena: models, method of solving and unusual features Vsevolod Vladimirov AGH University of Science and technology, Faculty of Applied Mathematics ´ Krakow, August 10, 2010 Summer School: KPI, 2010 Nonlinear transport phenomena 1 / 38
  • 2. THE AIM OF THIS LECTURE IS TO PRESENT THE SIMPLEST PATTERNS SUPPORTED BY TRANSPORT EQUATIONS WHAT ARE PATTERNS? Summer School: KPI, 2010 Nonlinear transport phenomena 2 / 38
  • 3. THE AIM OF THIS LECTURE IS TO PRESENT THE SIMPLEST PATTERNS SUPPORTED BY TRANSPORT EQUATIONS WHAT ARE PATTERNS? Summer School: KPI, 2010 Nonlinear transport phenomena 2 / 38
  • 4. Examples of patterns: Solution to the system of reaction-diffusion equation ut = Du ∆u + a(1 − u) vt = Dv ∆v + u v 2 − k v u(0, x) = f (x), v(0, x) = g(x), 2 ∂2 ∆= i=1 ∂x2 i periodically initiated cordial impulse; formation of the train of solitary waves; tsunami; tornado etc., etc. Summer School: KPI, 2010 Nonlinear transport phenomena 3 / 38
  • 5. Examples of patterns: Solution to the system of reaction-diffusion equation ut = Du ∆u + a(1 − u) vt = Dv ∆v + u v 2 − k v u(0, x) = f (x), v(0, x) = g(x), 2 ∂2 ∆= i=1 ∂x2 i periodically initiated cordial impulse; formation of the train of solitary waves; tsunami; tornado etc., etc. Summer School: KPI, 2010 Nonlinear transport phenomena 3 / 38
  • 6. Examples of patterns: Solution to the system of reaction-diffusion equation ut = Du ∆u + a(1 − u) vt = Dv ∆v + u v 2 − k v u(0, x) = f (x), v(0, x) = g(x), 2 ∂2 ∆= i=1 ∂x2 i periodically initiated cordial impulse; formation of the train of solitary waves; tsunami; tornado etc., etc. Summer School: KPI, 2010 Nonlinear transport phenomena 3 / 38
  • 7. Examples of patterns: Solution to the system of reaction-diffusion equation ut = Du ∆u + a(1 − u) vt = Dv ∆v + u v 2 − k v u(0, x) = f (x), v(0, x) = g(x), 2 ∂2 ∆= i=1 ∂x2 i periodically initiated cordial impulse; formation of the train of solitary waves; tsunami; tornado etc., etc. Summer School: KPI, 2010 Nonlinear transport phenomena 3 / 38
  • 8. Examples of patterns: Solution to the system of reaction-diffusion equation ut = Du ∆u + a(1 − u) vt = Dv ∆v + u v 2 − k v u(0, x) = f (x), v(0, x) = g(x), 2 ∂2 ∆= i=1 ∂x2 i periodically initiated cordial impulse; formation of the train of solitary waves; tsunami; tornado etc., etc. Summer School: KPI, 2010 Nonlinear transport phenomena 3 / 38
  • 9. Examples of patterns: Solution to the system of reaction-diffusion equation ut = Du ∆u + a(1 − u) vt = Dv ∆v + u v 2 − k v u(0, x) = f (x), v(0, x) = g(x), 2 ∂2 ∆= i=1 ∂x2 i periodically initiated cordial impulse; formation of the train of solitary waves; tsunami; tornado etc., etc. Summer School: KPI, 2010 Nonlinear transport phenomena 3 / 38
  • 10. Examples of patterns: Solution to the system of reaction-diffusion equation ut = Du ∆u + a(1 − u) vt = Dv ∆v + u v 2 − k v u(0, x) = f (x), v(0, x) = g(x), 2 ∂2 ∆= i=1 ∂x2 i periodically initiated cordial impulse; formation of the train of solitary waves; tsunami; tornado etc., etc. Summer School: KPI, 2010 Nonlinear transport phenomena 3 / 38
  • 11. Our aim to discuss the patterns supported by the various transport equations in case of one spatial variable. Advantages: ability to present and analyze the analytical solutions to non-linear models the possibility to analyze the problem by means of qualitative theory methods. By patterns we temporarily mean: non-monotonic solution to the modelling system, maintaining their shape during the evolution; or the non-monotonic solutions evolving in a self-similar mode or the non-monotonic solutions tending to infinity in finite time (blowing-up patterns) Summer School: KPI, 2010 Nonlinear transport phenomena 4 / 38
  • 12. Our aim to discuss the patterns supported by the various transport equations in case of one spatial variable. Advantages: ability to present and analyze the analytical solutions to non-linear models the possibility to analyze the problem by means of qualitative theory methods. By patterns we temporarily mean: non-monotonic solution to the modelling system, maintaining their shape during the evolution; or the non-monotonic solutions evolving in a self-similar mode or the non-monotonic solutions tending to infinity in finite time (blowing-up patterns) Summer School: KPI, 2010 Nonlinear transport phenomena 4 / 38
  • 13. Our aim to discuss the patterns supported by the various transport equations in case of one spatial variable. Advantages: ability to present and analyze the analytical solutions to non-linear models the possibility to analyze the problem by means of qualitative theory methods. By patterns we temporarily mean: non-monotonic solution to the modelling system, maintaining their shape during the evolution; or the non-monotonic solutions evolving in a self-similar mode or the non-monotonic solutions tending to infinity in finite time (blowing-up patterns) Summer School: KPI, 2010 Nonlinear transport phenomena 4 / 38
  • 14. Our aim to discuss the patterns supported by the various transport equations in case of one spatial variable. Advantages: ability to present and analyze the analytical solutions to non-linear models the possibility to analyze the problem by means of qualitative theory methods. By patterns we temporarily mean: non-monotonic solution to the modelling system, maintaining their shape during the evolution; or the non-monotonic solutions evolving in a self-similar mode or the non-monotonic solutions tending to infinity in finite time (blowing-up patterns) Summer School: KPI, 2010 Nonlinear transport phenomena 4 / 38
  • 15. Our aim to discuss the patterns supported by the various transport equations in case of one spatial variable. Advantages: ability to present and analyze the analytical solutions to non-linear models the possibility to analyze the problem by means of qualitative theory methods. By patterns we temporarily mean: non-monotonic solution to the modelling system, maintaining their shape during the evolution; or the non-monotonic solutions evolving in a self-similar mode or the non-monotonic solutions tending to infinity in finite time (blowing-up patterns) Summer School: KPI, 2010 Nonlinear transport phenomena 4 / 38
  • 16. Our aim to discuss the patterns supported by the various transport equations in case of one spatial variable. Advantages: ability to present and analyze the analytical solutions to non-linear models the possibility to analyze the problem by means of qualitative theory methods. By patterns we temporarily mean: non-monotonic solution to the modelling system, maintaining their shape during the evolution; or the non-monotonic solutions evolving in a self-similar mode or the non-monotonic solutions tending to infinity in finite time (blowing-up patterns) Summer School: KPI, 2010 Nonlinear transport phenomena 4 / 38
  • 17. Heat transport: the balance equation d ρ u(t, x)d x = − q dσ + f (u(t, x); t, x)d x, (1) dt Ω ∂Ω Ω where q is the density of the heat flux on the boundary ∂ Ω. On virtue of the Fick law, is as follows: q = −κ u(t, x); u(t, x) is the energy per unit mass (the temperature); ρ is the heat capacity per unit volume; κ(u; t, x) is the heat transport coefficient; f (u(t, x); t, x) is the voluminal heat source. Summer School: KPI, 2010 Nonlinear transport phenomena 5 / 38
  • 18. Heat transport: the balance equation d ρ u(t, x)d x = − q dσ + f (u(t, x); t, x)d x, (1) dt Ω ∂Ω Ω where q is the density of the heat flux on the boundary ∂ Ω. On virtue of the Fick law, is as follows: q = −κ u(t, x); u(t, x) is the energy per unit mass (the temperature); ρ is the heat capacity per unit volume; κ(u; t, x) is the heat transport coefficient; f (u(t, x); t, x) is the voluminal heat source. Summer School: KPI, 2010 Nonlinear transport phenomena 5 / 38
  • 19. Heat transport: the balance equation d ρ u(t, x)d x = − q dσ + f (u(t, x); t, x)d x, (1) dt Ω ∂Ω Ω where q is the density of the heat flux on the boundary ∂ Ω. On virtue of the Fick law, is as follows: q = −κ u(t, x); u(t, x) is the energy per unit mass (the temperature); ρ is the heat capacity per unit volume; κ(u; t, x) is the heat transport coefficient; f (u(t, x); t, x) is the voluminal heat source. Summer School: KPI, 2010 Nonlinear transport phenomena 5 / 38
  • 20. Heat transport: the balance equation d ρ u(t, x)d x = − q dσ + f (u(t, x); t, x)d x, (1) dt Ω ∂Ω Ω where q is the density of the heat flux on the boundary ∂ Ω. On virtue of the Fick law, is as follows: q = −κ u(t, x); u(t, x) is the energy per unit mass (the temperature); ρ is the heat capacity per unit volume; κ(u; t, x) is the heat transport coefficient; f (u(t, x); t, x) is the voluminal heat source. Summer School: KPI, 2010 Nonlinear transport phenomena 5 / 38
  • 21. Using the Green-Gauss-Ostrogradsky theorem, we are able to write down the following identities: − q dσ ≡ κ u(t, x) dσ = [κ u(t, x)] d x. ∂Ω ∂Ω Ω So, the balance equation can be rewritten as ∂ u(t, x) ρ − [κ u(t, x)] − f [u(t, x); t, x] d x = 0. Ω ∂t Since the volume Ω is arbitrary, fulfillment of the balance equation is possible providing that the integrand is equal to zero In the case of one spatial variable, we get this way the equation ∂ u(t, x) ∂ ∂ u(t, x) ˜ = κ (u(t, x), t, x) ˜ + f (u(t, x), t, x) , ∂t ∂x ∂x (2) ˜ κ = κ/ρ, f = f /ρ. ˜ Summer School: KPI, 2010 Nonlinear transport phenomena 6 / 38
  • 22. Using the Green-Gauss-Ostrogradsky theorem, we are able to write down the following identities: − q dσ ≡ κ u(t, x) dσ = [κ u(t, x)] d x. ∂Ω ∂Ω Ω So, the balance equation can be rewritten as ∂ u(t, x) ρ − [κ u(t, x)] − f [u(t, x); t, x] d x = 0. Ω ∂t Since the volume Ω is arbitrary, fulfillment of the balance equation is possible providing that the integrand is equal to zero In the case of one spatial variable, we get this way the equation ∂ u(t, x) ∂ ∂ u(t, x) ˜ = κ (u(t, x), t, x) ˜ + f (u(t, x), t, x) , ∂t ∂x ∂x (2) ˜ κ = κ/ρ, f = f /ρ. ˜ Summer School: KPI, 2010 Nonlinear transport phenomena 6 / 38
  • 23. Using the Green-Gauss-Ostrogradsky theorem, we are able to write down the following identities: − q dσ ≡ κ u(t, x) dσ = [κ u(t, x)] d x. ∂Ω ∂Ω Ω So, the balance equation can be rewritten as ∂ u(t, x) ρ − [κ u(t, x)] − f [u(t, x); t, x] d x = 0. Ω ∂t Since the volume Ω is arbitrary, fulfillment of the balance equation is possible providing that the integrand is equal to zero In the case of one spatial variable, we get this way the equation ∂ u(t, x) ∂ ∂ u(t, x) ˜ = κ (u(t, x), t, x) ˜ + f (u(t, x), t, x) , ∂t ∂x ∂x (2) ˜ κ = κ/ρ, f = f /ρ. ˜ Summer School: KPI, 2010 Nonlinear transport phenomena 6 / 38
  • 24. Using the Green-Gauss-Ostrogradsky theorem, we are able to write down the following identities: − q dσ ≡ κ u(t, x) dσ = [κ u(t, x)] d x. ∂Ω ∂Ω Ω So, the balance equation can be rewritten as ∂ u(t, x) ρ − [κ u(t, x)] − f [u(t, x); t, x] d x = 0. Ω ∂t Since the volume Ω is arbitrary, fulfillment of the balance equation is possible providing that the integrand is equal to zero In the case of one spatial variable, we get this way the equation ∂ u(t, x) ∂ ∂ u(t, x) ˜ = κ (u(t, x), t, x) ˜ + f (u(t, x), t, x) , ∂t ∂x ∂x (2) ˜ κ = κ/ρ, f = f /ρ. ˜ Summer School: KPI, 2010 Nonlinear transport phenomena 6 / 38
  • 25. Self-similar solutions to the heat equation. Statement 1. Any physical law can be written down in the form a = ϕ (a1 , a2 , ..., an ; an+1 , ...an+m ) , where a1 , a2 , ..., an are the physical quantities expressed in the basic units, a, an+1 , ...an+m are the physical quantities expressed in derived units. Example of basic units: the length[L], the time [T ], the mass [M ] Examples of physical quantities expressed in derived units: dx L d2 x M · L V = , F =m 2 , etc., etc. dt T dt T2 Summer School: KPI, 2010 Nonlinear transport phenomena 7 / 38
  • 26. Self-similar solutions to the heat equation. Statement 1. Any physical law can be written down in the form a = ϕ (a1 , a2 , ..., an ; an+1 , ...an+m ) , where a1 , a2 , ..., an are the physical quantities expressed in the basic units, a, an+1 , ...an+m are the physical quantities expressed in derived units. Example of basic units: the length[L], the time [T ], the mass [M ] Examples of physical quantities expressed in derived units: dx L d2 x M · L V = , F =m 2 , etc., etc. dt T dt T2 Summer School: KPI, 2010 Nonlinear transport phenomena 7 / 38
  • 27. Self-similar solutions to the heat equation. Statement 1. Any physical law can be written down in the form a = ϕ (a1 , a2 , ..., an ; an+1 , ...an+m ) , where a1 , a2 , ..., an are the physical quantities expressed in the basic units, a, an+1 , ...an+m are the physical quantities expressed in derived units. Example of basic units: the length[L], the time [T ], the mass [M ] Examples of physical quantities expressed in derived units: dx L d2 x M · L V = , F =m 2 , etc., etc. dt T dt T2 Summer School: KPI, 2010 Nonlinear transport phenomena 7 / 38
  • 28. Self-similar solutions to the heat equation. Statement 1. Any physical law can be written down in the form a = ϕ (a1 , a2 , ..., an ; an+1 , ...an+m ) , where a1 , a2 , ..., an are the physical quantities expressed in the basic units, a, an+1 , ...an+m are the physical quantities expressed in derived units. Example of basic units: the length[L], the time [T ], the mass [M ] Examples of physical quantities expressed in derived units: dx L d2 x M · L V = , F =m 2 , etc., etc. dt T dt T2 Summer School: KPI, 2010 Nonlinear transport phenomena 7 / 38
  • 29. Self-similar solutions to the heat equation. Statement 1. Any physical law can be written down in the form a = ϕ (a1 , a2 , ..., an ; an+1 , ...an+m ) , where a1 , a2 , ..., an are the physical quantities expressed in the basic units, a, an+1 , ...an+m are the physical quantities expressed in derived units. Example of basic units: the length[L], the time [T ], the mass [M ] Examples of physical quantities expressed in derived units: dx L d2 x M · L V = , F =m 2 , etc., etc. dt T dt T2 Summer School: KPI, 2010 Nonlinear transport phenomena 7 / 38
  • 30. Statement 2. Derived physical quantities a, an+1 , ...an+m can be expressed through the basic ones in the following form: a = ar1 ar2 ...arn Π, 1 2 n rj rj j an+j = a11 a22 ...arn Πj , n j = 1, 2, ...., m, where Π, Π1 , ...., Πm are dimensionless parameters. So, the physical law a = F (a1 , ...an ; ...an+j , ....) is equivalent to 1 rj rj rj Π= F a1 , a2 , ..., an ;...a11 a22 ...ann Πj ... . (3) ar1 1 ar2 2 ...arn n Summer School: KPI, 2010 Nonlinear transport phenomena 8 / 38
  • 31. Statement 2. Derived physical quantities a, an+1 , ...an+m can be expressed through the basic ones in the following form: a = ar1 ar2 ...arn Π, 1 2 n rj rj j an+j = a11 a22 ...arn Πj , n j = 1, 2, ...., m, where Π, Π1 , ...., Πm are dimensionless parameters. So, the physical law a = F (a1 , ...an ; ...an+j , ....) is equivalent to 1 rj rj rj Π= F a1 , a2 , ..., an ;...a11 a22 ...ann Πj ... . (3) ar1 1 ar2 2 ...arn n Summer School: KPI, 2010 Nonlinear transport phenomena 8 / 38
  • 32. The essence of Π theorem Theorem. The RHS of the formula (3) depends, at most, on the m dimensionless parameters Π1 , ...., Πm , and does not depend on the dimension quantities a1 , a2 , ..., an . 1 So, the physical law a = F (a1 , ...an ; ...an+j , ....) in dimensionless variables takes the form to Π = Ψ (Π1 , Π2 , ...Πm ) . (4) 1 A. Samarskij, A. Mikhailov, Mathematical Modelling: Ideas, Methods, Examples, Mosclw, 1997, Ch. V. Summer School: KPI, 2010 Nonlinear transport phenomena 9 / 38
  • 33. The essence of Π theorem Theorem. The RHS of the formula (3) depends, at most, on the m dimensionless parameters Π1 , ...., Πm , and does not depend on the dimension quantities a1 , a2 , ..., an . 1 So, the physical law a = F (a1 , ...an ; ...an+j , ....) in dimensionless variables takes the form to Π = Ψ (Π1 , Π2 , ...Πm ) . (4) 1 A. Samarskij, A. Mikhailov, Mathematical Modelling: Ideas, Methods, Examples, Mosclw, 1997, Ch. V. Summer School: KPI, 2010 Nonlinear transport phenomena 9 / 38
  • 34. The essence of Π theorem Theorem. The RHS of the formula (3) depends, at most, on the m dimensionless parameters Π1 , ...., Πm , and does not depend on the dimension quantities a1 , a2 , ..., an . 1 So, the physical law a = F (a1 , ...an ; ...an+j , ....) in dimensionless variables takes the form to Π = Ψ (Π1 , Π2 , ...Πm ) . (4) 1 A. Samarskij, A. Mikhailov, Mathematical Modelling: Ideas, Methods, Examples, Mosclw, 1997, Ch. V. Summer School: KPI, 2010 Nonlinear transport phenomena 9 / 38
  • 35. Applications. Point explosion (linear case) u t = κ ux x , (5) ∞ u(0, x) = 0, x ∈ R, x = 0, u(t, x) d x = Q = const. −∞ The solution to (5) can be expressed as u = f (t, κ, Q; x). Basic units: E · L2 t[T ], κ , Q[E · L]. T Derived units: Q √ u = √ Π, x = κ t Π1 . κt Thus, passing to dimensionless variables we get √ κt √ Π= f (t, κ, Q; κ t Π1 ) = ϕ (Π1 ) . Q Summer School: KPI, 2010 Nonlinear transport phenomena 10 / 38
  • 36. Applications. Point explosion (linear case) u t = κ ux x , (5) ∞ u(0, x) = 0, x ∈ R, x = 0, u(t, x) d x = Q = const. −∞ The solution to (5) can be expressed as u = f (t, κ, Q; x). Basic units: E · L2 t[T ], κ , Q[E · L]. T Derived units: Q √ u = √ Π, x = κ t Π1 . κt Thus, passing to dimensionless variables we get √ κt √ Π= f (t, κ, Q; κ t Π1 ) = ϕ (Π1 ) . Q Summer School: KPI, 2010 Nonlinear transport phenomena 10 / 38
  • 37. Applications. Point explosion (linear case) u t = κ ux x , (5) ∞ u(0, x) = 0, x ∈ R, x = 0, u(t, x) d x = Q = const. −∞ The solution to (5) can be expressed as u = f (t, κ, Q; x). Basic units: E · L2 t[T ], κ , Q[E · L]. T Derived units: Q √ u = √ Π, x = κ t Π1 . κt Thus, passing to dimensionless variables we get √ κt √ Π= f (t, κ, Q; κ t Π1 ) = ϕ (Π1 ) . Q Summer School: KPI, 2010 Nonlinear transport phenomena 10 / 38
  • 38. Applications. Point explosion (linear case) u t = κ ux x , (5) ∞ u(0, x) = 0, x ∈ R, x = 0, u(t, x) d x = Q = const. −∞ The solution to (5) can be expressed as u = f (t, κ, Q; x). Basic units: E · L2 t[T ], κ , Q[E · L]. T Derived units: Q √ u = √ Π, x = κ t Π1 . κt Thus, passing to dimensionless variables we get √ κt √ Π= f (t, κ, Q; κ t Π1 ) = ϕ (Π1 ) . Q Summer School: KPI, 2010 Nonlinear transport phenomena 10 / 38
  • 39. Applications. Point explosion (linear case) u t = κ ux x , (5) ∞ u(0, x) = 0, x ∈ R, x = 0, u(t, x) d x = Q = const. −∞ The solution to (5) can be expressed as u = f (t, κ, Q; x). Basic units: E · L2 t[T ], κ , Q[E · L]. T Derived units: Q √ u = √ Π, x = κ t Π1 . κt Thus, passing to dimensionless variables we get √ κt √ Π= f (t, κ, Q; κ t Π1 ) = ϕ (Π1 ) . Q Summer School: KPI, 2010 Nonlinear transport phenomena 10 / 38
  • 40. Corollary.Ansatz Q u = √ ϕ(ξ), ξ = Π1 κt should strongly simplify the point explosion problem (5). In fact, inserting this ansatz into the heat transport equation, we get d 2 ϕ[ξ] + ξ ϕ[ξ] + ϕ[ξ] ≡ ¨ ˙ (ξϕ[ξ] + 2 ϕ[ξ]) = 0, ˙ (6) dξ with the additional condition (as well expressed in the dimensionless variables): ∞ ϕ[ξ] dξ = 1. (7) −∞ Summer School: KPI, 2010 Nonlinear transport phenomena 11 / 38
  • 41. Corollary.Ansatz Q u = √ ϕ(ξ), ξ = Π1 κt should strongly simplify the point explosion problem (5). In fact, inserting this ansatz into the heat transport equation, we get d 2 ϕ[ξ] + ξ ϕ[ξ] + ϕ[ξ] ≡ ¨ ˙ (ξϕ[ξ] + 2 ϕ[ξ]) = 0, ˙ (6) dξ with the additional condition (as well expressed in the dimensionless variables): ∞ ϕ[ξ] dξ = 1. (7) −∞ Summer School: KPI, 2010 Nonlinear transport phenomena 11 / 38
  • 42. Corollary.Ansatz Q u = √ ϕ(ξ), ξ = Π1 κt should strongly simplify the point explosion problem (5). In fact, inserting this ansatz into the heat transport equation, we get d 2 ϕ[ξ] + ξ ϕ[ξ] + ϕ[ξ] ≡ ¨ ˙ (ξϕ[ξ] + 2 ϕ[ξ]) = 0, ˙ (6) dξ with the additional condition (as well expressed in the dimensionless variables): ∞ ϕ[ξ] dξ = 1. (7) −∞ Summer School: KPI, 2010 Nonlinear transport phenomena 11 / 38
  • 43. Statement 3. Solution to the problem d (ξϕ[ξ] + 2 ϕ[ξ]) = 0, ˙ dξ ∞ ϕ[ξ] dξ = 1 −∞ is given by the formula 1 ξ2 ϕ[ξ] = √ e− 4 . 4π Corollary. The only solution to the linear point explosion problem (5) is given by the formula Q x2 u= √ e− 4 κ t . (8) 4κπt Summer School: KPI, 2010 Nonlinear transport phenomena 12 / 38
  • 44. Statement 3. Solution to the problem d (ξϕ[ξ] + 2 ϕ[ξ]) = 0, ˙ dξ ∞ ϕ[ξ] dξ = 1 −∞ is given by the formula 1 ξ2 ϕ[ξ] = √ e− 4 . 4π Corollary. The only solution to the linear point explosion problem (5) is given by the formula Q x2 u= √ e− 4 κ t . (8) 4κπt Summer School: KPI, 2010 Nonlinear transport phenomena 12 / 38
  • 45. Figure: Summer School: KPI, 2010 Nonlinear transport phenomena 13 / 38
  • 46. Figure: Summer School: KPI, 2010 Nonlinear transport phenomena 14 / 38
  • 47. Figure: Summer School: KPI, 2010 Nonlinear transport phenomena 15 / 38
  • 48. Remark 1. For Q = 1 the previous formula written as 1 x2 E=√ e− 4 κ t 4πκt defines so called fundamental solution of the operator ˆ ∂ ∂2 L= − κ 2. ∂t ∂x Knowledge of this solution enables to solve the initial value problem. Theorem Solution to the initial value problem ut − κ u x x , u(0, x) = F (x) is given by the formula +∞ u(t, x) = E ∗ F (t, x) ≡ E (t, x − y) · F (y) d y −∞ provided that the integral in the RHS does exist. Summer School: KPI, 2010 Nonlinear transport phenomena 16 / 38
  • 49. Remark 1. For Q = 1 the previous formula written as 1 x2 E=√ e− 4 κ t 4πκt defines so called fundamental solution of the operator ˆ ∂ ∂2 L= − κ 2. ∂t ∂x Knowledge of this solution enables to solve the initial value problem. Theorem Solution to the initial value problem ut − κ u x x , u(0, x) = F (x) is given by the formula +∞ u(t, x) = E ∗ F (t, x) ≡ E (t, x − y) · F (y) d y −∞ provided that the integral in the RHS does exist. Summer School: KPI, 2010 Nonlinear transport phenomena 16 / 38
  • 50. Peculiarity of the solution Q x2 u= √ e− 4 κ t 4κπt to the linear point explosion: At any moment of time all the space ”knows ” that the point explosion took place. This is physically incorrect!!! Summer School: KPI, 2010 Nonlinear transport phenomena 17 / 38
  • 51. Peculiarity of the solution Q x2 u= √ e− 4 κ t 4κπt to the linear point explosion: At any moment of time all the space ”knows ” that the point explosion took place. This is physically incorrect!!! Summer School: KPI, 2010 Nonlinear transport phenomena 17 / 38
  • 52. Point explosion (nonlinear case) ut = κ [u ux ]x , (9) ∞ u(0, x) = 0, x ∈ R, x = 0, u(t, x) d x = Q = const. −∞ The solution to (9) can be expressed as u = f (t, κ, Q; x). Basic units: L2 t[T ], κ , Q[E · L]. E·T Derived units: Q2/3 u= Π, x = [κ Q t]1/3 Π1 . [κ t]1/3 Thus, passing to dimensionless variables we get [κ t]1/3 Π= f t, κ, Q; Π1 [κ Q t]1/3 = ϕ (Π1 ) . Q2/3 Summer School: KPI, 2010 Nonlinear transport phenomena 18 / 38
  • 53. Point explosion (nonlinear case) ut = κ [u ux ]x , (9) ∞ u(0, x) = 0, x ∈ R, x = 0, u(t, x) d x = Q = const. −∞ The solution to (9) can be expressed as u = f (t, κ, Q; x). Basic units: L2 t[T ], κ , Q[E · L]. E·T Derived units: Q2/3 u= Π, x = [κ Q t]1/3 Π1 . [κ t]1/3 Thus, passing to dimensionless variables we get [κ t]1/3 Π= f t, κ, Q; Π1 [κ Q t]1/3 = ϕ (Π1 ) . Q2/3 Summer School: KPI, 2010 Nonlinear transport phenomena 18 / 38
  • 54. Point explosion (nonlinear case) ut = κ [u ux ]x , (9) ∞ u(0, x) = 0, x ∈ R, x = 0, u(t, x) d x = Q = const. −∞ The solution to (9) can be expressed as u = f (t, κ, Q; x). Basic units: L2 t[T ], κ , Q[E · L]. E·T Derived units: Q2/3 u= Π, x = [κ Q t]1/3 Π1 . [κ t]1/3 Thus, passing to dimensionless variables we get [κ t]1/3 Π= f t, κ, Q; Π1 [κ Q t]1/3 = ϕ (Π1 ) . Q2/3 Summer School: KPI, 2010 Nonlinear transport phenomena 18 / 38
  • 55. Point explosion (nonlinear case) ut = κ [u ux ]x , (9) ∞ u(0, x) = 0, x ∈ R, x = 0, u(t, x) d x = Q = const. −∞ The solution to (9) can be expressed as u = f (t, κ, Q; x). Basic units: L2 t[T ], κ , Q[E · L]. E·T Derived units: Q2/3 u= Π, x = [κ Q t]1/3 Π1 . [κ t]1/3 Thus, passing to dimensionless variables we get [κ t]1/3 Π= f t, κ, Q; Π1 [κ Q t]1/3 = ϕ (Π1 ) . Q2/3 Summer School: KPI, 2010 Nonlinear transport phenomena 18 / 38
  • 56. Point explosion (nonlinear case) ut = κ [u ux ]x , (9) ∞ u(0, x) = 0, x ∈ R, x = 0, u(t, x) d x = Q = const. −∞ The solution to (9) can be expressed as u = f (t, κ, Q; x). Basic units: L2 t[T ], κ , Q[E · L]. E·T Derived units: Q2/3 u= Π, x = [κ Q t]1/3 Π1 . [κ t]1/3 Thus, passing to dimensionless variables we get [κ t]1/3 Π= f t, κ, Q; Π1 [κ Q t]1/3 = ϕ (Π1 ) . Q2/3 Summer School: KPI, 2010 Nonlinear transport phenomena 18 / 38
  • 57. Corollary. Ansatz Q x u= ϕ(ξ), ξ = Π1 = [κ Q t]1/3 [κ Q t]1/3 should strongly simplify the point explosion problem (9). In fact, inserting this ansatz into the heat transport equation, we get d (ξϕ[ξ] + 3 ϕ[ξ] ϕ[ξ]) = 0, ˙ (10) dξ with the additional condition (as well expressed in the dimensionless variables): ∞ ϕ[ξ] dξ = 1. (11) −∞ Summer School: KPI, 2010 Nonlinear transport phenomena 19 / 38
  • 58. Corollary. Ansatz Q x u= ϕ(ξ), ξ = Π1 = [κ Q t]1/3 [κ Q t]1/3 should strongly simplify the point explosion problem (9). In fact, inserting this ansatz into the heat transport equation, we get d (ξϕ[ξ] + 3 ϕ[ξ] ϕ[ξ]) = 0, ˙ (10) dξ with the additional condition (as well expressed in the dimensionless variables): ∞ ϕ[ξ] dξ = 1. (11) −∞ Summer School: KPI, 2010 Nonlinear transport phenomena 19 / 38
  • 59. Corollary. Ansatz Q x u= ϕ(ξ), ξ = Π1 = [κ Q t]1/3 [κ Q t]1/3 should strongly simplify the point explosion problem (9). In fact, inserting this ansatz into the heat transport equation, we get d (ξϕ[ξ] + 3 ϕ[ξ] ϕ[ξ]) = 0, ˙ (10) dξ with the additional condition (as well expressed in the dimensionless variables): ∞ ϕ[ξ] dξ = 1. (11) −∞ Summer School: KPI, 2010 Nonlinear transport phenomena 19 / 38
  • 60. Statement 4. Solution to the problem d (ξϕ[ξ] + 3 ϕ[ξ] ϕ[ξ]) = 0, ˙ dξ ∞ ϕ[ξ] dξ = 1 −∞ is given by the formula 1 6 ξΦ − ξ 2 if |ξ| < ξΦ , 2 ϕ[ξ] = 0 otherwise, where ξΦ = (9/2)1/3 . Corollary. The only solution to the nonlinear point explosion problem (9) is given by the formula 1/6   (9/2)1/3 − x2 t 2 Q  [κ Q t]2/3 if |x| < 9 (κ 2 Q) , u= 6 [κ Q t]1/3  0 otherwise  (12) Summer School: KPI, 2010 Nonlinear transport phenomena 20 / 38
  • 61. Statement 4. Solution to the problem d (ξϕ[ξ] + 3 ϕ[ξ] ϕ[ξ]) = 0, ˙ dξ ∞ ϕ[ξ] dξ = 1 −∞ is given by the formula 1 6 ξΦ − ξ 2 if |ξ| < ξΦ , 2 ϕ[ξ] = 0 otherwise, where ξΦ = (9/2)1/3 . Corollary. The only solution to the nonlinear point explosion problem (9) is given by the formula 1/6   (9/2)1/3 − x2 t 2 Q  [κ Q t]2/3 if |x| < 9 (κ 2 Q) , u= 6 [κ Q t]1/3  0 otherwise  (12) Summer School: KPI, 2010 Nonlinear transport phenomena 20 / 38
  • 62. Statement 4. Solution to the problem d (ξϕ[ξ] + 3 ϕ[ξ] ϕ[ξ]) = 0, ˙ dξ ∞ ϕ[ξ] dξ = 1 −∞ is given by the formula 1 6 ξΦ − ξ 2 if |ξ| < ξΦ , 2 ϕ[ξ] = 0 otherwise, where ξΦ = (9/2)1/3 . Corollary. The only solution to the nonlinear point explosion problem (9) is given by the formula 1/6   (9/2)1/3 − x2 t 2 Q  [κ Q t]2/3 if |x| < 9 (κ 2 Q) , u= 6 [κ Q t]1/3  0 otherwise  (12) Summer School: KPI, 2010 Nonlinear transport phenomena 20 / 38
  • 63. Figure: Summer School: KPI, 2010 Nonlinear transport phenomena 21 / 38
  • 64. Figure: Summer School: KPI, 2010 Nonlinear transport phenomena 22 / 38
  • 65. Figure: Summer School: KPI, 2010 Nonlinear transport phenomena 23 / 38
  • 66. Figure: Summer School: KPI, 2010 Nonlinear transport phenomena 24 / 38
  • 67. Conclusion. Nonlinearity of the heat transport coefficient causes the localization of the heat wave within the finite domain < −L(t), L(t) >, where 1/6 9 (κ t Q)2 L(t) = = C t1/3 . 2 Summer School: KPI, 2010 Nonlinear transport phenomena 25 / 38
  • 68. Unlimited growth of temperature and the complete localization of the heat energy ut = κ [uσ ux ]x , (13) −n u(t, 0) = A (−t) , t < 0, n > 0, x ∈ R+ , lim u(t, x) = 0(14) t→−∞ In general, the problem is self-similar and the ansatz A x u= ϕ(ξ), ξ=√ (−t)n Aσ κ t1−n σ reduces the initial problem to corresponding ODE. If we assume that n = σ = 1, than ξ will not depend on t: A x u= ϕ(ξ), ξ=√ . (−t) Aκ Summer School: KPI, 2010 Nonlinear transport phenomena 26 / 38
  • 69. Unlimited growth of temperature and the complete localization of the heat energy ut = κ [uσ ux ]x , (13) −n u(t, 0) = A (−t) , t < 0, n > 0, x ∈ R+ , lim u(t, x) = 0(14) t→−∞ In general, the problem is self-similar and the ansatz A x u= ϕ(ξ), ξ=√ (−t)n Aσ κ t1−n σ reduces the initial problem to corresponding ODE. If we assume that n = σ = 1, than ξ will not depend on t: A x u= ϕ(ξ), ξ=√ . (−t) Aκ Summer School: KPI, 2010 Nonlinear transport phenomena 26 / 38
  • 70. Unlimited growth of temperature and the complete localization of the heat energy ut = κ [uσ ux ]x , (13) −n u(t, 0) = A (−t) , t < 0, n > 0, x ∈ R+ , lim u(t, x) = 0(14) t→−∞ In general, the problem is self-similar and the ansatz A x u= ϕ(ξ), ξ=√ (−t)n Aσ κ t1−n σ reduces the initial problem to corresponding ODE. If we assume that n = σ = 1, than ξ will not depend on t: A x u= ϕ(ξ), ξ=√ . (−t) Aκ Summer School: KPI, 2010 Nonlinear transport phenomena 26 / 38
  • 71. Unlimited growth of temperature and the complete localization of the heat energy ut = κ [uσ ux ]x , (13) −n u(t, 0) = A (−t) , t < 0, n > 0, x ∈ R+ , lim u(t, x) = 0(14) t→−∞ In general, the problem is self-similar and the ansatz A x u= ϕ(ξ), ξ=√ (−t)n Aσ κ t1−n σ reduces the initial problem to corresponding ODE. If we assume that n = σ = 1, than ξ will not depend on t: A x u= ϕ(ξ), ξ=√ . (−t) Aκ Summer School: KPI, 2010 Nonlinear transport phenomena 26 / 38
  • 72. The reduced equation ϕϕ + ϕ2 − ϕ = 0, ¨ ˙ together with the initial condition ϕ(0) = 1 (arising from the A condition u(t, 0) = (−t) ), has the solution √ √ (1 − ξ/ 6)2 , if 0 < ξ < √ 6, ϕ(ξ) = 0 if ξ ≥ 6, The corresponding solution to the BV problem (13) is as follows: A 2 √ 1− √ x , if 0 < x < 6 κ A, u(t, x) = −t 6κA √ 0 if x ≥ 6 κ A. Moral of the tale: [1]. Heat energy is completely localized in √ the segment 0 < x < 6 κ A0 ; [2]. The√temperature u(t, x) tends to +∞ in every point x ∈ [0, 6 κ A0 ) as t → 0−. Summer School: KPI, 2010 Nonlinear transport phenomena 27 / 38
  • 73. The reduced equation ϕϕ + ϕ2 − ϕ = 0, ¨ ˙ together with the initial condition ϕ(0) = 1 (arising from the A condition u(t, 0) = (−t) ), has the solution √ √ (1 − ξ/ 6)2 , if 0 < ξ < √ 6, ϕ(ξ) = 0 if ξ ≥ 6, The corresponding solution to the BV problem (13) is as follows: A 2 √ 1− √ x , if 0 < x < 6 κ A, u(t, x) = −t 6κA √ 0 if x ≥ 6 κ A. Moral of the tale: [1]. Heat energy is completely localized in √ the segment 0 < x < 6 κ A0 ; [2]. The√temperature u(t, x) tends to +∞ in every point x ∈ [0, 6 κ A0 ) as t → 0−. Summer School: KPI, 2010 Nonlinear transport phenomena 27 / 38
  • 74. The reduced equation ϕϕ + ϕ2 − ϕ = 0, ¨ ˙ together with the initial condition ϕ(0) = 1 (arising from the A condition u(t, 0) = (−t) ), has the solution √ √ (1 − ξ/ 6)2 , if 0 < ξ < √ 6, ϕ(ξ) = 0 if ξ ≥ 6, The corresponding solution to the BV problem (13) is as follows: A 2 √ 1− √ x , if 0 < x < 6 κ A, u(t, x) = −t 6κA √ 0 if x ≥ 6 κ A. Moral of the tale: [1]. Heat energy is completely localized in √ the segment 0 < x < 6 κ A0 ; [2]. The√temperature u(t, x) tends to +∞ in every point x ∈ [0, 6 κ A0 ) as t → 0−. Summer School: KPI, 2010 Nonlinear transport phenomena 27 / 38
  • 75. The reduced equation ϕϕ + ϕ2 − ϕ = 0, ¨ ˙ together with the initial condition ϕ(0) = 1 (arising from the A condition u(t, 0) = (−t) ), has the solution √ √ (1 − ξ/ 6)2 , if 0 < ξ < √ 6, ϕ(ξ) = 0 if ξ ≥ 6, The corresponding solution to the BV problem (13) is as follows: A 2 √ 1− √ x , if 0 < x < 6 κ A, u(t, x) = −t 6κA √ 0 if x ≥ 6 κ A. Moral of the tale: [1]. Heat energy is completely localized in √ the segment 0 < x < 6 κ A0 ; [2]. The√temperature u(t, x) tends to +∞ in every point x ∈ [0, 6 κ A0 ) as t → 0−. Summer School: KPI, 2010 Nonlinear transport phenomena 27 / 38
  • 76. Figure: Summer School: KPI, 2010 Nonlinear transport phenomena 28 / 38
  • 77. Figure: Summer School: KPI, 2010 Nonlinear transport phenomena 29 / 38
  • 78. Figure: Summer School: KPI, 2010 Nonlinear transport phenomena 30 / 38
  • 79. Blow-up regime caused by the source term incorporation Model equation: ut = κ [u ux ]x + q0 u2 . (15) Ansatz u(t, x) = P (t) Q(x) leads, after some algebraic manipulation, to the equations P −2 (t)P (t) = κ(Q(x) Q(x) ) + q0 Q(x)2 Q−1 (x) = C1 . (16) ˙ Theorem. The equation (15) has the following solution: q0 2 4 cos2 x , if |x| < 2 π0 κ , u(t, x) = 8κ q 3 q0 (tf − t) 0 otherwise. (17) The above solution 2 π2 κ is localized within the interval [−L, L], L = q0 ; tends to +∞ ∀ x ∈ (−L, L), as t → tf − 0. Summer School: KPI, 2010 Nonlinear transport phenomena 31 / 38
  • 80. Blow-up regime caused by the source term incorporation Model equation: ut = κ [u ux ]x + q0 u2 . (15) Ansatz u(t, x) = P (t) Q(x) leads, after some algebraic manipulation, to the equations P −2 (t)P (t) = κ(Q(x) Q(x) ) + q0 Q(x)2 Q−1 (x) = C1 . (16) ˙ Theorem. The equation (15) has the following solution: q0 2 4 cos2 x , if |x| < 2 π0 κ , u(t, x) = 8κ q 3 q0 (tf − t) 0 otherwise. (17) The above solution 2 π2 κ is localized within the interval [−L, L], L = q0 ; tends to +∞ ∀ x ∈ (−L, L), as t → tf − 0. Summer School: KPI, 2010 Nonlinear transport phenomena 31 / 38
  • 81. Blow-up regime caused by the source term incorporation Model equation: ut = κ [u ux ]x + q0 u2 . (15) Ansatz u(t, x) = P (t) Q(x) leads, after some algebraic manipulation, to the equations P −2 (t)P (t) = κ(Q(x) Q(x) ) + q0 Q(x)2 Q−1 (x) = C1 . (16) ˙ Theorem. The equation (15) has the following solution: q0 2 4 cos2 x , if |x| < 2 π0 κ , u(t, x) = 8κ q 3 q0 (tf − t) 0 otherwise. (17) The above solution 2 π2 κ is localized within the interval [−L, L], L = q0 ; tends to +∞ ∀ x ∈ (−L, L), as t → tf − 0. Summer School: KPI, 2010 Nonlinear transport phenomena 31 / 38
  • 82. Blow-up regime caused by the source term incorporation Model equation: ut = κ [u ux ]x + q0 u2 . (15) Ansatz u(t, x) = P (t) Q(x) leads, after some algebraic manipulation, to the equations P −2 (t)P (t) = κ(Q(x) Q(x) ) + q0 Q(x)2 Q−1 (x) = C1 . (16) ˙ Theorem. The equation (15) has the following solution: q0 2 4 cos2 x , if |x| < 2 π0 κ , u(t, x) = 8κ q 3 q0 (tf − t) 0 otherwise. (17) The above solution 2 π2 κ is localized within the interval [−L, L], L = q0 ; tends to +∞ ∀ x ∈ (−L, L), as t → tf − 0. Summer School: KPI, 2010 Nonlinear transport phenomena 31 / 38
  • 83. Blow-up regime caused by the source term incorporation Model equation: ut = κ [u ux ]x + q0 u2 . (15) Ansatz u(t, x) = P (t) Q(x) leads, after some algebraic manipulation, to the equations P −2 (t)P (t) = κ(Q(x) Q(x) ) + q0 Q(x)2 Q−1 (x) = C1 . (16) ˙ Theorem. The equation (15) has the following solution: q0 2 4 cos2 x , if |x| < 2 π0 κ , u(t, x) = 8κ q 3 q0 (tf − t) 0 otherwise. (17) The above solution 2 π2 κ is localized within the interval [−L, L], L = q0 ; tends to +∞ ∀ x ∈ (−L, L), as t → tf − 0. Summer School: KPI, 2010 Nonlinear transport phenomena 31 / 38
  • 84. Blow-up regime caused by the source term incorporation Model equation: ut = κ [u ux ]x + q0 u2 . (15) Ansatz u(t, x) = P (t) Q(x) leads, after some algebraic manipulation, to the equations P −2 (t)P (t) = κ(Q(x) Q(x) ) + q0 Q(x)2 Q−1 (x) = C1 . (16) ˙ Theorem. The equation (15) has the following solution: q0 2 4 cos2 x , if |x| < 2 π0 κ , u(t, x) = 8κ q 3 q0 (tf − t) 0 otherwise. (17) The above solution 2 π2 κ is localized within the interval [−L, L], L = q0 ; tends to +∞ ∀ x ∈ (−L, L), as t → tf − 0. Summer School: KPI, 2010 Nonlinear transport phenomena 31 / 38
  • 85. Figure: Summer School: KPI, 2010 Nonlinear transport phenomena 32 / 38
  • 86. Figure: Summer School: KPI, 2010 Nonlinear transport phenomena 33 / 38
  • 87. Figure: Summer School: KPI, 2010 Nonlinear transport phenomena 34 / 38
  • 88. Figure: Summer School: KPI, 2010 Nonlinear transport phenomena 35 / 38
  • 89. THE ROLE OF SELF-SIMILAR SOLUTIONS Summer School: KPI, 2010 Nonlinear transport phenomena 36 / 38
  • 90. Maximum principle. Comparison theorems ( initial value problem). Let us consider the Cauchy problem ut = [κ(u) ux ]x , (18) u(0, x) = u0 (x) ≥ 0, x ∈ R, κ(u) > 0. Theorem 1. Maximum of the solution u(t, x) at any time t > 0 does not exceed the maximum of the initial data: maxt>0, −∞<x<∞ u(t, x) ≤ max−∞<x<∞ u0 (x). 2 2 A. Samarskij, A. Mikhailov, Mathematical Modelling: Ideas, Methods, Examples, Moscow, 1997, A. Samarskij, V. Galaktionov et al, Blow-up regimes in quasilinear parabolic equations, Berlin: Walter de Gruyter, 1995. Summer School: KPI, 2010 Nonlinear transport phenomena 37 / 38
  • 91. Maximum principle. Comparison theorems ( initial value problem). Let us consider the Cauchy problem ut = [κ(u) ux ]x , (18) u(0, x) = u0 (x) ≥ 0, x ∈ R, κ(u) > 0. Theorem 1. Maximum of the solution u(t, x) at any time t > 0 does not exceed the maximum of the initial data: maxt>0, −∞<x<∞ u(t, x) ≤ max−∞<x<∞ u0 (x). 2 2 A. Samarskij, A. Mikhailov, Mathematical Modelling: Ideas, Methods, Examples, Moscow, 1997, A. Samarskij, V. Galaktionov et al, Blow-up regimes in quasilinear parabolic equations, Berlin: Walter de Gruyter, 1995. Summer School: KPI, 2010 Nonlinear transport phenomena 37 / 38
  • 92. Maximum principle. Comparison theorems ( initial value problem). Let us consider the Cauchy problem ut = [κ(u) ux ]x , (18) u(0, x) = u0 (x) ≥ 0, x ∈ R, κ(u) > 0. Theorem 1. Maximum of the solution u(t, x) at any time t > 0 does not exceed the maximum of the initial data: maxt>0, −∞<x<∞ u(t, x) ≤ max−∞<x<∞ u0 (x). 2 2 A. Samarskij, A. Mikhailov, Mathematical Modelling: Ideas, Methods, Examples, Moscow, 1997, A. Samarskij, V. Galaktionov et al, Blow-up regimes in quasilinear parabolic equations, Berlin: Walter de Gruyter, 1995. Summer School: KPI, 2010 Nonlinear transport phenomena 37 / 38
  • 93. Maximum principle. Comparison theorems ( initial value problem). Let us consider the Cauchy problem ut = [κ(u) ux ]x , (18) u(0, x) = u0 (x) ≥ 0, x ∈ R, κ(u) > 0. Theorem 1. Maximum of the solution u(t, x) at any time t > 0 does not exceed the maximum of the initial data: maxt>0, −∞<x<∞ u(t, x) ≤ max−∞<x<∞ u0 (x). 2 2 A. Samarskij, A. Mikhailov, Mathematical Modelling: Ideas, Methods, Examples, Moscow, 1997, A. Samarskij, V. Galaktionov et al, Blow-up regimes in quasilinear parabolic equations, Berlin: Walter de Gruyter, 1995. Summer School: KPI, 2010 Nonlinear transport phenomena 37 / 38
  • 94. Corollary (comparison theorem). Let u1 (t, x), u2 (t, x), u3 (t, x), are the solutions to the equation ut = [κ(u) ux ]x , u(0, x) = u0 (x) ≥ 0, x ∈ R, κ(u) > 0, corresponding to the initial data u1 (x), u2 (x), and u3 (x) 0 0 0 such that u1 (x) ≤ u2 (x) ≤ u3 (x) 0 0 0 for −∞ < x < ∞. Then u1 (t, x) ≤ u2 (t, x) ≤ u3 (t, x) for all −∞ < x < ∞, and t > 0. Summer School: KPI, 2010 Nonlinear transport phenomena 38 / 38
  • 95. Corollary (comparison theorem). Let u1 (t, x), u2 (t, x), u3 (t, x), are the solutions to the equation ut = [κ(u) ux ]x , u(0, x) = u0 (x) ≥ 0, x ∈ R, κ(u) > 0, corresponding to the initial data u1 (x), u2 (x), and u3 (x) 0 0 0 such that u1 (x) ≤ u2 (x) ≤ u3 (x) 0 0 0 for −∞ < x < ∞. Then u1 (t, x) ≤ u2 (t, x) ≤ u3 (t, x) for all −∞ < x < ∞, and t > 0. Summer School: KPI, 2010 Nonlinear transport phenomena 38 / 38
  • 96. Corollary (comparison theorem). Let u1 (t, x), u2 (t, x), u3 (t, x), are the solutions to the equation ut = [κ(u) ux ]x , u(0, x) = u0 (x) ≥ 0, x ∈ R, κ(u) > 0, corresponding to the initial data u1 (x), u2 (x), and u3 (x) 0 0 0 such that u1 (x) ≤ u2 (x) ≤ u3 (x) 0 0 0 for −∞ < x < ∞. Then u1 (t, x) ≤ u2 (t, x) ≤ u3 (t, x) for all −∞ < x < ∞, and t > 0. Summer School: KPI, 2010 Nonlinear transport phenomena 38 / 38
  • 97. Corollary (comparison theorem). Let u1 (t, x), u2 (t, x), u3 (t, x), are the solutions to the equation ut = [κ(u) ux ]x , u(0, x) = u0 (x) ≥ 0, x ∈ R, κ(u) > 0, corresponding to the initial data u1 (x), u2 (x), and u3 (x) 0 0 0 such that u1 (x) ≤ u2 (x) ≤ u3 (x) 0 0 0 for −∞ < x < ∞. Then u1 (t, x) ≤ u2 (t, x) ≤ u3 (t, x) for all −∞ < x < ∞, and t > 0. Summer School: KPI, 2010 Nonlinear transport phenomena 38 / 38