The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models...

29
General Introduction Analysis Numerical Simulations References The Discrete Unbounded Coagulation-Fragmentation Equation with Growth, Decay and Sedimentation Luke O. Joel University of KwaZulu-Natal South Africa Supervisors Prof. J. Banasiak Dr. S. Shindin L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Transcript of The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models...

Page 1: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

The Discrete UnboundedCoagulation-Fragmentation Equation with

Growth, Decay and Sedimentation

Luke O. JoelUniversity of KwaZulu-Natal

South Africa

Supervisors

Prof. J. Banasiak

Dr. S. Shindin

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 2: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Content

1 General IntroductionIntroductionFrom Literature

2 AnalysisLinear PartNonlinear Part

3 Numerical SimulationsTruncated EquationSimulations

4 ReferencesReferencesThank you

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 3: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

IntroductionFrom Literature

Focus of Presentation

1 In this presentation, we study the discretecoagulation–fragmentation models with growth, decay andsedimentation. We demonstrate the existence and uniquenessof classical global solutions, using the theory of linear andsemilinear semigroups of operators, provided the linearprocesses are sufficiently strong.

2 The conclusions obtained from the theories are supported bynumerical simulations.

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 4: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

IntroductionFrom Literature

Applications

In the applications of coagulation-fragmentation models to lifesciences, the clusters consist of living organisms and canchange their size not only due to the coalescence or splitting,but also due to internal demographic processes such as deathor birth of organisms inside.

Specifically, in the phytoplankton dynamics, the removal ofwhole clusters due to their sedimentation is an importantprocess that is responsible for rapid clearance of the organicmaterial from the surface of the sea.

The removal of clusters of suspended solid particles from amixture is also important in water treatment, biofuelproduction, or beer fermentation.

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 5: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

IntroductionFrom Literature

The Phytoplankton

Phytoplankton serves as the oxygen-producing foundation of many marinefood chains and humans (about 50% of the world’s oxygen).

Aggregates of phytoplankton are formed due to collisions of smalleraggregates and due to their stickiness resulting from the presence oforganic glues such as carbohydrates and Transparent ExopolymerParticles (TEP).

Jackson (1990) in modelling the dynamics of Phytoplankton,

d

dtC1 = µC1 − αC1

∞∑i=1

Ciβ1,i −C1wi

z,

d

dtCn =

α

2

n−1∑i=1

CiCn−iβi,n−i − αCn

∞∑i=1

Ciβn,i −Cnwn

z︸ ︷︷ ︸sedimentation of aggregate

, n ≥ 2.

He suggested a generalization of the above model.

Hence, the model should include fragmentation, growth and death terms

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 6: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

IntroductionFrom Literature

The full Model

The full model, [Banasiak J.], would be:

dfidt

= (gi−1fi−1 − gi fi )︸ ︷︷ ︸growth of a cell

+(di+1fi+1 − di fi )︸ ︷︷ ︸death of a cell (decay)

−si fi︸ ︷︷ ︸death of aggregate (sedimentation)

−ai fi +∞∑

j=i+1

ajbi,j fj︸ ︷︷ ︸Fragmentation of aggregate

+1

2

i−1∑j=1

ki−j,j fi−j fj −∞∑j=1

ki,j fi fj︸ ︷︷ ︸coagulation of aggregate

, i ≥ 1,

fi (0) = f 0i , i ≥ 1.

(1)

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 7: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

IntroductionFrom Literature

Assumptions

As clusters can fragment into two or more smaller pieces(multiple fragmentation) but not into bigger clusters, we notethe following:

a1 = 0, ai ≥ 0 ∈ R, i ≥ 2,

b1,j = 0,

bi ,j = 0, j ≥ i .

(2)

Naturally, for this system to conserve mass, the followingassumption has to be imposed

j−1∑i=1

ibi ,j = j , j ≥ 2. (3)

From the physical point of view, both the fragmentation termand the coagulation term should be symmetric.

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 8: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

IntroductionFrom Literature

Literature

The discrete model is not as popular in the literature as the continuous model.We decided to contribute to the discrete case.

For the continuous case, some authors (Arlotti & Banasiak, 2004; Banasiak &

Lamb, 2003 & 2009; Banasiak, 2006; Banasiak & Oukouomi, 2009; Cai et al,

1991; Edwards et al, 1990; Huang et al, 1991; Mischler & Scher, 2016; Omari,

2011; Poka, 2012; ) have worked on equation (1) with either

the decay, sedimentation and coagulation terms are notincluded, orthe growth and sedimentation terms are not included

Smith et al (2011), worked on the discrete equation of (1) but with a boundedcoagulation term and without the growth and sedimentation term using thediscrete form of the mass loss condition

j−1∑n=1

nbn,j = j(1− λj ) for j ≥ 2. (4)

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 9: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Linear PartNonlinear Part

Abstract Cauchy Problem

Formulating the equation as an Abstract Cauchy Problem (ACP):

df

dt= Ypf + Kpf , f (0) = (fn)∞n . (5)

where Yp is the linear part, and Kp is the nonlinear part

[Ypf ]i = gi−1fi−1 − (gi + di + si + ai )fi , i ≥ 1, a1 = 0,

+ di+1fi+1 +∞∑

j=i+1

ajbi,j fj , i ≥ 2. (6)

[Kpf ]i =1

2

i−1∑j=1

ki−j,j fi−j fj −∞∑j=1

ki,j fi fj , i ≥ 1 (7)

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 10: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Linear PartNonlinear Part

Growth-sedimentation-fragmentation equation

We shall use the fact that equation (6) can be written as thegrowth-sedimentation-fragmentation model

dfndt

= gn−1fn−1 − (gn + an + sn)fn +∞∑

i=n+1

aibn,i fi , n ≥ 1,

fn(0) = f inn , n ≥ 1, (8)

where an = an + dn, n ≥ 2, (with a1 = 0) and

bn,i =

{an+1bn,n+1+dn+1

an+1+dn+1, i = n + 1,

aibn,iai+di

, i ≥ n + 2.(9)

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 11: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Linear PartNonlinear Part

We note that the fragmentation part of this model no longer isconservative as

i−1∑n=1

nbn,i = i

(1− di

i(ai + di )

), i ≥ 2, (10)

so it corresponds to the model with the so-called discrete mass-loss withmass-loss fraction λn = dn/n(an + dn), ( see Cai, 1991), mathematicallyanalysed in (Smith, 2011).

The analysis of the pure fragmentation equation most often is carried outin the space X1 := `1

1 with the norm

‖f ‖[1] =∞∑n=1

n|fn| (11)

which, for a nonnegative f , gives the mass of the ensemble.

However, it is much better to consider (1) in the spaces with finite highermoments, Xp := `1

p, with the norm

‖f ‖[p] =∞∑n=1

np|fn|, p ≥ 1. (12)

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 12: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Linear PartNonlinear Part

Operators

We consider the operators (Tp,D(Tp)), (G−p ,D(Gp)),(D+

p ,D(Dp)) and (Bp,D(Bp)) defined by

[Tpf ]i = −θi fi , [G−p f ]i = gi−1fi−1,

[D+p f ]i = di+1fi+1, [Bpf ]i =

∞∑j=i+1

ajbi ,j fj , i ≥ 1,

where θi := ai + gi + di + si , i ≥ 1, and g0 = a1 = 0. Further, wedenote

∆(p)i := ip −

i−1∑j=1

jpbj ,i , i ≥ 2, p ≥ 0. (13)

Then the following holds (see Banasiak et al, 2018 for the details):

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 13: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Linear PartNonlinear Part

Generation of Analytic Semigroup

Theorem (1)

If for some p > 1

lim infi→∞

aiθi

∆(p)i

ip> 0, (14)

then for any p > 1 the sum(Yp,D(Yp)) = (Tp + Gp + Dp + Bp,D(Tp)) generates a positiveanalytic C0-semigroup {Sp(t)}t≥0 in Xp.

Proof: - the diagonal operator generates a substochastic analyticsemigroup- perturbation by [G−p f ]- perturbation by [Bpf ], and- by Arendt-Rhandi theorem, the sum generates an analyticsemigroup

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 14: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Linear PartNonlinear Part

Local Mild Solution

We assume that all the conditions of Theorem 1 are satisfied. In addition, weimpose the following bound on the coefficients of the coagulation kernel

ki,j ≤ κ((1 + θi )α + (1 + θj)

α), i , j ≥ 1, 0 < α < 1. (15)

Then, we proved the following theoremwhere the norm of the intermediate spaces, Xp,α, is given by the expression

‖f ‖p,α =∞∑i=1

ip(1 + θi )α|fi |, 0 < α < 1. (16)

Lemma (1)

Assume for some p > 1 conditions (14) and (15) are satisfied. Then for eachf0 ∈ X+

p,α and some T > 0, the initial value problem (1) has a uniquenon-negative mild solution f ∈ C([0,T ],Xp,α).

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 15: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Linear PartNonlinear Part

Local Classical Solution

Theorem (2)

Assume that conditions (14) and (15) are satisfied. Then, for eachf0 ∈ Xp,α there is T = T (f0) > 0 such that the initial valueproblem (1) has a unique non-negative classical solutionf ∈ C ([0,T ],Xp,α) ∩ C 1((0,T ),Xp) ∩ C ((0,T ),Xp,1).

Proof: - We use the variation of constant formula-We prove the map is bounded, locally lipschitz continuous and itis a contraction- We prove the differentiability of the mild solution

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 16: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Linear PartNonlinear Part

Global Solution

Lemma (2)

Assume that f0 ∈ X+p,α and for some ω1

gi − dii− si ≤ ω1 (17)

Then, under the assumptions of Theorem (2), the local solutionsatisfies

‖f ‖1 ≤ eω1t‖f0‖1, t ∈ (0,T (f0)). (18)

Theorem (3)

Under the assumptions of Theorem (2) and Lemma (2), anysolution of (1) with f0 ∈ X+

p,α, p ≥ 1, is global in time.

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 17: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Truncated EquationSimulations

Truncated Equation

In numerical simulations, we approximate the original infinite dimensionalsystem (1) by the following finite dimensional counterpart:

duidt

= gi−1ui − θiui + di+1ui+1 +N∑

j=i+1

ajbi,juj

+1

2

i−1∑j=1

ki−j,jui−juj −N∑j=1

ki,juiuj +δN,iN

N∑j=1

N∑n=N+1−j

jkn,junuj ,

ui (0) = u0,i , 1 ≤ i ≤ N.

(19)

The quadratic penalty term ensures that the discrete coagulation process isconservative – this property is important when dealing with purecoagulation-fragmentation models.

We show that if u(N) is the solution of the truncated problem (19) with the

initial condition u(N)0 , then the sequence INu

(N) approaches f as the truncationindex N increases.

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 18: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Truncated EquationSimulations

Solvability of the Truncated Equation

Theorem (4)

Assume (14), (15) and (17) hold. The truncated problem in (19)is locally solvable, i.e. for each p > 1 there exists some T > 0 suchthat for each N

u(N) ∈ C ([0,T ],Xp,α) ∩ C 1((0,T ),Xp) ∩ C ((0,T ),Xp,1), (20)

and the respective norms of u(N) are bounded independently of N.

If, in addition, the initial datum u(N)0 is non-negative, (20) holds

for any fixed T > 0. Finally, if for some q > p − 1, q ≥ 0 we have

f0 ∈ X+q+1,α and limN→∞ ‖INu(N)

0 − f0‖p,α = 0, then INu(N) → f in

C ([0,T ],Xp,α) as N →∞.

Proof: See Banasiak et al, 2018L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 19: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Truncated EquationSimulations

Fragmentation and Coagulation Kernels

In our simulations, we make use of the following two fragmentation kernels:

bi,j =2

j − 1, (21a)

bi,j =iσ(j − i)σ

αj, αj =

1

j

j−1∑i=1

i1+σ(j − i)σ , σ > −1. (21b)

And the coagulation process is driven by one of the unbounded kernels:

ki,j = k1(i1/3 + j1/3)73 , (22a)

ki,j = k2(i + k3)(j + k3), (22b)

where k1, k2 and k3 are positive constants.

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 20: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Truncated EquationSimulations

Parameters

The transport, the sedimentation and the fragmentation ratesare chosen to be

gi = giα, di = diβ, si = siγ , ai = aiδ,

for all i ≥ 1, except for d1 = a1 = 0.

In view of Theorem (1), in the calculations below it isassumed that either

max{α, β, γ} ≤ δ, p > 1, (23a)

ormax{β, δ} ≤ γ, p = 1, (23b)

The conditions ensure that the associated semigroups{Sp(t)}t≥0, equipped with either of the fragmentation kernels(21a) or (21b), are analytic in Xp, p ≥ 1.

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 21: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Truncated EquationSimulations

The pure coagulation-fragmentation

To begin, we consider (1) with g = d = s = 0, fragmentationkernel (21a) and coagulation kernel (22a). Here, the

coagulation coefficients satisfy ki ,j = O(i79 + j

79 ) hence

Threorem (3) applies, provided δ > 79 .

In our simulations, we let: N = 200, a = 1, δ = 1 andk1 = 5 · 10−3. Since N is fixed, we shorten the notationsetting uN = u. As the initial conditions, we take

un(0) = 10, 5 ≤ n ≤ 20 and un(0) = 0 otherwise

and integrate (19) in time interval [0, 1] using ode15s built-inMatlab ODE solver.

The results of the simulations are shown below.

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 22: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Truncated EquationSimulations

The pure coagulation-fragmentation contd

2040

6080

0

0.5

0

20

nt

un(t)

(a) Number of clusters

2040

6080

0

0.5

0

100

200

nt

nun(t)

(b) Distribution of cluster masses

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 23: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Truncated EquationSimulations

The pure coagulation-fragmentation contd 2

0 0.2 0.4 0.6 0.8 10

50

100

150

t

‖u‖ 0

(c) Total number of particles

0 0.2 0.4 0.6 0.8 10

500

1,000

1,500

2,000

t

‖u‖ 1

(d) The total mass

0 0.2 0.4 0.6 0.8 10

2

4

6

·104

t

‖u‖ 2

(e) Higher moments, p = 2

0 0.2 0.4 0.6 0.8 10

1

2

3

4

·106

t

‖u‖ 3

(f) Higher moments, p = 3

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 24: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Truncated EquationSimulations

The fragmentation-coagulation equation withgrowth-decay-sedimentation

We consider the complete model (1), withg = d = s = a = 1, β = γ = 0 and α = δ = 1.

The fragmentation and the coagulation processes arecontrolled respectively by the kernels (21a) and (22a), withk1 = 5 · 10−3.

The truncation index N, the time interval [0,T ] and the initialcondition u0 are chosen to be the same as in Examples 1.

The death and the sedimentation processes dominate andyield a slow decay in the moments as time increases.

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 25: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Truncated EquationSimulations

The fragmentation-coagulation equation withgrowth-decay-sedimentation contd

2040

6080

0

0.5

0

10

20

nt

un(t)

(g) Number of clusters

2040

6080

0

0.5

0

100

200

nt

nun(t)

(h) Distribution of cluster masses

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 26: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Truncated EquationSimulations

The fragmentation-coagulation equation withgrowth-decay-sedimentation contd 2

0 0.2 0.4 0.6 0.8 10

50

100

150

t

‖u‖ 0

(i) Total number of particles

0 0.2 0.4 0.6 0.8 10

500

1,000

1,500

2,000

t

‖u‖ 1

(j) The total mass

0 0.2 0.4 0.6 0.8 10

2

4

6

·104

t

‖u‖ 2

(k) Higher moments, p = 2

0 0.2 0.4 0.6 0.8 10

1

2

3

4

·106

t

‖u‖ 3

(l) Higher moments, p = 3

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 27: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

Truncated EquationSimulations

Strong Sedimentation case

2040

6080

0

0.5

0

10

20

nt

un(t)

(m) Number of clusters

2040

6080

0

0.5

0

100

200

nt

nun(t)

(n) Distribution of cluster masses

0 0.2 0.4 0.6 0.8 10

50

100

150

t

‖u‖ 0

(o) Total number of particles

0 0.2 0.4 0.6 0.8 10

500

1,000

1,500

2,000

t

‖u‖ 1

(p) The total mass

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 28: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

ReferencesThank you

References

[1] Banasiak J., Joel L. O. and Shindin S. (2018) The Discrete Unbounded Coagulation-Fragmentation Equationwith Growth, Decay and Sedimentation. ArXiv eprint, arXiv:1809.00046, August 2018.

[2] Banasiak J., Joel L.O. and Shindin S. (2017) Analysis and simulations of the discrete fragmentation equationwith decay, Mathematical Methods in the Applied Sciences

[3] Banasiak J. (2001), On an extension of the Kato-Voigt perturbation theorem for substochastic semigroups andits application, Taiwanese Journal of Mathematics, vol 5.

[4] Banasiak J., and Lamb W. (2003), On the application of substochastic semigroup theory to fragmentationmodels with mass loss, Journal of mathematical analysis and applications, vol 284, pages 9–30.

[5] Banasiak J., Arlotti L. (2006), Perturbations of positive semigroups with applications, Springer Science &Business Media

[6] Cai M., Edwards B. F. and Han H. (1991), Exact and asymptotic scaling solutions for fragmentation with massloss, Physical Review A, vol 43, pages 656 - 662.

[7] Jackson, G. A. (1990). A model of the formation of marine algal flocs by physical coagulation processes. DeepSea Research Part A. Oceanographic Research Papers, 37(8):1197 1211.

[8] Smoluchowski M., Drei vortrage uber diffusion, brownsche bewegung und koagulation von kolloidteilchen.Zeitschrift fur Physik, 17:557-585, 1916.

[9] Pazy, A. (1983), Semigroups of linear operators and applications to partial differential equations, Appliedmathematical sciences, Springer, New York, Berlin.

[10] Smith A. L., (2011), Mathematical analysis of discrete coagulation-fragmentation equations, PhD Theses,University of Strathclyde.

[11] Voigt J. (1987), On substochastic C0-semigroups and their generators, Transport Theory and StatisticalPhysics, vol 16, pages 453–466, Taylor & Francis.

[12] Ziff R. M. and McGrady E. D. (1985), The kinetics of cluster fragmentation and depolymerisation, Journal ofPhysics A: Mathematical and General, vol 18.

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation

Page 29: The Discrete Unbounded Coagulation-Fragmentation Equation ... · coagulation{fragmentation models with growth, decay and sedimentation. We demonstrate the existence and uniqueness

General IntroductionAnalysis

Numerical SimulationsReferences

ReferencesThank you

Thank you

Thank you!!!

L. O. Joel, J. Banasiak and S. Shindin The Discrete Coagulation-Fragmentation Equation