A parabolic system modeling microbial competition in an unmixed bio-reactor (Q1924448)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | A parabolic system modeling microbial competition in an unmixed bio-reactor |
scientific article |
Statements
A parabolic system modeling microbial competition in an unmixed bio-reactor (English)
0 references
10 July 1997
0 references
The authors consider a system of reaction diffusion equations of the following type: \[ S_t=d_0\Delta S-\sum^m_1\gamma^{-1}_iu_if_i(S),\quad x\in\Omega,\quad t>0,\tag{1} \] \[ (u_i)_t=d_i\Delta u_i+u_i(f_i(S)-k_i)\quad\text{for }i=1,\dots,m. \] There are inhomogeneous boundary conditions: \[ \partial_nS(t,x)+r_0(x)S(t,x)=S^0(x),\quad\partial_nu_i(t,x)=0,\quad x\in\partial\Omega,\quad t>0.\tag{2} \] Here \(\Omega\subseteq \mathbb{R}^N\) is a bounded smooth domain and \(\partial_nS(t,x)\) the derivative at \(x\in\partial\Omega\) with respect to the outer unit normal \(n(x)\) at \(x\in\partial\Omega\). Finally, nonnegative initial conditions are given: (3) \(S(0,x)=S_0(x)\), \(u_i(0,x)=u_{i0}(x)\), \(x\in\Omega\). The functions \(r_i(x)\), \(f_i(S)\), \(f(S)\), \(S^0(x)\) are subject to positivity and smoothness conditions; the numbers \(d_i\), \(\gamma_i\), \(k_i\) are positive. As to \(f\) in particular it is assumed that \(f: \mathbb{R}_+\to \mathbb{R}_+\), \(f(0)=0\) and \(f\in C^1(\mathbb{R}_+)\). Examples of particular importance are \(f(S)=mS(a+S)^{-1}\) (the Monod function) and \(f(S)=mS(a+S+K^{-1}S^2)^{-1}\) (non-monotone Andrews function). At some places it is assumed that \(f'(S)>0\) on some interval \([0,L]\) for suitable \(L\). The system (1), (2), (3) has its roots in population dynamics. The \(u_i\), \(i=1,\dots,m\) represent concentrations of competing species while \(S\) is the concentration of a nutrient. In the introduction, the authors explain at length the biological background of (1), (2), (3). Then they consider (1), (2), (3) as a semidynamical system on \(X_+=C^{m+1}_+\), where \(C_+\) is the cone of nonnegative functions in \(C(\overline\Omega)\), and where \(X=C(\overline\Omega)^{m+1}\) is endowed with the usual supremum norm. About this dynamical system, they prove a series of results. A principal one is the following Theorem: Equations (1), (2), (3) generate a semidynamical system \(T(t)\), \(t\geq 0\) on \(X_+\) which is dissipative and possesses a global attractor \(A\) having finite Hausdorff dimension; if \((S,u_1,\dots,u_m)\in A\) then \(0\leq S\leq S_*\). Here, \(S_*\) is the solution of the boundary value problem: \(\Delta S_*=0\), \(x\in\Omega\), \(\partial_nS_*+r_0(x) S_*=S^0(x)\), \(x\in\partial\Omega\). The proof requires quite a series of preparatory estimates, too involved to be discussed here. The proof provides actually more information than stated in the theorem and gives rise to further results. The paper concludes with a discussion of equilibria of (1), (2) and of their properties.
0 references
Monod function
0 references
Andrews function
0 references
semidynamical system
0 references
global attractor
0 references
Hausdorff dimension
0 references