Stationary solutions of a free boundary problem modeling the growth of tumors with Gibbs-Thomson relation (Q2634206): Difference between revisions

In this paper, the following free boundary problem in the unknowns \((\sigma,p,\Omega)\) is studied: \[ \begin{aligned}\Delta\sigma=\lambda\sigma,\quad \Delta p=-\mu(\sigma-\tilde\sigma) &\quad \text{in }\Omega,\\ \sigma=\bar\sigma(1-\gamma\kappa),\quad p=\bar p,\quad \partial_n p=0&\quad\text{on }\partial\Omega.\end{aligned} \] This system can be viewed as a stationary model for tumour growth, where \(\sigma\) denotes the nutrient concentation and \(p\) denotes the pressure inside the tumour region \(\Omega\subset\mathbb{R}^3\). Apart from the mean curvature \(\kappa\) and the normal vector field \(n\) which depend on \(\Omega\), all other variables above are positive constants: in this model, the central role is played by the cell-to-cell adhesiveness \(\gamma\) since it acts as a bifurcation parameter. The main results of this paper are as follows: There exists a constant \(\theta_\ast\in (0,1)\) which can be computed explicity and depends on \(\gamma\) only, such that, if \(0<\frac{\tilde\sigma}{\bar\sigma}<\theta_\ast\), there exist two radial solutions of the system above (meaning that \(\Omega\) is a ball with center \(0\) and \(\sigma\), \(p\) are radially symmetric functions). Conversely, if \(\frac{\tilde\sigma}{\bar\sigma}>\theta_\ast\), radial solutions do not exist; if \(\frac{\tilde\sigma}{\bar\sigma}=\theta_\ast\), there is a unique radial solution. In the first case where two radial solutions exist, infinitely many branches of non-radial solutions exist bifurcating from each radial solution, where \((\gamma_k)\) is a null sequence of bifurcation points \(\gamma=\gamma_k\).