Global classical solutions for a free-boundary problem modeling combustion of solid propellants (Q2488792)

In its rescaled formulation the problem here studied has the form \[ \begin{alignedat}{2} U_t&=U_{yy},\quad & y&<s(t),\;t<0,\\ U_y(-\infty, t)&=0,\quad &t&>0,\\ U_y(s(t),t)&=Q(U(s(y),t)),\quad &t&>0,\\ U(y,0)&=U_{in}(y),\quad &y&\leq0,\\ -\dot{s}(t)&=R(U(s(t),t)),\quad &t&>0,\\ s(0)&=0\end{alignedat} \] where \(R\), \(Q\) are nonnegative functions, continuous and bounded with their derivatives up to the second order, and \(U_{in}\) is continuous, bounded and vanishing to \(-\infty\). Existence and uniqueness for sufficiently small time is proved by means of classical arguments. Next, supposing that \(U_{in}\) has an exponential decay to \(-\infty\), the solution is shown to exist for all times. Finally, under some additional assumptions on \(Q\), \(R\), both \(U\) and \(U_y\) multiplied by \(\exp[\eta(s(t)-y)]\), for a suitable positive constant \(\eta\), are proved to remain bounded. This suggests that the travelling wave front associated to the original free boundary problem is stable.