1: \begin{abstract}

2: \begin{center}

3: \begin{minipage}{0.9\textwidth}

4:   {\small

5:   \textbf{Abstract.} 

6:   In this paper we introduce a new, simple and efficient numerical scheme for the

7:   implementation of the freezing method for capturing similarity solutions in

8:   partial differential equations.

9:   The scheme is based on an IMEX-Runge-Kutta approach for a method of lines

10:   (semi-)discretization of the freezing partial differential

11:   algebraic equation (PDAE).  We prove second order convergence for the time

12:   discretization at smooth solutions in the ODE-sense and we present

13:   numerical experiments that show second order convergence for the full

14:   discretization of the PDAE.

15:   

16:   As an example serves the multi-dimensional Burgers' equation.  By considering

17:   very different sizes of viscosity, Burgers' equation can be considered as a

18:   prototypical example of general coupled hyperbolic-parabolic PDEs.  

19:   Numerical experiments show that our method works perfectly well for all sizes

20:   of viscosity, suggesting that the scheme is indeed suitable for capturing

21:   similarity solutions in general hyperbolic-parabolic PDEs by direct

22:   forward simulation with the freezing method.

23:   }

24: \end{minipage}

25: \end{center}

26: %\end{abstract}

27: