1: \begin{abstract}

2: 	In this paper we discuss the use of implicit Runge-Kutta schemes for the time discretization 

3: 	of optimal control problems with evolution equations.

4: 	The specialty of the considered discretizations is  that 

5: 	the discretizations schemes for the state and adjoint state are chosen 

6: 	such that discretization and optimization commute. 

7: 	It is well known that for Runge-Kutta schemes with this property additional order conditions are necessary.

8: 	We give sufficient conditions for which class of schemes these additional order condition are automatically fulfilled.

9: 	The focus is especially on implicit Runge-Kutta schemes of

10: 	Gauss, Radau IA, Radau IIA, Lobatto IIIA, Lobatto IIIB and Lobatto IIIC collocation type up to order six.

11: 	Furthermore we also use a SDIRK (singly diagonally implicit Runge-Kutta) method to demonstrate, that for general implicit Runge-Kutta methods the additional order conditions are not automatically fulfilled.

12: 

13: 	Numerical examples illustrate the predicted convergence rates.

14: 

15: 	\ 

16: 

17: 	\textbf{Mathematical Subject Classification (2010)}

18:         49M25, % OC             -- Num meth             - discrete approx

19:         49M05, % OC             -- Num meth             - based on necessary conditions

20:         65M15, % Num Ana        -- PDE, IVP + IBVP      - error bounds

21:         65M60, % Num Ana        -- PDE, IVP + IBVP      - FE...Galerkin Methods

22:         49M29. % OC             -- Num meth             - involving duality

23: 

24: 

25: 	\textbf{Keywords:}

26:         Optimal control problem,

27:         Parabolic partial differential equation,

28: 	Implicit Runge-Kutta schemes.

29: 

30: \end{abstract}

31: