1: \begin{abstract}

2: A local convergence rate is established for

3: a Gauss orthogonal collocation method applied to optimal control

4: problems with control constraints.

5: If the Hamiltonian possesses a strong convexity property, then

6: the theory yields convergence for problems whose optimal state and

7: costate possess two square integrable derivatives.

8: The convergence theory is based on a stability result for the sup-norm

9: change in the solution of a variational inequality relative to a 2-norm

10: perturbation,

11: and on a Sobolev space bound for the error in interpolation at the

12: Gauss quadrature points and the additional point $-1$.

13: The tightness of the convergence theory is examined using a numerical example.

14: \end{abstract}

15: