1: \begin{abstract}

2: Gradient schemes is a framework which enables the unified convergence analysis of many different methods -- such as finite elements (conforming, non-conforming and mixed) and finite volumes methods -- for $2^{\rm nd}$ order diffusion equations. We show in this work that the gradient schemes framework can be extended to variational inequalities involving mixed Dirichlet, Neumann and Signorini boundary conditions. 

3: This extension allows us to provide error estimates for numerical approximations of such models,

4: recovering known convergence rates for some methods, and establishing new convergence rates for

5: schemes not previously studied for variational inequalities. The general framework we develop

6: also enables us to design a new numerical method for the obstacle and Signorini problems,

7: based on hybrid mimetic mixed schemes. We provide

8: numerical results that demonstrate the accuracy of these schemes, and confirm our theoretical rates

9: of convergence.

10: \end{abstract}

11: