1: \begin{abstract}

2: Virtual element methods is a new promising finite element methods using general polygonal meshes.  Its optimal {\it a priori} error estimates are

3: well established in the literature.   In this paper, we take a different viewpoint.   We try to  uncover the superconvergent

4: property of the virtual element methods by doing some local  post-processing only  on the degrees of freedom.   Using linear virtual element method

5: as an example, we propose a universal recovery procedure to improve the accuracy of gradient approximation for numerical methods using general

6: polygonal meshes.  Its capability of serving  as {\it a posteriori } error estimators in adaptive methods is also investigated.  Compared to the existing

7: residual-type {\it a posteriori} error estimators for the  virtual element methods, the recovery-type {\it a posteriori} error estimator based on the proposed

8: gradient recovery technique is much  simpler in implementation and asymptotically exact.   A series of  benchmark tests are presented to numerically

9: illustrate the superconvergence of recovered gradient and validate the asymptotical exactness of the recovery-based {\it a posteriori } error estimator.

10: \vskip .3cm

11: {\bf AMS subject classifications.} \ {Primary 65N30, 65N12; Secondary 65N15, 53C99}

12: \vskip .3cm

13: 

14: {\bf Key words.} \ {Gradient recovery, superconvergence, polynomial preserving, virtual element method, recovery-based, a posteriori error estimator, polygonal mesh}

15: \end{abstract}

16: