1: \begin{abstract}

2: We propose, analyze mathematically, and study numerically a novel approach for the 

3: finite element approximation of the spectrum of second-order elliptic operators. 

4: {The main idea is to reduce the stiffness of the problem by subtracting a least-squares penalty on the gradient jumps across the mesh interfaces

5: from the standard stiffness bilinear form.}

6: %

7: This penalty bilinear form is similar to the known technique

8: used to stabilize finite element approximations in various contexts{. 

9: The penalty term is designed to dampen the high frequencies in the spectrum and so it is weighted here by a negative coefficient.

10: %

11: T}he resulting approximation technique is called softFEM {since it reduces the stiffness of the problem}. The two key advantages of

12: softFEM over the standard Galerkin FEM 

13: are to improve the approximation of the eigenvalues in the upper part of

14: the discrete spectrum and to reduce the condition number of the stiffness matrix.

15: We derive a sharp upper bound

16: on the softness parameter weighting the stabilization bilinear form so as to maintain

17: coercivity for the softFEM bilinear form. Then we prove that softFEM delivers the

18: same optimal convergence rates as the standard Galerkin FEM approximation for the 

19: eigenvalues and the eigenvectors. We next compare the discrete eigenvalues obtained 

20: when using Galerkin FEM and softFEM. Finally, a detailed analysis of linear softFEM

21: for the 1D Laplace eigenvalue problem delivers a sensible choice for the softness

22: parameter. With this choice, the stiffness reduction ratio scales linearly with the

23: polynomial degree. Various numerical experiments illustrate the benefits of using softFEM

24: over Galerkin FEM.

25:  \textbf{Mathematics Subjects Classification}: 65N15, 65N30, 65N35, 35J05

26: \end{abstract}

27: