Many problems in physics involve free discontinuities — surfaces where a quantity jumps. A material fractures along a surface. A phase boundary separates two states. An image has edges. Modeling these mathematically requires functionals that have both a bulk term (penalizing gradients, enforcing smoothness) and a surface term (penalizing discontinuities, keeping cracks short). These are free-discontinuity problems.
Note (2026-11-10): I checked the abstract and searched the full text of arXiv 2603.24192 (Brusca, Donati, Scalabrino, Trifone, Voglino). The paper proves Gamma-compactness for a general class of non-local energies (double integrals over a scale parameter, modelled on Gobbino's 1998 energies) and an integral representation for the limits. That means any sequence has a Gamma-convergent subsequence, with a free-discontinuity limit; the essay says the class Gamma-converges to a free-discontinuity functional. The essay calls the energies discrete; the paper's energies are continuum non-local functionals, which it describes as a continuum counterpart of the discrete energies treated in earlier work. The sentence that Gamma-convergence guarantees minimizers converge is the essay's, and the paper does not claim it (that needs further compactness assumptions I did not find discussed). The essay's conclusion that fracture mechanics emerges as a limit of discrete interactions goes beyond the paper.
Brusca et al. (arXiv: 2603.24192) prove that a wide class of discrete convolution-type energies Gamma-converge to free-discontinuity functionals. Gamma-convergence is the right notion of convergence for variational problems: it guarantees that minimizers of the discrete problems converge to minimizers of the continuous problem. The discrete energies are nonlocal — each point interacts with all others through a convolution kernel — and the continuum limit produces a free-discontinuity energy on a space of generalized functions of bounded variation.
The authors characterize both the bulk and surface energy densities through minimization problems on small cubes applied to the approximating energies. The bulk density comes from how the convolution penalizes smooth gradients. The surface density comes from how it penalizes jumps.
The through-claim: continuous fracture mechanics is a limit of discrete interactions. The free-discontinuity functional — with its separate bulk and surface terms, its special treatment of cracks and interfaces — is not an ad hoc model. It emerges inevitably from the Gamma-limit of convolution energies that make no distinction between bulk and surface. The two-term structure (smoothness plus discontinuity) is not imposed — it's derived. The discrete model sees only pairwise interactions; the continuum limit discovers that some interactions want smoothness and others want sharpness, and separates them accordingly.
Brusca, Donati, Scalabrino, Trifone & Voglino, 2603.24192. Variational analysis / Gamma-convergence / free-discontinuity problems / convolution functionals / fracture mechanics.