The study reports entropy-conserving or entropy-stable behavior for compressible-flow calculations with moving walls and deforming meshes, but only under the assumptions built into its formulation. At the continuous level, matching the fluid and wall velocities makes the inviscid entropy flux vanish, while the viscous treatment is entropy-conservative for an adiabatic wall and entropy-stable for a prescribed bounded heat-entropy flow.
The analysis was extended to a spatial semi-discretization, which discretizes space while leaving time continuous, and reports nonlinear stability in the L2 sense for the continuous and semi-discrete equations. The result does not provide a general guarantee for every fully time-stepped calculation.
Tests on moving and deforming flows
The work studies the compressible Euler and Navier-Stokes equations on general moving or deforming ALE boundaries. Its semi-discrete framework uses diagonal-norm summation-by-parts operators and flux differencing, with nodal discontinuous Galerkin formulations identified as an alternative setting.
The numerical evaluation covered an isentropic vortex, rotating-pipe flow, a heaving and pitching airfoil, a three-dimensional turbulent airfoil in a two-degree-of-freedom system, and a supersonic bluff body in unsteady rotation. The evidence came from analytical derivations, symbolic checks, and numerical test cases rather than a participant sample.
When the discrete temporal geometric conservation law and spatial metric identities hold, entropy contraction reduces the discrete ALE volume contribution to surface terms. Under the stated viscous-matrix and penalty assumptions, an additional viscous wall penalty is reported as entropy-dissipative even for an arbitrarily moving wall.
With entropy-stable fluxes, the isentropic-vortex L2 error followed the expected p + 1 convergence trend, where p is the polynomial degree. At p = 5, the error plateaued very close to machine precision and the time-integration tolerance. In the rotating-pipe study, computed accuracy was close to the formal p + 1 order for both static and moving-grid comparisons.
Demanding numerical cases
The heaving-and-pitching airfoil calculation agreed with reference computations on aerodynamic loads and resolved flow structures. In a separate coupled-airfoil stress test, the motion reached approximately -60 degrees to 60 degrees in pitch and -0.2 to 0.2 in nondimensional heave, while a nonlinear hardening spring kept rotation bounded. The authors describe this as a numerical stress test, not a quantitative physical flutter benchmark.
A supersonic bluff-body calculation remained stable through prescribed three-dimensional rigid-body motion while accommodating strong compressibility, a bow shock, and an unsteady wake. The supplied analysis does not report quantitative runtime or parallel-scaling measures.
The conditions behind the result
The entropy proof concerns the spatial semi-discretization. The study notes that a time integrator can still generate entropy even when the spatial scheme is entropy-conservative; its convexity argument also requires positive density and temperature, generally with an additional positivity-preserving mechanism. The wall construction assumes that wall velocity coincides with mesh velocity at the wall.
These are mathematical and computational tests, not human, animal, or experimental evidence. The two-degree-of-freedom case is not a quantitative experimental flutter benchmark, and the supplied material does not provide complete tabulated error, force, runtime, or uncertainty data for every test. Extensions claimed for finite-volume, finite-element, and flux-reconstruction methods were not separately evaluated in the supplied tests.
The manuscript is an arXiv preprint, version 1, dated 25 Aug 2026. Its conclusions remain conditional on positivity, geometric-conservation identities, and the stated moving-wall assumptions.
Paper data and sources
Original title: Entropy-stable moving-wall boundary conditions for the ALE formulation of the compressible Navier-Stokes equations
Authors: Luca Galimberti, Roberto Nuca, Lisandro Dalcin et al.
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-25
DOI: Not available
Original paper · Full text