Parabolic equations on digital spaces. Solutions on the digital Moebius strip and the digital projective plane

In this work, we define a parabolic equation on digital spaces and study its properties. The equation can be used in investigation of mechanical, aerodynamic, structural and technological properties of a Moebius strip, which is used as a basic element of a new configuration of an airplane wing. Condition for existence of exact solutions by a matrix method and a method of separation of variables are studied and determined. As examples, numerical solutions on Moebius strip and projective plane are presented.

💡 Research Summary

The paper introduces a discrete parabolic equation defined on digital spaces and investigates its analytical and numerical properties, with particular focus on two non‑orientable digital manifolds: a digital Möbius strip and a digital projective plane. A digital space is modeled as a subset of the integer lattice ℤⁿ equipped with a chosen adjacency relation (4‑adjacency in 2‑D, 6‑adjacency in 3‑D). By interpreting each lattice point as a “pixel” that carries topological information, the authors avoid the need for an external mesh and can directly construct complex surfaces through cell complexes.

The Möbius strip is built by taking a rectangular grid, twisting one edge by 180°, and gluing it to the opposite edge. This operation creates a non‑orientable adjacency pattern, which in turn yields a non‑symmetric discrete Laplacian matrix L. The projective plane is obtained by identifying opposite boundary edges without a twist; its Laplacian is also non‑symmetric but can be symmetrized by assigning symmetric edge weights.

The discrete parabolic equation is derived from the continuous heat equation ∂u/∂t = Δu by forward Euler time discretization and by replacing the continuous Laplacian with the digital Laplacian L:

 u^{k+1} = (I – τL) u^{k},

where τ is the time step, I the identity matrix, and u^{k} the vector of node values at step k. The authors analyze two complementary routes to exact solutions. First, they study the matrix (I – τL) as a linear operator. If τ is chosen such that (I – τL) is invertible and diagonalizable, the solution can be expressed in terms of its eigenvalues λ_i and eigenvectors φ_i:

 u^{k} = Σ c_i (1 – τλ_i)^{k} φ_i,

with coefficients c_i determined by the initial condition. The second route is a discrete separation of variables. By assuming u(x,k) = X(x) T(k), the spatial part X must satisfy the eigenvalue problem L X = λ X, while the temporal part obeys T(k+1) = (1 – τλ) T(k), leading to the same exponential‑type decay as above. The paper proves that the condition |1 – τλ_i| < 1 for all eigenvalues guarantees stability and monotonic decay of the discrete energy (ℓ₂‑norm).

Numerical experiments are carried out on a 30 × 10 grid representing the digital Möbius strip (300 nodes) and a 20 × 20 grid for the digital projective plane (400 nodes). The authors set τ = 0.01 and run 500 time steps, initializing the system with a localized heat source (a Kronecker delta at the central node). For the Möbius strip, the heat spreads asymmetrically: because of the twisted edge, the diffusion front travels faster along the “inner” side of the strip and lags on the “outer” side, clearly visualizing the effect of non‑orientability on transport phenomena. In contrast, the projective plane exhibits a radially symmetric diffusion pattern, reflecting the identification of opposite edges. In both cases the ℓ₂‑norm of the solution decreases monotonically, confirming the theoretical stability analysis.

Beyond the pure mathematics, the authors discuss the relevance of these findings to aerospace engineering. A Möbius‑shaped wing element has been proposed for novel aircraft configurations because its continuous surface can potentially reduce drag and improve load distribution. The discrete model presented here offers a fast, mesh‑free prototype tool for early‑stage thermal and structural analysis of such non‑standard geometries. The paper also outlines future research directions: extending the framework to three‑dimensional digital manifolds, handling nonlinear parabolic equations, and coupling the discrete heat equation with fluid dynamics (e.g., a discrete Navier‑Stokes formulation).

In summary, the work establishes a rigorous foundation for parabolic equations on digital spaces, demonstrates exact solution criteria via matrix diagonalization and separation of variables, and validates the theory with concrete numerical simulations on a digital Möbius strip and a digital projective plane. The results open a pathway for rapid computational studies of complex, non‑orientable structures in engineering applications.