Definition
A coupling technique that enforces compatibility between models or discretizations by projecting variables, fluxes, or traces onto a shared representation or a compatible discrete basis.
Principle
Principle
Formulate projection operators (e.g., L2, H1, mortar-type) that impose weak or strong continuity by minimizing projection error in a chosen norm and matching conserved interface quantities, possibly using variational formulations.
Demonstration
Demonstration
Illustrative example: projecting fluid pressure and normal flux onto interface basis functions so a Navier–Stokes solver and a porous-media solver exchange consistent interface data through a variational projection.
Misapplication
Misapplication
Projecting only pointwise nodal values without accounting for the mismatch of function spaces or quadrature support, which can introduce instabilities, loss of conservation, or aliasing errors.
Consequence
Consequence
Projection coupling yields variationally consistent interfaces, can reduce spurious reflections and numerical mismatch, and often leads to sparse, well-structured operators; stability and accuracy depend on projection choice and basis compatibility.
Reversal
Reversal
The reverse concept is a lifting or reconstruction without projection that injects interface coefficients into local fields; such lifts can amplify unresolved components and expose information loss of prior projection.
Boundary
Boundary
Applicable when a shared representation or compatible discrete bases can be defined; excludes purely heuristic averaging or ad hoc interpolation that lack projection properties or variational grounding.
Semantic Tension
Semantic Tension
Tension exists with mortar methods, Nitsche-type couplings, and direct interpolation: projection coupling stresses a variational projection operator, whereas alternatives use Lagrange multipliers, penalties, or strong matching.
Synthesis
Synthesis
Projection coupling is the practice of enforcing model compatibility by mapping interface data into a common variational subspace via projection operators, trading off error minimization, conservation, and computational structure; typical use-cases include fluid–structure interfaces and multi-mesh coupling.