This 2D cross-section resolves the fracture along its own length instead of treating it as one lumped aperture. The conduit is split into 40 cells; water flowing under a fixed head drop must carry the same discharge q through every cell in series (mass conservation), so a narrow cell anywhere along the path dominates the total resistance — exactly like resistors in series:
q = ρ·g·i·L / (12·μ·Σᵢ(Δxᵢ / wᵢ³)) — series cubic-law flow
v(x) = q / w(x) — local velocity, varies along x
Saturation still marches along the conduit with the real Plummer–Wigley–Parkhurst kinetic trigger (linear far from equilibrium, collapsing to a steep 4th-order law above 90% saturation), but each cell now sees its own local velocity, so the near-inlet water — always the least saturated — dissolves rock fastest right there:
dw(x)/dt = 2·k₁·f(c(x))·M_CaCO3 / ρ_rock — local wall recession, both walls
The result is a genuine 2D effect a single-aperture model can't show: dissolution concentrates near the inlet, so that end widens dramatically faster than the outlet — a real "funnel" shape opens up. Once the inlet segment is wide enough that it no longer limits the flow, the bottleneck shifts to the next cell downstream, and a widening front sweeps from inlet to outlet like a wave, rather than the whole conduit thickening uniformly at once — the geometry of how real karst breakthrough fronts actually propagate (Groves & Howard 1994; Dreybrodt 1996).
- Initial aperture — smaller starting cracks take far longer to break through, and the inlet/outlet asymmetry is more extreme.
- CO₂ aggressiveness — raises the equilibrium Ca²⁺ ceiling, letting water carry away more rock before saturating.
- Hydraulic gradient — the head drop per metre driving flow through the whole series of cells.
Numerically the model self-verified against the original lumped 3D formulation: at matched initial conditions the length-averaged local recession rate and outlet saturation agree with the lumped model to within floating-point precision, and the integral of local rock removal along the conduit matches the solute mass actually carried out at the outlet — the two are the same physics, resolved at different spatial detail.