embedded3 A3-i: the advancing-wall friction closure (RTX_E3_WALL_ADVANCING=1, host + device, per-face normal-velocity table; f(0) = 1 exactly) — REFUTED as built on the slab ladder (flag shear 34.2 / 33.6 / 31.1 vs 28.9 / 32.3 / 30.1: spread 3.4 -> 3.1, now from above; worst mass residual 1e-8 -> 1e-4); knob stays off, default digit-identical

Co-Authored-By: Claude Fable 5.1 <[email protected]>

crates/specialized/rtx-cfd/src/kernels/cuda/e3_cut.cu

@@ -29,8 +29,15 @@ struct E3Cut {
     const unsigned int *fold_ptr, *fold_idx; /* CSR: the slaves of every cell */
     double *cell_flux;                  /* scratch per cell */
     const double *s_u, *s_v, *s_w;      /* open-part centroid shifts per face, 3 interleaved (S2-5) */
+    const double *vn_u, *vn_v, *vn_w;   /* the wall's normal velocity into the fluid per face (A3-i) */
 };
 
+/* f(xi) = xi / (1 - exp(-xi)), f(0) = 1 exactly (closure.rs advancing_factor). */
+__device__ __forceinline__ double adv_factor(double xi)
+{
+    return fabs(xi) < 1e-8 ? 1.0 + 0.5 * xi : xi / (-expm1(-xi));
+}
+
 /* Face index of component c at lattice (i, j, k); −1 outside (z wraps when periodic). */
 __device__ __forceinline__ int cut_face(const E3Params& g, int c, int i, int j, int k)
 {
@@ -311,6 +318,11 @@ __device__ double cut_face_update(const E3Params& g, const E3Ptrs& f, const E3Cu
        open face away from the body along the wall normal's dominant axis
        (order 2; the neighbour's old value explicit). */
     double c1 = 1.0 / distance, shear_explicit = 0.0;
+    /* bit 9 of wall_order: the advancing-wall closure on the one-point coefficient (A3-i) */
+    if (g.wall_order & 512) {
+        const double* vnt = c == 0 ? m.vn_u : (c == 1 ? m.vn_v : m.vn_w);
+        c1 = adv_factor(vnt[fidx] * distance / g.nu) / distance;
+    }
     if ((g.wall_order & 15) >= 2 && a_w > 0.0) {
         double nw[3] = { wall[0] / a_w, wall[1] / a_w, wall[2] / a_w };
         int d = 0;