07 July 2014 09:03:08 PM FEM2D_BVP_SERENE_PRB C++ version Test the FEM2D_BVP_SERENE library. TEST01 Solve - del ( A del U ) + C U = F on the unit square with zero boundary conditions. A1(X,Y) = 1.0 C1(X,Y) = 0.0 F1(X,Y) = 2*X*(1-X)+2*Y*(1-Y) U1(X,Y) = X * ( 1 - X ) * Y * ( 1 - Y ) Number of X grid values NX = 5 Number of Y grid values NY = 5 I J X Y U Uexact Error 0 0 0 0 0 0 0 1 0 0.25 0 0 0 0 2 0 0.5 0 0 0 0 3 0 0.75 0 0 0 0 4 0 1 0 0 0 0 0 1 0 0.25 0 0 0 2 1 0.5 0.25 0.0486306 0.046875 0.00175562 4 1 1 0.25 0 0 0 0 2 0 0.5 0 0 0 1 2 0.25 0.5 0.0486306 0.046875 0.00175562 2 2 0.5 0.5 0.0589888 0.0625 0.00351124 3 2 0.75 0.5 0.0486306 0.046875 0.00175562 4 2 1 0.5 0 0 0 0 3 0 0.75 0 0 0 2 3 0.5 0.75 0.0486306 0.046875 0.00175562 4 3 1 0.75 0 0 0 0 4 0 1 0 0 0 1 4 0.25 1 0 0 0 2 4 0.5 1 0 0 0 3 4 0.75 1 0 0 0 4 4 1 1 0 0 0 l1 norm of error = 0.000501605 L2 norm of error = 0.000805291 Seminorm of error = 0.0123351 TEST02 Basis function checks. The matrix Aij = V(j)(X(i),Y(i)) should be the identity. 1 -0 0 0 -0 0 0 -0 0 1 0 -0 0 -0 0 -0 -0 0 1 -0 0 -0 0 -0 -0 0 0 1 -0 0 -0 0 -0 0 -0 0 1 0 -0 0 -0 0 -0 0 0 1 0 0 -0 -0 0 -0 0 -0 1 0 0 -0 0 -0 0 -0 0 1 The vectors dVdX(1:8)(X,Y) and dVdY(1:8)(X,Y) should both sum to zero for any (X,Y). Random evaluation point is (0.436837,4.91264) dVdX dVdY 0 -0.102179 0.137834 1 1.07713 0.341423 2 -0.974948 0.933378 3 -0.0835485 -1.4266 4 -0.00466728 0.49322 5 0.0492007 -0.341423 6 -0.0445334 0.260839 7 0.0835485 -0.398672 Sum: -9.71445e-17 1.11022e-16 TEST03 Solve - del ( A del U ) + C U = F on the unit square with zero boundary conditions. A1(X,Y) = 0.0 C1(X,Y) = 1.0 F1(X,Y) = X * ( 1 - X ) * Y * ( 1 - Y ) U1(X,Y) = X * ( 1 - X ) * Y * ( 1 - Y ) This example is contrived so that the system matrix is the WATHEN matrix. Number of X grid values NX = 5 Number of Y grid values NY = 5 Wathen elementary mass matrix: Col: 0 1 2 3 4 Row 0: 6 -6 2 -8 3 1: -6 32 -6 20 -8 2: 2 -6 6 -6 2 3: -8 20 -6 32 -6 4: 3 -8 2 -6 6 5: -8 16 -8 20 -6 6: 2 -8 3 -8 2 7: -6 20 -8 16 -8 Col: 5 6 7 Row 0: -8 2 -6 1: 16 -8 20 2: -8 3 -8 3: 20 -8 16 4: -6 2 -8 5: 32 -6 20 6: -6 6 -6 7: 20 -6 32 I J X Y U Uexact Error 0 0 0 0 0 0 0 1 0 0.25 0 0 0 0 2 0 0.5 0 0 0 0 3 0 0.75 0 0 0 0 4 0 1 0 0 0 0 0 1 0 0.25 0 0 0 2 1 0.5 0.25 0.0488281 0.046875 0.00195313 4 1 1 0.25 0 0 0 0 2 0 0.5 0 0 0 1 2 0.25 0.5 0.0488281 0.046875 0.00195312 2 2 0.5 0.5 0.0585938 0.0625 0.00390625 3 2 0.75 0.5 0.0488281 0.046875 0.00195312 4 2 1 0.5 0 0 0 0 3 0 0.75 0 0 0 2 3 0.5 0.75 0.0488281 0.046875 0.00195313 4 3 1 0.75 0 0 0 0 4 0 1 0 0 0 1 4 0.25 1 0 0 0 2 4 0.5 1 0 0 0 3 4 0.75 1 0 0 0 4 4 1 1 0 0 0 l1 norm of error = 0.000558036 L2 norm of error = 0.00078125 Seminorm of error = 0.0124347 Wathen matrix: Col: 0 1 2 3 4 Row 0: 6 -6 2 -6 -8 1: -6 32 -6 20 20 2: 2 -6 6 -8 -6 3: -6 20 -8 32 16 4: -8 20 -6 16 32 5: 2 -8 3 -6 -8 6: -8 16 -8 20 20 7: 3 -8 2 -8 -6 Col: 5 6 7 Row 0: 2 -8 3 1: -8 16 -8 2: 3 -8 2 3: -6 20 -8 4: -8 20 -6 5: 6 -6 2 6: -6 32 -6 7: 2 -6 6 FEM2D_BVP_SERENE_PRB Normal end of execution. 07 July 2014 09:03:08 PM