Convergence of the elliptic-elliptic mapped system is hampered by the integration of the microscopic function in the right hand side of the macroscopic equation.
I think the routine is mathematically correct, but perhaps suffers from numerical errors. This has two
consequences:
The error does not decrease monotonically in the Picard iterations
The convergence is less than second order
Replacing the numerically integrated value with the exact value fixes these problems, so I think any other sources or errors are eliminated. Two options to solve this:
Convergence of the elliptic-elliptic mapped system is hampered by the integration of the microscopic function in the right hand side of the macroscopic equation. I think the routine is mathematically correct, but perhaps suffers from numerical errors. This has two consequences:
Replacing the numerically integrated value with the exact value fixes these problems, so I think any other sources or errors are eliminated. Two options to solve this: