Jacobi Arterial 639970
1. **Problem Statement:**
Solve the linear system $Ap = q$ where $A$ is a graph Laplacian-like matrix of size $n=500$ nodes representing arterial networks, using the Jacobi iterative method starting from initial guess $p^{(0)}$ equal to the mean pressure.
2. **Jacobi Method Formula:**
The Jacobi iteration updates the solution as:
$$p^{(k+1)} = D^{-1}(q - (L+U)p^{(k)})$$
where $A = D + L + U$, with $D$ the diagonal, $L$ the strictly lower, and $U$ the strictly upper parts of $A$.
3. **Handling Ill-conditioning:**
Since $\text{cond}(A) \approx 10^6$, the system is ill-conditioned due to vessel compliance.
Ill-conditioning slows convergence and can cause numerical instability.
4. **Step (a) - 50 Iterations:**
Perform 50 Jacobi iterations starting from $p^{(0)} = \text{mean pressure}$:
- Compute $p^{(k+1)}$ using the formula above.
- Monitor convergence but expect slow progress due to ill-conditioning.
5. **Step (b) - Preconditioning with Diagonal Scaling:**
Precondition by scaling the system with $D^{-1}$:
$$B = I - D^{-1}A$$
Jacobi iteration matrix is $B$.
The error at iteration $k$ is:
$$e^{(k)} = p^{(k)} - p^* = B^k e^{(0)}$$
where $e^{(0)} = p^{(0)} - p^*$ is the initial error.
6. **Step (c) - Validation Against CFD Simulation:**
Compare the Jacobi solution after iterations with CFD simulation results.
The error in pulse wave propagation should be less than 1%, confirming accuracy.
**Summary:**
- Use Jacobi iteration formula.
- Start from mean pressure.
- Perform 50 iterations.
- Precondition with diagonal scaling to improve convergence.
- Error evolves as $e^{(k)} = B^k e^{(0)}$.
- Validate solution with CFD, ensuring error <1%.