# Solver Convergence Review Checklist

## Purpose
This checklist provides a structured framework for evaluating whether a CFD or simulation solution has adequately converged. It is designed to help engineers identify false convergence, ensure conservation of mass and energy, and verify that engineering quantities of interest have stabilized.

## How to use this checklist
- **Review before interpreting:** Complete this checklist *before* extracting final data, creating plots, or making engineering decisions based on the simulation results.
- **Save with case files:** Copy this file into your case directory (e.g., `RUN-001-convergence-review.md`) and check off items as you verify them.
- **Attach to reports:** Include the completed checklist in design reviews or simulation reports as evidence of solution quality.

---

## 1. Project and Context Metadata
- **Project Name:** [e.g., HVAC Duct Optimization]
- **Case ID:** [e.g., RUN-012]
- **Run ID:** [e.g., ITER-2000]
- **Analyst:** [Name]
- **Reviewer:** [Name]
- **Date / Version:** [YYYY-MM-DD / v1.0]
- **Solver / Software / Version:** [e.g., commercial CFD software 2023.1, commercial CFD software 2024R1]

## 2. Simulation Type & Convergence Expectations
- **Steady vs. Transient:** [Steady-State / Transient]
- **Compressibility:** [Incompressible / Compressible]
- **Turbulence Model:** [e.g., k-epsilon Realizable]
- **Additional Physics:** [e.g., Multiphase VOF, Conjugate Heat Transfer]
- **Expected Timescale / Residence Time:** [e.g., Flow-through time is approx. 0.5s. Run should cover at least 3-5 flow-through times.]

## 3. Residual Review
*Note: Residuals alone are not sufficient to prove convergence, but they are the first indicator of numerical health.*
- [ ] **Residual Targets:** Have all primary equations (continuity, momentum, energy, turbulence) dropped by at least 3-4 orders of magnitude?
- [ ] **Residual Slope/Trend:** Are the residuals trending downwards or completely flat? (Upward trending or wildly oscillating residuals indicate instability).
- [ ] **Oscillatory Behavior:** If residuals are oscillating, is the amplitude bounded and centered around a low value?
- [ ] **Variable-Specific Residuals:** Are there specific equations (e.g., species transport, turbulent dissipation) struggling to converge? Document which ones.
- [ ] **Normalization:** Are the residuals normalized or scaled? (Be aware of solver-specific normalization methods which can mask high absolute errors).

## 4. Engineering Monitor Review
*Monitors of physical quantities of interest (QoIs) must stabilize.*
- [ ] **Forces / Moments:** Have aerodynamic forces (Lift, Drag) or moments stabilized to within acceptable tolerance (e.g., < 1% variation over the last 100-500 iterations)?
- [ ] **Pressure Drop:** Has the pressure drop across the domain/component stabilized?
- [ ] **Mass Flow:** Are inlet and outlet mass flow monitors stable?
- [ ] **Heat Transfer:** Has the global heat transfer rate or volume-averaged temperature stabilized?
- [ ] **Point/Surface Monitors:** Have specific point probes (e.g., velocity at a critical location) or surface averages stabilized?

## 5. Conservation Checks
*Global conservation must be satisfied regardless of residual levels.*
- [ ] **Mass Imbalance:** Is the net mass flow rate (Inlet Mass Flow + Outlet Mass Flow) less than 0.1% of the total mass flux?
- [ ] **Energy Imbalance:** For heat transfer problems, does Heat In = Heat Out (within < 1% error)?
- [ ] **Species Imbalance:** If solving for species, is the global mass fraction conserved?
- [ ] **Momentum / Force Checks:** Do the integrated wall forces roughly balance the change in momentum flux across the domain (if applicable/verifiable)?

## 6. Boundary Condition Consistency
- [ ] **Inlet / Outlet Behavior:** Are the flow profiles at the inlet and outlet physically realistic and consistent with the intended BCs?
- [ ] **Recirculation at Openings:** Is there unexpected flow re-entering the domain at pressure outlets? (If yes, the outlet may need to be extended).
- [ ] **Backflow Warnings:** Did the solver report excessive backflow during the run? Did it subside by the end?
- [ ] **Physical Plausibility:** Are the calculated values at boundaries (e.g., outlet temperature) within physically possible bounds?

## 7. Solver Stability & Numerical Behavior
- [ ] **Courant / CFL Behavior:** Did the Courant number remain within stable limits for the chosen scheme?
- [ ] **Time Step (or Pseudo-Time Step):** If adjusted during the run, did the solution stabilize at the final, intended time step?
- [ ] **Under-Relaxation:** Were under-relaxation factors (URFs) reduced significantly to prevent divergence? (If URFs are extremely low, the solution may appear converged while still slowly drifting).
- [ ] **Boundedness / Clipping Warnings:** Check the solver log for warnings about variables being limited (e.g., "temperature limited to 1.0 in 5000 cells", "turbulent viscosity ratio limited"). Are these isolated or widespread?
- [ ] **Divergence Warnings:** Were there any floating-point exceptions or "reversed flow" warnings that persisted to the final iteration?
- [ ] **Scheme Changes:** If first-order upwind was used for initialization, was the solution successfully transitioned to second-order (or higher) and allowed to re-converge?

## 8. Transient Review (If Applicable)
- [ ] **Periodicity:** Has the transient behavior reached a statistically stationary or periodic state?
- [ ] **Averaging Window:** Is the time-averaging window long enough to capture multiple low-frequency shedding cycles?
- [ ] **Sampled Cycles:** Have at least 3-5 distinct vortex shedding or rotational cycles been simulated after the initial startup transient?
- [ ] **Monitor Drift:** Is there a slow, low-frequency drift in the running time-average of key monitors?

## 9. Mesh-Related Convergence Checks
- [ ] **Localized Poor Cells:** Are convergence issues (e.g., high local residuals, clipping) isolated to known bad cells (high skewness, low orthogonal quality)?
- [ ] **y+ Hotspots:** Are there regions where the wall y+ violates the chosen wall treatment strategy, potentially causing wall shear/heat flux oscillations?
- [ ] **Gradient Alignment:** Are regions of high flow gradients adequately resolved, or are coarse cells causing numerical diffusion and instability?
- [ ] **Mesh Sensitivity:** Has a grid convergence study (e.g., GCI) been performed, or is this run part of one?

## 10. Physical Plausibility Checks
- [ ] **Flow Direction:** Does the bulk flow direction make sense?
- [ ] **Separation / Reattachment:** Are separation points and wake sizes physically plausible?
- [ ] **Pressure / Temperature Ranges:** Are the min/max absolute pressure and temperature within expected physical limits?
- [ ] **Velocity Magnitudes:** Are peak velocities realistic (e.g., checking for unexpected supersonic regions in low-speed flows)?
- [ ] **Sanity Check:** Does the result align with rough hand calculations, 1D flow models, or expected non-dimensional numbers (e.g., Reynolds, Mach, Nusselt)?

## 11. False-Convergence Warning Signs
- [ ] **Drifting Monitors:** Residuals are low (<1e-4), but drag/lift monitors are still slowly climbing or falling. (Solution: Run longer or check URFs).
- [ ] **High Imbalances:** Monitors are perfectly flat, but mass conservation error is > 2%. (Solution: Check mesh quality, BC setup, or solver precision).
- [ ] **Widespread Clipping:** The solver log shows continuous clipping of turbulence quantities or temperatures. (Solution: Often points to mesh quality issues or bad BCs).
- [ ] **Unstable Local Regions:** The global forces are stable, but a small, critical local region (e.g., a cooling channel) is oscillating wildly. (Solution: Switch to transient or refine mesh).

## 12. Final Interpretation Readiness
Based on the checks above, what is the status of this run?
- [ ] **Ready to Interpret:** Convergence criteria met. Solution is credible.
- [ ] **Needs More Iterations/Time:** Monitors are still drifting; residuals have not flattened.
- [ ] **Needs Mesh/BC/Model Changes:** Solution is unstable, clipping, or physically implausible.
- [ ] **Needs Validation/Reference:** Setup is questionable; requires comparison against experimental or reference data before trusting.

## 13. Reviewer Signoff
- **Analyst Notes:** [e.g., Drag monitor stabilized at iteration 1200. Mass imbalance is 0.05%. Ready for review.]
- **Reviewer Notes:** [e.g., Convergence looks solid. The small oscillation in the wake is expected for this steady assumption.]
- **Decision:** [Approved / Rejected / Run Longer]
- **Signoff / Date:** [Name / YYYY-MM-DD]

---

**Disclaimer:** This document is an educational workflow aid and template provided by the konio project. It does not replace official solver documentation, formal validation procedures, project-specific QA protocols, or professional engineering judgment. No specific simulation results, safety, or certification claims are guaranteed by the use of this template.