The final stage of a CFD study is to explain what the results support.
Bring together the physical assumptions, numerical checks, and flow observations to produce a clear engineering conclusion.
Learning objective
Write a concise simulation report that connects conclusions to evidence and states the remaining limitations.
Key idea
The strength of an engineering conclusion depends on the evidence supporting it.
Understand the idea
Begin by restating the question, geometry, and operating conditions.
Explain the important flow features and connect them to measurable quantities. Distinguish observations from interpretations: a region of high velocity is an observation; an explanation of why it occurs requires physical reasoning.
Keep verification and validation distinct:
- Verification examines whether the equations are being solved correctly and assesses numerical error.
- Validation examines how well the model represents physical behaviour through comparison with suitable experimental evidence.
Agreement between two simulations is a useful comparison, but it does not by itself establish experimental validation.
Your conclusions should also reflect the limitations of the MRF approximation and the selected turbulence, boundary-condition, and mesh treatments.
Engineering example
Create a copy of the baseline case and change one documented operating parameter, such as rotation speed.
Record what changed and what remained fixed. Compare both cases using consistent numerical checks, output definitions, and plotting scales.
A difference in the predicted result becomes useful only when you can explain the conditions under which that comparison was made.
Practical task
Prepare your final report in this order:
- Engineering question and objectives.
- Geometry and operating conditions.
- Physical models and assumptions.
- Mesh description and quality checks.
- Solver settings and convergence evidence.
- Pressure and velocity results.
- Controlled comparison, if completed.
- Numerical uncertainty and model limitations.
- Engineering conclusions.
- Software versions, case release, and references.
Clearly identify checks that were performed and those that remain outstanding.
Checkpoint
Can a converged simulation with plausible-looking flow fields be described as validated?
Not on that basis alone. Validation requires comparison with suitable experimental evidence for the intended application, considering the uncertainties in both the simulation and the measurements.
References
- Verification Assessment — NASA
- Validation Assessment — NASA
- The starter release and the logs supporting your report.