Start with a physical question, change one quantity, and inspect the result. These diagrams explain the controls; the wind-tunnel views contain the calculated fields.
01 / WHICH EXPERIMENT?
Two questions, two models.
Interactive: change geometry and viscosity, then observe low-Re viscous flow. Standard SU2: study pressure and section forces for a fixed, fully turbulent NACA 0012 case.
The different Reynolds numbers, boundaries and equations make them different experiments. An edited interactive shape is not transferred into SU2.
Positive incidence raises the nose relative to the trailing edge in the displayed left-to-right flow. Camber is the curvature of the mean line; thickness is a separate geometric quantity. Moving a handle edits a stationary shape and restarts the flow.
Geometry illustration only; moving this slider runs no flow calculation.
03 / VISCOSITY AND SCALE
Match the regime.
Reynolds number compares inertia with viscosity. Equal shapes alone do not define comparable flows. This calculator uses SI units; it does not change either solver.
In the interactive model, U∞ and the original reference length stay fixed; its Re control changes viscosity. The SU2 values above describe the numerical benchmark, not standard sea-level air.
04 / PRESSURE AND FORCES
Read the whole distribution.
Cp = (p − p∞)/(½ρ∞U∞²). A negative value indicates pressure below the reference. The surface chart puts lower pressure upwards; it does not invert the flow.
Pressure acts normally and viscous wall shear acts tangentially. Their integrated resultant has lift and drag components relative to the free stream. A single high-speed patch does not determine total lift.
Instantaneous streamlines follow the velocity directions at one moment. Moving tracers follow the time history of the calculated field. In unsteady flow these can differ.
SU2 here computes a stationary mean field: its moving markers show advection through that field. They do not depict time-resolved turbulent eddies or vortex shedding.
SU2 checks log₁₀ RMS pressure, both velocity and turbulence residuals; CL and CD ranges over the final 100 iterations; and relative net farfield mass imbalance.
Passing these checks establishes this run’s numerical completion. Experimental validation and mesh independence remain separate. The live model instead shows instantaneous diagnostics and may remain unsteady.
Start with NACA 0012 at zero incidence. Save a snapshot, then change camber or trailing-edge deflection while keeping Re and the calculation grid fixed. Compare similar convective times and the same display layer. For SU2, load the 0° and 4° examples and compare their native pressure distributions.
A positive trailing-edge setting bends the aft section downwards smoothly. It represents a sealed morphing shape, not a hinged or slotted multi-element flap. The low-Re model does not establish an aircraft stall angle.
Keep your experiment reproducible.
Design JSON stores the editable geometry and conditions. Loading it restarts the live flow. Result JSON includes a captured field, numerical status and model identity. Field CSV exports normalised samples from the live lattice or the interpolated SU2 display grid. Surface Cp CSV uses SU2’s native surface samples.
Saved results are local to this browser and device. New SU2 calculations require the local CPU service; the precomputed examples and comparisons work without that service.