Underestimated Displacements and Velocity

[MUHAMMAD NOUMAN RAFIQ]

Hi ,

I’m using PIVlab for a BOS (background-oriented schlieren) analysis with one static reference image and a sequence of dynamic images. I form pairs as (A+B), (A+C), …, (A+end).

Issue: With a time separation of Δt = 1 s per pair, the measured displacements are very small, so the velocity vectors end up numerically equal to the displacements. When I recompute the same data assuming 1000 fps (Δt = 1 ms), PIVlab reports a maximum velocity magnitude of about 0.2 m/s, which is far below the actual flow speed (~1.7 m/s). How can I resolve this discrepancy?

Setup details

• Calibration: 388.77 pixels = 40 mm

• Multigrid interrogation windows: 128 → 64 → 32 → 16 px, 50% overlap each pass

• Correlation robustness: tried both “standard” and “extreme

[thielickeoptolution]

Afaik, this is not how BOS velocity analysis works. When I understand correctly, you would have to export the optical density gradient maps, then do PIV on these ones to get a velocity estimate.