Context: Physical modeling of large river systems requires immense space.

Application: If a $10 \text{ km}$ wide river is modeled at a $1:1000$ scale, the model is $10 \text{ m}$ wide. However, if the real river is $5 \text{ m}$ deep, an undistorted model would be only $5 \text{ mm}$ deep. At this depth, surface tension and viscous forces dominate, completely destroying the Froude similarity required for open channel flow. Therefore, engineers use distorted models, applying a different scale for the vertical and horizontal dimensions (e.g., $1:100$ vertical and $1:1000$ horizontal) to maintain sufficient depth in the model while keeping it a practical size.

Context: Testing wind loads on skyscrapers using small models.

Application: Wind forces are governed by Reynolds number similitude. However, achieving exactly the same Reynolds number for a small model requires extremely high wind speeds in the tunnel, often reaching supersonic levels where compressibility changes the flow physics entirely. Instead, engineers rely on the fact that for blunt bodies with sharp edges (like buildings), the flow separation points are fixed by the geometry, and the drag coefficient becomes independent of the Reynolds number once a certain critical threshold is crossed. They test the model above this threshold to predict forces on the full-scale building.