It shows where air goes around the shape. It cannot tell you how the real car behaves at speed. A 1:64 model is 64 times smaller than the car and sits in air moving far slower than a car on a road, so the flow around it is not a scaled-down copy of the real thing. That is why desktop wind tunnels - ours included - are flow-visualisation displays, not measurement instruments.
That is not a knock on them. It is just the honest version of what the fog is showing you.
Why a small model is not a small car
How air behaves around an object depends on a single ratio engineers call the Reynolds number: roughly, speed multiplied by length, divided by a property of the air. Two shapes only share the same flow pattern if that number matches.
| Real car | 1:64 model on a desk | |
|---|---|---|
| Length | ~4.6 m | ~72 mm |
| Air speed (illustrative) | 30 m/s, about motorway speed | A few metres per second |
| Reynolds number (order of magnitude) | ~9 million | ~20,000-30,000 |
The model's number is several hundred times smaller. To close that gap with length fixed at 72 mm, the air would have to move about 64 times faster than the real car: roughly 1,900 m/s for motorway speed, more than five times the speed of sound. No desktop tunnel does that, and no desktop tunnel could - so the flow you see is a different regime from the one a real car lives in.
What the fog does show well
| You can see | What it looks like |
|---|---|
| Where flow follows the body | Fog lines stay smooth and close over a rounded roof or bonnet |
| Where flow breaks away | Fog lifts off and curls at sharp edges, a cut-off tail or the back of a wing |
| Wakes | A tumbling, messy region behind the car and behind the wheels |
| Relative differences between shapes | A wedge and a box behave visibly differently in the same airflow |
That is genuinely useful for understanding why cars are shaped the way they are, and it is what makes a model look like it is moving on camera.
What it cannot show
- Numbers. Drag and downforce figures do not transfer from a model at desk speed to the real car.
- Exactly where real flow separates. At low Reynolds numbers, flow tends to detach earlier than it would on the real car, so treat what you see as exaggerated rather than accurate.
- Ground effect. Under a real car the road is moving relative to the body. In a static tunnel it is not, so the underbody flow is not representative.
- Cooling and internal flow. A 1:64 casting has no working radiator or ducts.
How real aerodynamicists deal with scale
They attack the same ratio from both sides: much larger models, faster air, moving floors under the model, and computer simulation alongside the tunnel. Motorsport teams test models that are a large fraction of full size, not toys. The physics that rules out a 1:64 model is the same physics that makes those tunnels so expensive.
Reading the fog
- Smooth, parallel lines hugging the surface mean the flow is attached.
- Lines breaking into swirls mean separation - look for where it starts.
- Fog that hangs and collects behind the car marks a slow, recirculating wake.
- Compare, do not measure. Swap two cars under the same settings and watch the difference; that comparison is the honest use of the tool.
What about tufts?
Real cars are often tested with tufts - short lengths of wool taped to the bodywork that lie flat in attached flow and flutter where it separates. It is a proper technique at full size. At 1:64 it mostly does not work, because even the finest thread is far too stiff and heavy relative to the model to follow the air.
So what is a desktop tunnel for?
Seeing shape and airflow together, teaching why a car looks the way it does, and making a parked model read as moving on camera. Our WindChaser is built for exactly that; how to shoot with one is in using a desktop wind tunnel for diecast photos, and what is in the fog - and whether it is safe on a car you care about - is in is fog safe for diecast cars. Every published figure is on the specs page.
FAQ
Does a desktop wind tunnel show real aerodynamics?
It shows where air flows around the model's shape, but not how the real car behaves at speed. A 1:64 model in slow air sits in a very different flow regime from a full-size car on a road, so it is a visualisation tool, not a measurement instrument.
Why doesn't a small model behave like the real car?
Flow patterns only match when speed multiplied by length matches. A 1:64 model would need air moving about 64 times faster than the real car's speed - roughly 1,900 m/s for motorway speed, more than five times the speed of sound.
What can you learn from fog over a model car?
Where the flow follows the body, where it breaks away at edges and wings, and how the wake forms. It is best used to compare two shapes under the same settings rather than to read exact behaviour.
How do racing teams test aerodynamics with models?
With much larger models, faster airflow, moving floors under the model and computer simulation. The scale problem that rules out a 1:64 model is why professional tunnels are so large and expensive.
Can you tuft-test a 1:64 car?
Not usefully. Tufts - short threads taped to the body - work on real cars, but at 1:64 even the finest thread is too stiff and heavy relative to the model to follow the airflow.