I Owned the Right Number and Used It on the Wrong Thing

Single-Mirror Schlieren Imaging of Heat and Shockwaves

The razor slid into the cone of light exactly the way six different tutorials promised, and the screen went the colour of dishwater. Not bright on one side and dark on the other. Not the clean guillotine cut where half the field snaps to black. Just a flat, sullen grey that dimmed a little more with every millimetre I nudged the blade, like a lamp on a rheostat with no opinions.

That is precisely what a schlieren cutoff is not supposed to do.

The pitch is almost indecent in its simplicity. A pinprick of light. A concave mirror down at the far end of the garage. A knife-edge parked at the focus so it shaves away the light that travelled straight, leaving only the rays that got bent by warm air, or cold gas, or the shock front off a toy cap gun. Bend a ray a few arc-seconds and it either clears the blade or gets eaten, and that tiny difference blooms into visible contrast. You get to photograph the density of the air itself. No smoke.

That last part is why I was down there at all. Two nights earlier the wind tunnel had finally shown me a wing stalling — real flow separation peeling off the leading edge — and I’d managed to photograph it as a smear of grey nothing, because seeded smoke under flat light washes out to porridge. Schlieren was supposed to be the escape from grey. And here I was, in a different room, staring at the identical grey. The universe has a filing system and I keep landing in the same drawer.

The mirror, at least, wasn’t the crack from last spring. My own attempt at grinding one is still a story that ends with a fault line running centre-to-edge and me looking at it too long. This was a bought 150 mm f/8 primary, long-focus, well-figured, exactly the kind of mirror the technique wants. So when the field refused to cut, I did the reasonable thing and blamed the good mirror. Cleaned it. Re-seated it in the cell. Swapped the LED for a brighter one and stopped it down with foil and a pin. Changed razor blades — a nick in the edge can smear the cutoff. Nothing. Grey, greyer, black. No plume, no shimmer, no soldering iron ghost.

I spent the better part of two hours convinced the optics were bad before the arithmetic caught up with me.

A concave mirror doesn’t image a nearby point source<<<SLEPP_ARTICLE_BODY>>> The razor blade slid up into the point of light exactly the way six different tutorials had promised, and the screen went the colour of dishwater. Not bright on one side and dark on the other. Not the clean guillotine cut where half the field drops into shadow while the other half stays lit. Just a sullen, uniform grey that dimmed a little more with every turn of the micrometer, like a lamp on a rheostat. No structure. No plume. No nothing.

A schlieren cutoff is supposed to snap. You creep the knife-edge in, the field stays evenly lit, and then at one precise position it tips over — the whole frame darkens together and any sliver of moving air suddenly writes itself across the image in bright and dark. I got the darkening. I never got the writing.

The setup is almost insultingly simple on paper. A bright pinhole of light. A long-focus concave mirror a few metres away that bounces the light back to a focus. A razor blade parked at that focus to shave off the undeflected rays. Warm air drifting through the beam bends light by a few arc-seconds — enough that some rays clear the blade and some don’t, and that difference paints density gradients you cannot otherwise see. Toepler was doing this in 1864 to hunt streaks in optical glass. The German word Schlieren literally means those streaks.

I had reasons to expect this to work. The mirror is a store-bought 150 mm f/8 primary — I’m not trusting my own ground glass yet, not since the first blank I tried to grind cracked from centre to edge inside twenty minutes because my stand wasn’t level. And the whole appeal was a rebuke to last week’s wind tunnel, where I finally watched a wing stall but the smoke photographed as flat, washed-out grey. Schlieren promised the opposite: clean air, no seeding, no smoke, sharp contrast pulled out of nothing. And here I was, staring at grey again. I had escaped grey smoke and walked straight into a grey screen.

So I debugged. Brighter source. Cleaned the mirror with a puff of air and a lens brush. Swapped the razor for a fresh one in case the edge was nicked. Rotated the blade vertical, then horizontal, on the theory I was cutting along the wrong axis for a rising plume. I held a soldering iron in the beam — 340 °C of guaranteed convection — and got a faint, mushy smear if I squinted and lied to myself. I started to suspect the mirror was garbage, which is the point in any project where I should stop and go make tea.

The mistake was geometric, and it was embarrassingly mine.

I had placed the source and the knife-edge at the mirror’s focal length — about 1.2 metres — because “focus” is the word everyone uses and 1.2 m is where a distant star would focus. But a nearby point source doesn’t image at the focal length. It images at the centre of curvature, twice the focal length, roughly 2.4 m out. The light was crossing well behind my blade. I wasn’t cutting the image of the source at all; I was chopping into a fat, out-of-focus cone, which is exactly why the field faded like a dimmer instead of tipping like a switch.

The infuriating part is that I know this geometry. It is the Foucault knife-edge test — the same razor-at-the-centre-of-curvature trick I’d use to figure a telescope mirror, the same Léon Foucault whose pendulum I had swinging down the stairwell three days ago. Same man, same blade, same 2f, and I still set it up at f like a tourist.

I owned the correct number. I just used it on the wrong quantity.

I dragged the tripod back to 2.4 m. Re-found the returning point of light, a hot little star hanging in mid-air. Eased the blade in — and for about four seconds the plume off the soldering iron ghosted upward across the frame, a lazy braid of hot air rolling over on itself, unmistakably there. Then it smeared. Off-axis, the single-mirror rig throws coma, and my source and blade were sitting too far apart, so the point stretched into a comet and the contrast bled out sideways.

I did not get a photograph. I got four seconds of proof and a headache, and then it was past eleven and the garage was cold. The braid is still there tonight, rising off everything warm in the room, whether or not I’ve got the blade in the right place to see it.