Injector design — impinging, coaxial, swirl injectors: Visual walkthrough

We will lean on ideas already in the vault: Bernoulli Equation (how spinning liquid gains speed), Atomization and the d-squared Law (why a thin cone matters), and Combustion Instability (why we care about the spray at all).

Step 1 — Two motions, drawn separately

WHY. Any motion in space, however curvy, can always be split into simple straight-line pieces along chosen directions. This is the whole trick of physics: decompose, then recombine. We choose two directions that match the injector's own geometry.

The axial direction — straight down the chamber axis, the way the rocket will push its exhaust. Call the blob's speed in this direction $v_{x}$ (the little $x$ is for "axis").
The tangential direction — sideways, along the circle the blob was spinning on inside the vortex chamber. Call this speed $v_{t}$ ( $t$ for "tangential", meaning "along the edge of the circle").

Figure — Injector design — impinging, coaxial, swirl injectors

Look at the amber blob: the cyan arrow points forward ( $v_{x}$ ), the white arrow points sideways ( $v_{t}$ ). Neither arrow alone is the real motion — the real motion is still to come.

Step 2 — Where the sideways motion comes from: the vortex chamber

WHY this design. If you squirt water straight at a drain it just plunges through. If you squirt it along the wall of a round bowl, it swirls — exactly like water going down a plughole. The swirl injector deliberately makes the plughole vortex so the exiting liquid is already spinning.

The amber inlet arrow enters along the wall, not toward the centre. Follow the cyan spiral: the blob rides the wall, spinning faster as it nears the small exit (this speed-up is a Bernoulli effect — narrower path, faster flow). By the time it reaches the exit it carries a real tangential speed $v_{t}$ .

Step 3 — Combine the two arrows: the velocity triangle

WHY tip-to-tail. Velocities add like displacements: if in one second you slide forward by $v_{x}$ and sideways by $v_{t}$ , your net move is the straight line from start to finish — the diagonal of the rectangle those two make. That diagonal is the resultant velocity.

The blue right triangle has a cyan leg ( $v_{x}$ , forward), a white leg ( $v_{t}$ , sideways), and an amber diagonal ( $v$ , true motion). The little square in the corner marks the $9 0^{\circ}$ between forward and sideways — they are perpendicular by construction, which is exactly why Pythagoras applies.

Step 4 — The angle $ϕ$ is born, and it is a tangent

WHY the tangent, and not sine or cosine? We want a number that captures "how much sideways per unit forward" — the steepness of the tilt. On the right triangle from Step 3:

The opposite side to $ϕ$ (the leg facing the angle) is the sideways leg $v_{t}$ .
The adjacent side (the leg touching the angle, along the axis) is the forward leg $v_{x}$ .

The ratio opposite ÷ adjacent is the definition of the tangent — "how far you rise sideways for each step you take forward". That ratio is precisely the tilt's steepness, so tangent is the natural tool. Sine would need the diagonal length; cosine too; but we can read the tilt from the two legs alone, and that ratio is $tan$ .

The amber wedge at the corner is $ϕ$ . Notice how it sits between the axis (adjacent, cyan) and the true-velocity diagonal (amber). This is the exact same right-triangle reasoning you would use for $θ = arctan (A_{y} / A_{x})$ on any vector — here $v_{t}$ plays the role of the "up" component and $v_{x}$ the "along" component.

Step 5 — From one tilted blob to a whole hollow cone

WHY hollow (not filled). The liquid rode the wall of the vortex chamber, so it all exits from the rim of the exit hole, not the centre. The centre is empty (an air core). Rim + constant tilt = a thin conical shell, open in the middle — the classic hollow-cone spray.

Each white arrow is one blob leaving at half-angle $ϕ$ ; rotate that arrow around the axis and it sweeps the cyan cone. The amber dashed circle shows where the sheet lands — a ring, not a dot, proving the cone is hollow.

Step 6 — Edge cases: turn the spin all the way down, then all the way up

Case A — no swirl, $v_{t} = 0$ . Top of the fraction is zero, so $tan ϕ = 0$ , so $ϕ = 0^{\circ}$ . Picture: the blob flies straight down the axis — a pencil-thin jet, no cone at all. This is what a plain (non-swirl) orifice does. ✓ makes sense.

Case B — huge swirl, $v_{t} ≫ v_{x}$ . The fraction blows up toward infinity, and $tan ϕ \to \infty$ means $ϕ \to 9 0^{\circ}$ . Picture: the cone flattens into a nearly flat disc sprayed sideways — great thinning, but it would scrub the chamber walls. Designers stop short of this. ✓ the limit behaves.

Case C — the danger zone, $ϕ \to 9 0^{\circ}$ . As $ϕ$ approaches $9 0^{\circ}$ , $tan ϕ$ grows without bound — a tiny extra bit of spin now swings the cone open enormously. So near- $9 0^{\circ}$ designs are twitchy and prone to wall wetting. This is why real swirl injectors sit comfortably around $ϕ \approx 4 5^{\circ}$ – $6 0^{\circ}$ , where the tangent changes gently.

Three sprays side by side: no-spin pencil (left), balanced cone (middle, the useful design), flat disc (right). Watch the amber $ϕ$ wedge open from $0^{\circ}$ to nearly $9 0^{\circ}$ as $v_{t} / v_{x}$ climbs.

Step 7 — A number to hold onto

The one-picture summary

Everything on one frame: the tangential inlet spins the blob (giving $v_{t}$ ), the blob also moves forward ( $v_{x}$ ), the two combine tip-to-tail into the true velocity $v$ tilted by $ϕ = arctan (v_{t} / v_{x})$ , and sweeping that tilt around the axis paints the hollow cone.

Recall Feynman: the whole walkthrough in plain words

Imagine pouring water into a round sink so it swirls before going down the plughole. It leaves the hole spinning sideways and dropping forward at the same time. Draw those two motions as two arrows — one forward, one sideways — and the real path is the slanted diagonal between them. How slanted? "Sideways-per-forward" — that's exactly what a tangent measures, so the tilt angle $ϕ$ satisfies $tan ϕ = v_{t} / v_{x}$ . Because the water leaves from the rim of the hole (it rode the wall), and every drop tilts by the same $ϕ$ , sweeping them round makes a thin hollow cone. Spin a lot → wide flat cone → very fine mist. Spin none → a boring straight pencil. That's the swirl injector in one breath.

Recall Quick self-test

Why is it $tan$ and not $sin$ ? ::: Because we read the tilt from the two legs (opposite $v_{t}$ , adjacent $v_{x}$ ) alone, and opposite/adjacent is the tangent. What happens to the cone as $v_{t} \to 0$ ? ::: It collapses to a straight axial jet, $ϕ \to 0^{\circ}$ . Why is the spray hollow, not solid? ::: The liquid rides the vortex-chamber wall and exits from the rim, leaving an empty air core in the middle.