Atomization Nozzles: Pressure, Air-Assist and Ultrasonic Compared

Atomization looks simple until the droplet size matters. A coating that needs 50 µm droplets and a humidifier that needs 10 µm are different machines wearing the same name, and picking the wrong one means reworking the whole station. The physical question is always the same: how much energy goes into breaking the liquid into smaller droplets, and where does that energy come from. Three methods answer it differently, and the difference shows up in droplet size, air consumption and cost.

This guide compares the three ways industrial buyers actually atomize liquid, pressure (hydraulic) atomizing, air-assist (pneumatic) atomizing and ultrasonic atomizing, in the order a specification decision usually goes: what droplet size the process needs, which method can deliver it, what each method costs to run, and what to put on the purchase spec so the supplier quotes the right thing. If you are comparing atomization nozzles from different suppliers, the tables below are the framework to do it with.

The Snapshot

• Method sets the droplet band before anything else: pressure (hydraulic) atomizers typically produce 100 to 500 µm droplets, air-assist designs 10 to 80 µm, and ultrasonic heads 1 to 50 µm set by frequency.
• Droplet size falls as the energy put into the liquid rises, but flow still follows Q = K times the square root of P. Doubling pressure raises flow about 41 percent and only cuts droplet size when pressure roughly quadruples.
• Air-assist buys the fine end with a permanent air bill: published designs consume about 5 to 100 Nm³/h of free air per nozzle, and the air-to-liquid ratio, not the nozzle shape, is the main lever on droplet size.
• Match the method to the process target before the hardware: dust suppression wants 50 to 150 µm, evaporative cooling 30 to 80 µm, coating 20 to 60 µm, humidification 5 to 30 µm, and impact cleaning 200 to 500 µm.

What Droplet Size Actually Means

Water atomization is the most common atomization duty in industry. Dust suppression, evaporative cooling, humidification and fire suppression all rely on it, and the physics is the same whether the liquid is water or a solvent.

To define atomization precisely: it is the mechanical breakup of a liquid into droplets, driven by pressure, air, or mechanical energy. To define atomizing in practical terms: it is the difference between a stream and a spray, the point where surface tension loses to kinetic energy.

The atomized liquid leaves the nozzle as a sheet or stream that breaks into droplets within a few diameters of the orifice. The breakup length is roughly 5–10 orifice diameters for plain hydraulic nozzles, and nearly zero for air-assist designs where the gas does the shredding.

Droplet size is reported as Sauter mean diameter (SMD), the diameter of a droplet whose volume-to-surface ratio equals the average of the whole spray. It matters because surface area drives evaporation, absorption and reaction rate. Halve the droplet diameter and you roughly double the total surface area for the same liquid volume. So a 10 µm mist evaporates much faster than a 100 µm spray even at the same flow.

Droplet size (SMD)	Typical behavior
< 20 µm	Fine mist, stays airborne, evaporates quickly
20 – 80 µm	Fine spray, settles slowly, good for coating
80 – 200 µm	Medium spray, visible wetting, cooling duty
200 – 500 µm	Coarse spray, direct impact cleaning

SMD is one number, though, and no spray is a single size. Every real spray is a distribution of drops, and the useful figures are the percentiles: Dv10, Dv50 and Dv90, the diameters below which 10%, 50% and 90% of the spray volume falls. Two nozzles can share the same SMD and behave completely differently in the process if one throws a wide spread of drop sizes and the other a tight one.

Distribution term	What it means	Why it matters
SMD (D32)	Volume-to-surface mean diameter	Drives evaporation, absorption, reaction rate
Dv50	Median: half the volume below this size	The typical drop the process “sees”
Dv90	90% of the volume below this size	Coarse tail: runs, misses, incomplete drying
Span	(Dv90 − Dv10) ÷ Dv50, width of the distribution	Wide span means some drops too fine, some too coarse

Air-assist and ultrasonic atomizers generally produce tighter distributions than plain pressure nozzles, which is part of why they are chosen for coating and dosing duty: the tail of coarse drops that causes runs and misses is smaller. If a supplier quotes only SMD, ask for Dv90 as well. The coarse tail is usually the number that decides whether the spray actually works in your process.

The process decides the target range first. Dust suppression wants 50–150 µm so droplets fall with the dust. Evaporative cooling wants 30–80 µm so they evaporate before landing. Coatings want 20–60 µm for even coverage without runs. Humidification wants 5–30 µm so the mist stays airborne until it evaporates, and cleaning wants 200–500 µm so the drops carry momentum to the surface. Write the target range and the acceptable coarse tail down before looking at any nozzle. The target is the specification, and the nozzle is the answer to it.

How Each Method Puts Energy Into the Liquid

Atomization is the liquid losing a fight against its own surface tension. The force that holds a drop together scales with surface tension and drop size; the force that tears it apart is the energy the nozzle puts in, and the three methods differ only in where that energy comes from.

• Pressure atomizing converts pump pressure into kinetic energy inside the orifice: the liquid accelerates through a small bore, and the velocity difference between the fast core and the slower edge of the jet tears the sheet into drops. The pump pays for the energy, continuously.
• Air-assist atomizing uses compressed air as the energy carrier: a high-velocity air stream shears the liquid into ligaments and drops at the nozzle tip. The compressor pays, continuously.
• Ultrasonic atomizing vibrates a liquid film at high frequency until the surface waves grow tall enough to break off drops. The electronics pay, and the energy bill is tiny. The practical limits are flow, not power.

The physics that decides how fine a spray gets is the ratio of disruptive energy to the liquid’s resistance, surface tension and viscosity. Thicker liquids need more energy per drop. That single rule explains most field behavior: a pressure nozzle tuned for water coarsens on a viscous chemical, air-assist needs higher air pressure (or an external-mix tip) on viscous duty, and ultrasonic heads have a practical viscosity ceiling. When a “bad nozzle” appears on a line, it is usually this balance that changed, a new liquid, a colder batch, a higher surface-tension additive, not the hardware.

Pressure Atomization: Simple, Cheap, Coarse

A pressure atomizer forces liquid through a small orifice at high pressure. The energy comes from the pump, and droplet size falls as pressure rises: roughly, halving the droplet size requires quadrupling the pressure. Typical pressure nozzles run 5–70 bar and produce 100–500 µm droplets.

The strength is simplicity: no compressed air, no controls, just a nozzle and a pump. It is the right answer in plants with no compressor, on huge flow duties, and anywhere the process tolerates a medium-to-coarse spray. The limit is droplet size. Below about 100 µm, pressure atomizers need pressures that strain pumps, seals and piping, and the orifice gets small enough to clog on anything but filtered liquid.

Two operating facts matter when you tune one. First, flow follows the square-root law: Q ∝ √ΔP, so doubling the pressure raises flow by about 41%. If the spray “grew” after a pressure change, that is the flow, not the pattern. Second, the spray angle is set by the swirl-chamber geometry, not the pressure: across the working band it stays close to constant, and it collapses only when the pressure drops below what the swirl needs to keep spinning. Pressure changes flow, not angle.

The practical turndown is the catch. Because flow and atomization quality are locked to the same pressure, throttling a pressure nozzle down to save liquid coarsens the spray and narrows the angle at the same time. A pressure nozzle gives roughly 2:1 to 4:1 of usable turndown before the spray degrades, while an air-assist nozzle can hold droplet size across a 10:1 liquid range by adjusting only the air. If the duty varies through the day, that difference matters more than the price of the nozzle.

Wear is the slow trap. An eroded orifice silently passes more flow and produces coarser drops, so a line that “suddenly” oversprays is usually wearing out, not mis-set. Measure the orifice, not just the pressure. The economics are also quietly nasty at the fine end: a hydraulic nozzle chasing droplets below 100 µm spends its money in high-pressure pumps and piping, which is where the air-assist alternative starts to look cheap.

Air-Assist Atomization: The Middle Ground

An air-atomizing nozzle adds compressed air at the orifice to shear the liquid into droplets. Air pressures of 1–6 bar produce 10–80 µm droplets while the liquid itself can stay at low pressure. The air-to-liquid ratio (ALR) controls the result: more air means smaller droplets and more air consumption.

ALR (air:liquid by mass)	Typical SMD
1:1	80 – 120 µm
3:1	40 – 70 µm
10:1	15 – 40 µm
20:1	8 – 20 µm

Two mix geometries cover the liquid range. Internal-mix designs atomize at lower air pressure and suit clean, low-viscosity liquids; external-mix designs keep the streams separate until the tip, so they tolerate dirty, viscous or abrasive liquids that would clog an internal chamber. For most buyers the rule is short: clean thin liquid → internal, everything else → external. The chamber mechanics are covered in detail in the internal- vs external-mix guide.

The reason air-assist dominates precision duty is decoupling. Droplet size is governed by the air; liquid pressure only sets the flow. That means you can hold a fixed droplet size while the liquid flow swings over a wide turndown, typically 10:1 or more, and the liquid feed can be pumped, siphoned or gravity-fed depending on the flow. A small siphon-fed unit on a drum atomizes without any liquid pump at all.

The trade-off is compressed air cost. Here is the arithmetic that decides whether air-assist is cheap or ruinous: at 10:1 ALR by mass, every litre of liquid needs about 8 Nm³ of free air (10 kg of air per kg of liquid at roughly 1.2 kg/m³). A 10 L/h duty therefore draws around 80 Nm³/h. Compressed air generation costs on the order of 0.1 kWh per Nm³ at typical plant efficiency, so that one nozzle is roughly 8 kWh/h of compressor load before anything else. At 2 L/h it is a rounding error; at 500 L/h it is a compressor purchase. That is why air-assist is specified where droplet size matters more than air cost, and avoided on huge flow duties. The full costing method is in the air consumption guide.

One field rule that never changes: the air must be clean and dry. Oil mist and condensate foul the mixing passages within days, and a pressure that sags under demand makes droplet size wander with the duty cycle. Filter and regulator at minimum, a dryer where the compressor room is humid. The nozzle is only as good as the air line behind it.

Ultrasonic Atomization: Fine and Fragile

Ultrasonic atomizers vibrate a liquid film at 20 kHz or higher until it breaks into droplets. The mechanism is capillary waves: the vibrating face sets up standing surface waves on the film, and when the wave amplitude passes a threshold, droplets shear off the wave crests. Droplet size tracks the wave length, which falls with frequency: roughly, d ∝ (σ / ρf²)^(1/3), where σ is surface tension, ρ is density and f is frequency. That is why a 20 kHz head lands in the tens of microns while the megahertz heads used in medical nebulizers reach 1–5 µm.

They produce 1–50 µm droplets at very low liquid pressure and almost no air. There is no air, no high-velocity jet and no impact. The spray leaves the surface slowly, which is exactly the point for delicate coatings and for processes where compressed air is not allowed near the product.

The limits are the other side of the same coin. A single ultrasonic head handles only a few litres per hour, so scaling up means adding heads, not buying a bigger one. The liquid has to reach the vibrating surface as a thin film, which rules out pumped high-pressure feed, and it has to be free of particulates. Particles lodge on the transducer face and kill atomization. Viscous liquids resist capillary-wave breakup, so the practical ceiling is well below what an air-atomizing nozzle handles with an external-mix tip. And because the droplets move slowly, a long fall path lets them re-coalesce or drift on room air currents, fine for a short controlled delivery, wrong for a process that needs drops to carry momentum to a target. Ultrasonic suits humidification, fine coating and medical or research misting, not process-scale flow. The fuller treatment is in the ultrasonic atomization guide.

Atomizing Pressure: Where It Fits

Raising atomizing pressure shrinks droplets and raises flow together, which is useful on pressure nozzles but creates a calibration trap: operators who raise pressure to improve atomization also change the flow rate and the impact force without meaning to. On air-assist systems the atomizing pressure refers to the air, and the liquid pressure can stay flat. The two are decoupled, which is exactly why air-assist dominates precision atomizing duty.

Over-atomization is the error nobody budgets for. Push the air pressure too high and the spray goes so fine it drifts, re-entrains into the airflow or evaporates before reaching the target, a process change that looks like a nozzle failure but is a pressure decision. “Finer” is not a direction; it is a target that has a correct value, and the correct value comes from the process, not from the gauge.

The Three-Way Selection Decision Table

Before the detail, the map. The working ranges below are engineering guidance for water-like liquids at typical operating conditions. They are the envelope each method lives in, not guarantees from any catalog:

Method	Droplet size (SMD)	Air consumption	Cost profile	Best-fit applications
Pressure (hydraulic)	100 – 500 µm (finer at high pressure)	None	Low equipment, no air cost; high pump cost at the fine end	Bulk cooling, dust suppression, tank and surface cleaning, spray drying feed
Air-assist (pneumatic)	10 – 80 µm (down to ~5 µm at high ALR)	5 – 100+ Nm³/h per nozzle	Moderate nozzle cost; continuous compressed-air cost	Coating, humidification, gas conditioning, dosing, release agents
Ultrasonic	1 – 50 µm, set by frequency	None	Highest per-unit cost; flow capped per head	Humidification, precision fine coating, medical and research misting

Read the table as a flow-and-cost budget, not a beauty contest. If the duty is coarse and the plant has no compressor, pressure wins and nothing else is close. If the process needs droplets below roughly 80 µm and you have air, air-assist is the workhorse that also gives you turndown. If you need a fine mist at a few litres per hour and air is not allowed near the product, ultrasonic. The most common mistake is buying the finest atomizer the budget allows and hoping the process tolerates it. Start from the droplet target, not from the finest brochure photo.

Common Atomization Misconceptions

Seven beliefs show up in every first specification review, and all of them cost money:

Misconception	What is actually true
“Higher pressure always means finer spray.”	Only on hydraulic nozzles, and only within limits: pressure also raises flow and impact. On air-assist, the air sets the droplet size, not the liquid pressure.
“Same SMD means the same spray.”	Distribution differs. Compare Dv90 and span, not just SMD.
“A finer mist is always better.”	Over-atomized spray drifts, evaporates before the target and skips coverage. Dust suppression and cleaning need drops with weight.
“Air-assist is always the expensive option.”	At small flows, air is cheap next to the pump and piping a high-pressure hydraulic system needs. At large flows the compressor bill flips the math.
“Ultrasonic scales up like a pump.”	Flow is capped per head. Scaling means multiple heads and multiple costs.
“Pressure changes the spray angle.”	Angle is geometry. Pressure changes flow; the angle only collapses below the design range.
“The nozzle is the whole system.”	Air prep, filtration and pressure stability decide field performance. Two “identical” installations differ only in what feeds the nozzle.

The pattern behind all seven: the spray is decided upstream of the nozzle, by the pressure, the air and the liquid, and the nozzle is the last, cheapest link. When a spray misbehaves, diagnose the upstream variables first.

Matching the Method to the Duty

Duty	Droplet target	Method
Dust suppression	50 – 150 µm	Pressure (simple)
Evaporative cooling	30 – 80 µm	Air-assist
Coating / release agent	20 – 60 µm	Air-assist
Humidification	5 – 30 µm	Ultrasonic or air-assist
Tank / surface cleaning	200 – 500 µm	Pressure
Spray drying feed	80 – 200 µm	Pressure or air-assist

If the droplet size is specified by the process, the method follows: coarse duty takes pressure, precision duty takes air-assist, tiny-flow fine duty takes ultrasonic. If the droplet size is not specified, measure what the line actually needs before buying. Most atomizing mistakes come from guessing the target range. Where two methods overlap, as humidification does, the tie-breakers are air availability, liquid flow rate, and whether fine droplets may drift onto the product. If you need to pin the target down in numbers first, the droplet size guide walks the measurement and the math.

What an OEM Buyer Should Put in the Spec

Once the method is chosen, the buying conversation should be short and numeric. For an OEM or line-builder purchase, put these lines on the request for quote:

1. Droplet target: the SMD range and the Dv90 cap, plus how they were measured (laser diffraction at what pressure, what distance). A supplier who quotes a single droplet number without conditions is quoting marketing.
2. Flow range and turndown: the minimum and maximum liquid flow the line must cover, and whether droplet size has to hold across the whole range. This alone disqualifies pressure nozzles for variable-duty precision work.
3. Air budget (air-assist): the air pressure available at the nozzle and the consumption at your operating point, checked against the compressor’s duty cycle. Ask for the consumption at your ALR, not the minimum in the brochure.
4. Liquid properties: viscosity, surface tension, solids content and temperature. All of them shift droplet size, and none of them appear on a pretty photo.
5. Pattern and distance: full cone, flat fan or round, the angle at a defined distance, and the mounting distance. “Fine mist” is not a pattern spec.
6. Wetted materials: the liquid-side materials against your chemistry. Do not pay for 316 where the liquid is water; do not accept brass where the liquid eats it.
7. Media preparation: the required filtration level for the liquid and the air cleanliness spec, written into the install scope so the plant actually provides it.
8. Verification: the performance curve at your conditions and, if the droplet range is critical, a test sample on your line before the production commitment.
9. Spares and life: the expected wear items and their cost. On high-air-duty, the tip and internals are consumables, and the spare price is part of the running cost.

An adjustable air-assist nozzle buys tuning range after install, which is genuinely useful while the process is still being learned, but it does not replace right-sizing the base unit to the air you actually have. Buyers who skip these lines end up paying the difference between the catalog and the line: the spec sheet costs nothing at the design stage and everything after a production run.

Frequently Asked Questions

Which method makes the smallest droplets? Ultrasonic, at the high-frequency end. Medical nebulizers reach 1–5 µm. Air-assist goes down to roughly 5–10 µm at high air-to-liquid ratios. Pressure nozzles realistically floor out around 50–100 µm before pump and clogging costs take over.

Can a pressure nozzle reach the fineness of an air-atomizing nozzle? Only with very high pressure and a very small orifice, and then you pay for it in pumps, piping and clogging. Below roughly 100 µm, air-assist is usually the cheaper total answer.

Why is compressed air so expensive in atomizing duty? Compressing air takes real energy, on the order of 0.1 kWh per Nm³ at typical plant efficiency, and consumption scales with ALR times liquid flow. At 10:1 ALR every litre of liquid costs about 8 Nm³ of air. Small flows, small cost; big flows, budget for it.

Does an air-atomizing nozzle need dry, filtered air? Yes. Oil mist and condensate foul the mixing passages within days, and a sagging supply makes droplet size wander with the duty cycle. Filter and regulator at minimum.

Can I convert an existing pressure-nozzle line to air-assist? Usually, if a compressor and a regulated air line exist. The liquid pump can stay or be replaced by siphon feed at low flows, and the nozzle body changes. The main new scope is the air piping and preparation, which is where most conversions fail.

Why did the spray get coarser this quarter? Check in this order: liquid temperature and viscosity change, worn orifice, then sagging air pressure. Cold batches atomize coarser; worn orifices pass more flow at the same pressure; a tired compressor makes the droplet size follow the duty cycle.

Is ultrasonic suitable for viscous liquids? Poorly. Capillary-wave breakup struggles as viscosity rises, and the film feed demands a clean thin liquid. For viscous duty, an air-assist nozzle with external mix is the better answer.

Which method is cheapest to run? Pressure has the lowest running cost whenever it can meet the droplet target: no air, simple hardware. Air-assist adds continuous compressor energy. Ultrasonic uses almost no energy but carries the highest per-head cost and a hard flow ceiling.

How do I verify droplet size before committing? Laser diffraction is the standard. Ask the supplier to test at your pressure and flow, not the catalog best case, and at minimum compare Dv90, because the coarse tail is what causes runs and misses.

What is the difference between SMD and Dv90? SMD is the volume-to-surface average that drives evaporation and reaction chemistry; Dv90 is the coarse tail that drives runs, misses and incomplete drying. Quote both, because they answer different questions.

If the droplet target lands in the 10–80 µm band, the BoreJet air atomizing range covers it nozzle-for-nozzle, with per-model air consumption, pattern and droplet data at defined conditions. Send the application team your droplet target, flow rate and liquid properties, and the method gets matched before you spend on hardware, and for the wider flow-pressure-pattern decision, the spray nozzle selection guide is the better starting point.