Table of Contents
A flat fan header looks simple: a pipe, a row of nozzles, liquid in, spray out. The simplicity hides the one number that decides whether the belt comes out evenly wetted or with dry stripes every few inches. That number is the spacing between nozzles, and it has to be calculated from the pattern, not copied from last year’s drawing.
This guide is for the engineer laying out a new wash, rinse or coating header, or fixing one that never covered right. We will treat flat spray nozzles as the working example, but the method applies to any sheet-forming nozzle in the family. The layout procedure has five steps, in order: fix the standoff, read the real coverage width, choose an edge type, set the overlap, then size the pump to the summed flow. Skip a step and the failure shows up on the belt, not on the drawing.
Start With the Standoff, Because It Is Usually Fixed
The first input is not the nozzle. It is the distance from the header to the belt or part. In most machines that distance is dictated by the enclosure, the roller height or the clearance the product needs. You rarely get to choose it; you have to design around it.
Why it comes first: coverage width is a function of standoff and spray angle. Until you know the standoff, you cannot pick an angle, and until you pick an angle you cannot say how many nozzles you need.
A practical note on measuring it: use the distance from the nozzle orifice to the target surface, not from the pipe centreline and not from the nozzle body. Most flat spray nozzles have a machined face, and the pattern originates at the orifice disc inside. Measure to the orifice plane and write that number down as the working standoff.
Coverage Width From Angle and Distance
The width a flat fan covers on a flat surface is:
width = 2 × standoff × tan(angle ÷ 2)
The table below gives the standard angles at three common standoffs. Move the header from 300 mm to 500 mm and every width grows in proportion; pull it to 200 mm and they all shrink.
This is why a “65° nozzle” is not a fixed-width device. The same tip covers 380 mm at 300 mm standoff and about 640 mm at 500 mm. When a layout fails, nine times out of ten someone quoted the 300 mm number while running the header at a different distance.
| Spray angle | Width at 200 mm | Width at 300 mm | Width at 500 mm | Where the width lands |
|---|---|---|---|---|
| 15° | 53 mm | 80 mm | 130 mm | Cutting and descaling lines |
| 25° | 89 mm | 130 mm | 220 mm | Spot washing, part rinsing |
| 40° | 145 mm | 220 mm | 360 mm | General line washing |
| 65° | 255 mm | 380 mm | 640 mm | Coating and rinse headers |
| 80° | 335 mm | 500 mm | 840 mm | Wide belt coverage |
| 110° | 570 mm | 860 mm | 1430 mm | Low-pressure rinse curtains |
| 145° | 1260 mm | 1900 mm | 3200 mm | Fogging, dust knockdown |
Widths are geometric values for a perpendicular aim on a flat surface; real patterns land a little narrower with viscous fluids and a little wider at the very top of the rated pressure band.
The same geometry works backwards. If the belt is 1000 mm wide and the standoff is fixed at 300 mm, one 110° nozzle covers 860 mm, not quite enough, while one 145° nozzle covers 1900 mm, wasting liquid past the edges. That mismatch is the moment most designers add a second nozzle; usually a pair of 65° or 80° tips at the right spacing is the cleaner answer, because it keeps the liquid on the belt.
The Overlap Rule That Removes the Seams
Once you know the real coverage width at your standoff, you space the nozzles so their patterns overlap. For tapered-edge fans the common starting point is to place them at 40–50% of the pattern width. So a nozzle covering 500 mm gets spaced about 200–250 mm from its neighbour. The thin end of each sheet then lands well inside the thick middle of the next one, and the row reads as a single continuous band.
Why 40–50% and not “just touch the edges”? Because a tapered fan is thick in the middle and thin at the ends. If you space nozzles so the thin ends just meet, every seam between nozzles is a low-density line, the pale stripes that show up on a coated belt. The thin edge of one sheet has to sit under the thick middle of its neighbour, and only a large fraction of the width guarantees that. Below roughly 30% overlap the seams start to appear; above roughly 60% you begin to double-dose the middle, though uniformity stays acceptable longer than you would expect. That tolerance band is why 40–50% is the standard home for wash and rinse headers.
If you are using even-edge fans instead, appropriate when each nozzle serves its own zone with no neighbour to blend into, you do not overlap at all; you butt them edge to edge. Mixing the two on one header is the classic source of banding, so decide the edge type up front and keep the whole row consistent.
Fan nozzles and flat fan nozzle spray variants follow the same spacing math; the edge profile and angle are what change the answer, not the name.
What “Even” Actually Means: the Uniformity Number
“Even coverage” is not a feeling, it is a number. On a moving belt the meaningful measure is the coefficient of variation (CV) of the liquid application rate across the width, the standard deviation of the local deposition divided by the mean, as a percentage. A CV of 5–10% reads as a visibly uniform band; a CV above 15–20% is where dry stripes and double-dosed lines start showing up on the product.
The uniformity you can achieve is set by three things, in order: the edge type and overlap of the row, the distribution across each individual sheet, and the pressure stability of the supply. A well-spaced row of tapered fans with a good internal distribution holds a CV in the single digits across the belt. The same row with 20% overlap instead of 45% drifts toward a CV in the mid-teens even though every part number is identical. So when a line “never covered evenly,” the first question is not what nozzle is on it. It is what overlap the row is actually running at.
Measuring it in the field is simple: run the line, catch the liquid in identical collection tubes across the belt for a timed interval, and divide the standard deviation by the mean. A strip of absorbent paper under a stationary header gives the same information in rougher form. The wet band should show a smooth gradient, not scalloped light patches between nozzles.
Edge Type Sets the Overlap Rule
The edge profile is not a cosmetic detail. It decides which overlap rule applies. Tapered-edge fans thin out gradually toward the ends so they can be blended; even-edge fans hold density almost to the boundary and then stop, so they cannot be blended without creating a seam.
| Property | Tapered-edge fan | Even-edge fan |
|---|---|---|
| Density profile | Thick middle, fading ends | Near-constant to the boundary, hard stop |
| Overlap rule | 40–50% of pattern width | None: butt edge to edge |
| Row behaviour | Blends into one continuous band | Leaves a low-density seam between pairs |
| Best duty | Multi-nozzle headers on belts | Single-zone cleaning, spot treatment |
| Typical failure | Seams reopen if spacing drifts | Banding if overlapped at all |
| Edge trimming | Outer nozzles often angled inward | Not needed: defined cut-off |
The practical trap is the mixed row. Someone tops up a header with “whatever was in the bin,” one even-edge tip lands in a row of tapered fans, and the line shows one repeating stripe at that position for months. The box label does not say which edge family the tip belongs to, so spec the edge type explicitly on every purchase order and keep a spare of exactly the same profile.
Counting Nozzles and Building the Flow Budget
With spacing decided, the nozzle count is just the belt width divided by the spacing, rounded up. A 1200 mm belt at 250 mm spacing needs five nozzles. Then the flow budget falls out: if each tip flows 4 L/min, the header draws 20 L/min. Multiply by the number of headers and you have the pump duty.
| Belt width | Spacing at 45% of 500 mm width | Nozzle count | Header flow at 4 L/min each |
|---|---|---|---|
| 600 mm | 225 mm | 3 | 12 L/min |
| 1200 mm | 225 mm | 6 | 24 L/min |
| 1500 mm | 225 mm | 7 | 28 L/min |
| 2400 mm | 225 mm | 11 | 44 L/min |
| 3000 mm | 225 mm | 14 | 56 L/min |
Counts are rounded up from belt width ÷ spacing; a header always needs enough tips for the full belt, even if the last nozzle runs slightly past the edge.
This is where layouts quietly fail in the field. A designer picks nozzles for the spray pattern and forgets to check that the existing pump can supply the total. Under-pressure the header and every nozzle in the row drifts off its rated flow, the patterns shrink, and the seams reopen. Size the pump to the summed flow at the working pressure, not to a single nozzle.
Two more budget items belong on the same sheet. The strainer mesh must be finer than the smallest orifice in the row, otherwise one chip plugs the smallest tip and the row loses its symmetry. And a dirty strainer can rob 0.5–1 bar from the header, several percent of flow on a square-root curve, so account for its pressure drop when you size the pump.
Matching Angle to the Job, Not the Habit
Narrow angles (15–25°) throw a thin, high-impact line, right for cutting, descaling or spot cleaning, wrong for covering a wide belt evenly. Wide angle fans (110–145°) lay a broad, gentle curtain that is perfect for low-pressure rinsing and dust knockdown but useless where you need to knock something off the surface.
For a belt-coverage duty, 65–80° flat spray nozzles are the usual home. They spread the film across a useful width without wasting liquid past the edges. If your current header uses 40° tips and shows dry seams, the fix is often switching to a wider angle and re-spacing, not adding more nozzles at the same angle.
| Duty | Angle family | Why |
|---|---|---|
| Descaling, deburring, cutting | 15–25° | Concentrated impact in a thin line |
| Spot washing, part rinsing | 25–40° | Targeted band at close range |
| Belt washing, rinsing | 65–80° | Good width, good impact per metre |
| Low-pressure rinse curtains | 110° | Wide coverage, gentle film |
| Dust knockdown, humidification | 145° | Maximum width at minimum impact |
Impact follows the same curve in reverse: double the angle at a fixed standoff and the same flow spreads over roughly twice the line, so impact per unit length drops by about half. If the wash must physically move contamination, check impact before widening the fan. The belt will be evenly wetted and still dirty.
A Worked Example: A 1500 mm Belt
Suppose a rinse belt is 1500 mm wide with a fixed 300 mm header standoff. Start from the coverage width each nozzle must deliver. An 80° flat spray nozzle covers about 500 mm at 300 mm, so spacing the row at roughly 45% of that width puts nozzles about 225 mm apart. The belt then needs seven tips (1500 ÷ 225, rounded up). At 4 L/min each, the header draws about 28 L/min.
Now push the same header to 500 mm standoff. The 80° tip now covers about 840 mm, spacing grows to roughly 380 mm, and the same belt needs only five nozzles. Same belt, same fluid, different standoff, and the nozzle count and pump duty both drop.
A Second Worked Example: Application Rate on a Coating Line
Belt coverage is only half the layout; the other half is how much liquid lands per square metre. Suppose a coating line runs a 1200 mm belt at 10 m/min and the process needs 0.5 L/m² of applied film. The applied area per minute is 1.2 m × 10 m = 12 m²/min, so the header must deliver about 6 L/min total: three nozzles at 2 L/min each, or six at 1 L/min each. The number of nozzles is set by the spacing math; the flow per nozzle is set by this application-rate budget. The two have to agree: pick the spacing first, then select tips whose flow at the working pressure sums to the rate the process demands.
One more check belongs in this budget. If the row is spaced for 45% overlap with 80° tips, the wet band is wider than the belt by roughly one overlap at each edge. That liquid is not wasted, but it is outside the counted area, so the real application rate on the belt is a little lower than the budget says. Trim the outer nozzles’ aim inward or add the edge loss to the pump duty.
Why a Cardboard Test Saves the Rebuild
Catalog widths are measured under ideal conditions: perpendicular aim, clean fluid, the quoted pressure. Your line is none of those. Before you drill the manifold or cut the pipe, build a throwaway rig, a length of rail, the actual nozzles, paper taped to a board at the real standoff, and fire it. Read the true width where the sheets land, and adjust spacing to what you see, not what the data sheet promised.
This ten-minute test catches three problems that otherwise surface after the header is welded in: aim that is a few degrees off perpendicular (which skews the effective width), pressure drop along a long manifold (end nozzles weaker than middle ones), and a pattern that is narrower than quoted because the fluid is more viscous than the test fluid the catalog used. Do the test twice if you can, once with water, once with the real fluid. The gap between the two readings is the drift the line will show when chemistry changes, and it tells you how much margin to build into the spacing.
Fluid Properties Change the Real Width
Catalog coverage widths assume water at the rated pressure. A more viscous fluid, a coating, a syrup, a slurry, forms a thicker sheet that can break up differently and land narrower than quoted. Surface tension and any entrained air shift droplet size too. This is exactly why the cardboard test matters: fire the actual fluid, not water, before you fix the spacing. A layout proven on water can drift a stripe once the real fluid flows, and the seam reopens for no reason you can see on the data sheet.
Manifold Pressure Drop on a Long Header
On a long pipe the nozzles nearest the inlet see higher pressure than those at the far end, so their flow and width run larger. Over a meter-long header that gradient can be enough to reopen seams at the tail. Size the feed so pressure drop along the manifold stays small, ideally under 3–5% of the working pressure, or feed from both ends on long rows. Otherwise every nozzle in the row is technically the right part, but the row still covers unevenly because the supply is not equal at each tip.
A Layout That Survives Commissioning
A header that covers the whole belt on day one still has to cover it in month six. Two things keep it honest: strainers sized to the smallest orifice, so one clogged tip does not break the row’s symmetry, and a spare-nozzle policy where every unit in a row is the same edge type and angle. When a tip is replaced with a different profile, the seam comes back.
Add a third habit: mark the header with its working standoff and nozzle spacing, right on the pipe with a paint pen. Maintenance crews shim, lower and re-aim headers constantly, and each small change moves the coverage in a way nobody logs; the marked drawing keeps a “minor adjustment” from silently undoing the whole layout.
Troubleshooting a Header That Won’t Cover
| Symptom | Likely cause | Check and fix |
|---|---|---|
| Pale stripes at regular intervals | Spacing too wide or overlap under 30% | Measure actual spacing against 40–50% of real width |
| One recurring dark line | Mixed edge types, or one tip oversized | Walk the row; confirm every tip is the same edge and flow |
| Weak coverage at the far end of the header | Manifold pressure drop | Gauge at the last tip vs the inlet; feed from both ends |
| Whole band narrower than spec | Standoff sagged, or pressure below rating | Re-measure orifice-to-belt distance; gauge at a tip |
| Band fine at start, fades mid-shift | Strainer loading up | Check strainer pressure drop; clean and log it |
| Streaks appear after a tip swap | Replacement was a different profile | Enforce same part number; keep spares per row |
The pattern of these failures is consistent: the nozzle is rarely the problem. Pressure, spacing, edge type and supply uniformity decide the coverage; the tip just executes the decision.
Frequently Asked Questions
How far apart should flat fan nozzles be on a header? For tapered-edge fans, 40–50% of the real coverage width at your standoff, a nozzle covering 500 mm goes roughly 200–250 mm from its neighbour. Even-edge fans go edge to edge.
What overlap gives the most even coverage? Around 45–50% for tapered fans. Below about 30% the seams reappear as pale bands; above about 60% you start double-dosing the middle.
How many nozzles do I need for a 2-metre belt? Divide the belt width by the spacing and round up. At 225 mm spacing a 2000 mm belt needs nine tips. The number changes with standoff and angle, so pin those first.
Does raising pump pressure fix dry stripes? No. Pressure raises flow and widens the fan slightly, but it does not change the overlap geometry. Dry stripes are a spacing or edge-type problem; fix the spacing, not the pump.
Can I mix tapered and even-edge nozzles on one header? No. Even-edge fans stop hard at their boundary, so one even-edge tip in a tapered row leaves a low-density seam at that position.
Why is the middle of the belt wetter than the edges? A tapered fan is thick in the middle by design. If the middle is visibly over-wet, the overlap is too large or the belt is narrower than the wet band. Trim the outer nozzles’ aim or reduce overlap.
How do I measure whether the header is even? Catch the liquid across the belt in identical tubes for a timed run and compare volumes. The coefficient of variation should sit in the single digits for a well-spaced tapered row.
What standoff should I design for? As short as the machine allows while still clearing the product. Shorter standoff means tighter pattern control and less drift sensitivity. The trade is more nozzles and more flow budget.
A Layout Checklist Before You Drill the Pipe
- Standoff pinned: measured orifice-to-belt, written on the drawing.
- Real width read: from the formula and confirmed on a cardboard rig, not from memory.
- Edge type chosen: tapered for blended rows, even-edge for single zones, one family per header.
- Spacing set: 40–50% of real width for tapered; edge to edge for even-edge.
- Count rounded up: belt width ÷ spacing, never rounded down.
- Flow budget summed: total header flow at working pressure, plus strainer drop.
- Pump checked: sized to the summed flow, not one nozzle.
- Strainers specified: finer than the smallest orifice in the row.
- Spares ordered: same part number, same edge, same angle, one per row.
If you are specifying a new line or reworking a streaky one, send us the belt width, standoff, fluid and target flow. We can sanity-check the angle, spacing and pump duty against the geometry above so the first build is the right one. For the wider picture on how the same overlap rules apply across pattern families, the spray nozzle selection guide is the next read, and the flow-rate calculation guide covers the pump-side math in detail. If the streaks are already on a coated product, why coating lines streak with flat fans walks the diagnostic from the product back to the header.
Next Step
Send the Duty. Get Sized Nozzles Back.
Send your flow, pressure, fluid and target coverage. We come back with nozzle options and figures, not a catalogue number.
Written by
Ray ChanIndustrial spray nozzle specialist. I size tank cleaning, atomizing, flat-fan and spiral nozzles against real duty conditions, flow, pressure, fluid and target, rather than catalogue numbers. Every guide here comes from actual sizing work.