Troubleshooting guide

Industrial Coatings

UV and conventional coatings on glass, metal, plastic, board, concrete and wood: what stops a line, and what to check first.

Uniform Industries · Greater Noida, India · ufindustries.com
Figures are typical for the class and carry their test method. The values that govern a supply are those on the Technical Data Sheet for the grade.

Before you start

How to use this guide

Work down the page. Cheap checks before expensive ones, and measure before you change anything.

Coating a part is not printing a web: the substrate arrived from someone else's process, so the first measurement of every shift is on the substrate, not on the coating.
Adhesion to glass, metal, ceramic and concrete is an interface property of your part plus our coating, so it is qualified on your own material, your own pre-treatment and your own part geometry, never from a data sheet.
A tape test at the end of the line proves nothing on an inorganic substrate; the test that counts is run after 24 h, and again after humidity or heat, because water is what breaks these interfaces.
Separate cure from adhesion before you change anything: confirm cure with solvent double rubs first, because at least half of what the floor calls an adhesion fault is an under-cured film being pulled off in one piece.
A bake is a schedule, not a dwell. On a conventional system the numbers that matter are peak metal temperature and time at that temperature, logged on the part with a data logger, not the set point on the oven controller.
Change one variable at a time and mark the parts, because on a spray or dip line the fault usually lives in one zone, one nozzle or one fixture and an across-the-board change hides it.
When a fault appears only on some parts, sort by rack position, by nozzle, by lamp and by substrate batch before you touch the coating.
Several checks in this guide put a person near a running lamp, a nitrogen blanketed tunnel or an open solvent. Guarding, extraction and the stated protection come first, and no measurement in this document is worth defeating an interlock for.

Industrial CoatingsHow to use this guide

The process

What happens, in order, and the one variable held at each step.

The process

What is controlled at each step

Industrial CoatingsThe process

Control variables

What to hold, and how to measure it

Variable	Target	How it is measured	How it drifts
Substrate wetting and surface energy	Glass and ceramic: a continuous water break free film for 30 s after rinsing. Plastics: 38 mN/m minimum, 42 to 46 mN/m for hard coats on PC and PMMA. Metal: no water break on the degreased panel.	Water break test on glass and metal. On plastics, wetting tension with test inks or dyne pens applied to the part about to be coated, not to a retained sample; state which system is used, because pens and inks do not read alike. Cite ASTM D2578 only where it applies, which is polyethylene and polypropylene film; on rigid mouldings in PC, PMMA and ABS treat the ink reading as a floor check and confirm with contact angle to ASTM D7334 when a number has to go into a specification.	Falls through a shift as degreaser bath loading rises and as treated plastic parts sit in stores. Corona and flame treatment decay over days; slip and mould release additives bloom back to the surface within a week. The first parts of the morning are usually the best parts of the day.
Steel preparation: blast grade, profile and soluble salts	Blast grade to the specification for the system, typically Sa 2.5 to ISO 8501-1 for an industrial build, with the profile inside the grade's stated range measured to ISO 8503. Soluble salt level to the limit agreed for the service condition; a few tens of mg/m2 of chloride is already too much for immersion or condensing service.	Blast grade judged against the ISO 8501-1 pictorial standards in the booth light. Profile by comparator or replica tape to ISO 8503, recorded per batch. Soluble salts by Bresle patch to ISO 8502-6 with conductivity to ISO 8502-9, at least three patches per batch and at least one on a weld or a corner where washing is worst.	Salt comes back with every wet or coastal delivery, with re-used abrasive and with handling. A blasted surface flash rusts within hours at high humidity, and abrasive that is recycled without separation carries salt from one job into the next.
Application viscosity	Spray 16 to 25 s, dip and flow coat 16 to 30 s set on the part, curtain 25 to 45 s, roller coat 30 to 60 s and often higher on a thixotropic grade, all on a 4 mm cup at the stated temperature. Flow coat floods the part and relies on the excess draining off, so it runs near dip viscosity, not at roller viscosity. Dip work is set on the part, not on the cup, then logged as a cup reading.	State the cup and the standard separately. The house cup is a DIN 4 mm cup to DIN 53211, sample and cup both at 25 +/- 1 C, calibration checked monthly. An ISO flow cup with a 4 mm jet is to ISO 2431 and does not give the same number, and ISO 2431 specifies 23 +/- 0.5 C, so where the house condition is 25 C record it as an agreed deviation; efflux time moves roughly 2 to 3 % per degree. Ford cups are to ASTM D1200 and Zahn cups to ASTM D4212. A target set on one cup is re-established on the other, never converted on the floor.	Rises over a shift as solvent flashes from an open pan and as the circulation pump puts work into the coating. A UV coating shear thins in the pump and reads low while running, then reads high when sampled cold. Sample from the same point every time.
Coating and part temperature	Coating 25 +/- 2 C at the applicator. Part within 5 C of the coating.	Contact or infra-red thermometer on the coating in the pan and on the part as it enters the applicator. Log both at the start, middle and end of a run.	Climbs steadily on a circulating system: pump work, lamp infra-red radiating back and hall temperature all add. A cold part out of stores in winter pulls viscosity up locally and gives orange peel on the first hour's work only.
Film weight or dry film thickness	Glass and plastic hard coats 3 to 8 micron dry. Metal decorative coatings 8 to 25 micron dry. Board coating 2.0 to 3.5 g/m2. Wood sealer 8 to 20 g/m2 and topcoat 6 to 12 g/m2 per pass. Concrete 60 to 200 micron per coat.	Wet film comb on the wet film, dry film by magnetic or eddy current gauge to ISO 2808 on metal, and by weighing a cut area of coated against uncoated stock on board. Calibrate the gauge on the same substrate class.	Climbs as viscosity climbs and as roller nip or curtain slot settings move under heat. Falls as a spray nozzle wears open and the pattern widens. Edges, radii, holes and weld crowns always carry less than the flat face, because surface tension pulls the film back from an edge while it is still liquid, and that is where corrosion and chemical attack start.
UV dose at the part surface	Indicative for this class: 200 to 500 mJ/cm2 for a clear industrial coating, 400 to 800 mJ/cm2 for a pigmented or heavy film, measured in the band the lamp actually emits.	Band matched radiometer carried through the line at running speed, on a fixture that puts the head where the worst part face sits. Match the head to the source: a UV-A head for a mercury arc or a 385 nm array, and a UV-V head, roughly 395 to 445 nm, for 395 and 405 nm arrays, because the UV-A band ends at 400 nm and under-reports them. A broadband head calibrated for a mercury arc will not report what an LED array delivers. Record the instrument, the head band and its calibration date with every trace.	Falls with lamp hours, dirty quartz, a tarnished reflector and a dusty LED window, long before anything looks different. On a three dimensional part the dose varies by a factor of three between the face and the side, and that variation is the real process window.
Oven cure: peak metal temperature and time at temperature	To the grade's schedule, quoted as peak metal temperature and minutes at that temperature, not as an oven set point and a conveyor time. A typical industrial stoving schedule sits in the range 140 to 200 C peak metal temperature for 10 to 20 min at temperature; the grade's own data sheet governs.	Oven profile with a multi-channel data logger travelling with the part, thermocouples attached to the heaviest section, the lightest section and a shadowed face. Record the ramp, the peak and the time above the stated temperature. Profile again after any change to loading, line speed, burner or extraction, and at least twice a year.	Moves with load density, with part mass, with extraction balance and with burner condition. A heavy rack shields its own centre, so the part in the middle of the load is the one that under-bakes while the controller reports the set point it was given.

Industrial CoatingsControl variables

Control variables

What to hold, and how to measure it

Variable	Target	How it is measured	How it drifts
Two-pack ratio and pot life	Hardener or silane addition to the ratio on the label, weighed. Pot life is grade specific and is taken from the product data sheet, commonly 2 to 8 h at 25 C, with hydrolysed silane primers at the short end and 2K isocyanates at the long end. Assume it roughly halves for every 10 C rise: a mix with 8 h at 25 C has about 4 h at 35 C and about 2 h at 45 C.	Weigh both components on a bench balance, do not use a ladle or a mark on a bucket. Take pot life from the data sheet and from the clock: write the mix time on the pot in marker and discard at the stated time whether or not the material still looks usable. Viscosity rise is a late and unreliable end point, and on a silane system it never arrives, because the mix can lose all of its reactive silane with no measurable change in cup time. Where a viscosity criterion is used at all on an isocyanate system, take it as a doubling of the initial efflux time and treat it as a back-stop, never as the control.	The reaction runs whether or not the line is running. Mixed material used beyond pot life looks and coats normally and never develops adhesion, which is why this fault surfaces at the customer rather than on the line.
Relative humidity, dew point margin and the coating hall	45 to 65 % RH at 25 C for silane primed glass and ceramic work. Below 70 % RH for solvent and water borne systems in the flash zone. In every case the substrate temperature is at least 3 C above the dew point at the time of application and until the film is through the flash zone.	Calibrated hygrometer at the line, not at the office wall, with a surface thermometer on the part, and the dew point margin taken by the ISO 8502-4 procedure. Log with the shift record.	Silane hydrolysis needs ambient water; below about 35 % RH the bridge does not form in the time available and adhesion is short. Above 75 % RH a solvent coating blushes and a water borne coating will not flash. A cold part out of winter stores can sit below the dew point at a hall RH of 55 %, so hold cold parts to hall temperature before coating. Monsoon and winter behave as different processes.
Line speed and flash time	Set so the film levels and the carrier is out before the lamp or the oven. Typically 30 to 90 s flash for a solvent reduced UV coating; none needed for a 100 % solids system; longer again on a conventional solvent coating going to a bake.	Time a marked part from applicator to lamp or oven with a stopwatch. Confirm the carrier is out by weighing a test panel before the cure stage.	Speeding the line cuts flash time and dose together, and the two faults look different on the part: retained solvent shows as blisters, pop and soft film, low dose as a tacky surface.
Circulation system cleanliness and light exclusion	No stray daylight or lamp light on the pan, the hoses or the header tank. Feed filtered to a rating set by the finish: typically 25 to 50 micron absolute for spray and thin hard coats, 50 to 100 micron for roller, curtain and heavy builds, with nominal or absolute stated on the specification. A coarser strainer on the return as well.	Find stray light in two safe passes. First, with the lamps off and the hall lights off, look for daylight ingress on the pan, the hoses and the tank, from roof lights, doors and translucent panels. Second, with the lamps on and all guarding, shutters and extraction in their normal running state, do not look at the lamp house: lay UV sensitive indicator cards or a photodiode probe on the pan lid, the hoses and the tank top and read them afterwards. Anyone inside the line while lamps run wears a UV face shield and UV blocking eyewear with skin covered, and the ozone extraction is confirmed running first. Check both filter baskets each shift and keep the removed material.	Gel particles build slowly, then appear all at once as seeds and craters. A rising catch of soft gel in the return strainer is the earliest warning you get. A UV coating that has gelled cannot be thinned back; it is discarded.

Industrial CoatingsControl variables

The mechanism

Why a coating lets go of glass and metal

The most misunderstood point in this category is that adhesion to an inorganic substrate is a wet interface problem, not a dry one. A coating held by physical forces alone passes every test you can run on day one and fails as soon as water reaches the interface, which is why the silane bridge and the humidity test both exist. Read the four panels left to right: the surface you are actually coating, what happens with physical adhesion only, what the silane bridge changes, and what the test schedule therefore has to be.

Industrial CoatingsThe mechanism

02 | 20 faults

Faults

Symptom, cause, check, change, confirm. The trap is the move to avoid.

Fault 1 of 20

Cissing, fisheyes and crawling in the wet film

WHAT YOU SEE

Small round craters with a raised rim, often with a speck at the centre, or the coating pulling back into islands and leaving bare substrate. It is usually worse on one rack position, one fixture or one side of the part.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Silicone or mould release on the substrateAt that spot the substrate's surface energy is below the surface tension of the wet coating, so the film cannot wet it: the liquid retracts into the surrounding area and drags a rim up with it, and the crater is frozen when the lamp or the oven fixes the film.
Compressed air carrying oil or water into the spray lineOil droplets land in the wet film and act as thousands of small release spots, which is why the pattern follows the spray fan.
Hand contact, skin oil and glove residue during loadingFingerprint oil is enough on glass and polished metal, and it survives a wipe with a dry cloth.
Silicone slip additive carried over from another coating in the same hall or the same circulation systemSilicone migrates readily; a few parts per million in the pan will crater a whole batch.
Degreaser bath loaded past its capacity, or condensation on a cold partA saturated bath leaves parts with a thin film of what it was meant to remove, and a water film breaks the coating in the same way an oil film does.

Industrial CoatingsCissing, fisheyes and crawling in the wet film

Fault 1 of 20 · continued

Cissing, fisheyes and crawling in the wet film

CHECK, IN THIS ORDER

Water break test on the cleaned substrate before any coating is applied: a continuous water film that holds 30 s is a pass, beading is a fail.
Blow the compressed air line onto a clean mirror or a sheet of white board for 30 s and look for oil mist; check the coalescing filter and the dryer dewpoint.
Sort the failed parts by rack position, fixture and nozzle, and record whether the craters follow the spray fan or the part geometry.
Wipe a test area with isopropanol on a lint free cloth and recoat by hand: if that area is clean, the fault is substrate contamination, not coating.
Check whether any silicone bearing product, slip additive or release spray is used anywhere in the same hall.

THEN CHANGE

Fix the cleaning stage before touching the coating: fresh degreaser, correct bath temperature, a proper rinse and a dry part.
Service the compressed air: coalescing filter, drain the receiver, confirm the dryer, and take the spray air from a dedicated line.
Introduce gloves at load and stop bare hand contact after the wash: clean cotton or nitrile for handling dry, clean, uncoated substrate, and nitrile only, changed the moment it is contaminated and never re-used, for anything that touches uncured coating. Uncured acrylate monomer is a skin irritant and a sensitiser, and sensitisation is permanent, so it is never handled bare handed and never in cotton, which holds it against the skin. Wash with soap and water, not with solvent.
Quarantine the circulation system if silicone carry-over is suspected, flush it, and change filters and hoses rather than chasing it with additive.
Only as a last resort, and with our agreement, add a flow and levelling additive: it hides small contamination and makes recoating and overcoating harder.

CONFIRM IT IS FIXED

Twenty consecutive parts through the full line with no crater visible at the inspection station: at least 1000 lux at the part, 1000 to 2000 lux being the working range, with a grazing low angle or specular light, the part turned through the reflection, at a fixed viewing distance and at the same station every time. Plus a water break free result on the cleaned substrate on the same shift. If a flow additive was used, confirm overcoat adhesion by ISO 2409 cross-cut on a topcoated part.

THE TRAP

Adding more flow additive or more solvent to make the coating level. It closes the craters on the next few parts and puts a low surface energy layer through the whole batch, which then fails intercoat adhesion under the topcoat and craters worse once the additive is consumed.

Industrial CoatingsCissing, fisheyes and crawling in the wet film (continued)

Fault 2 of 20

Coating lifts from glass after humidity or a wash

WHAT YOU SEE

The part passes a tape test at the end of the line and fails days later, after a humidity cabinet, a dishwasher cycle or a warm store. The coating comes away cleanly, with no coating left on the glass.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

No silane bridge, so the coating is held by physical adhesion onlyA clean glass surface carries silanol groups and an adsorbed water layer; organic binders cannot bond to it covalently, and water diffusing through the film hydrates the interface and lifts the coating. This is the mechanism the whole fault rests on, and it is the first thing to establish on a system that has never passed a humidity test.
Two-pack or primer used beyond its pot lifeThe silane has already hydrolysed and condensed with itself in the pot, so nothing is left to react with the glass. The coating applies and cures normally and never develops adhesion. On a line that was passing last week and is failing this week, look here first.
Glass not free of oils, cutting fluid or packing interleaf residueAny organic layer sits between the silane and the silanol groups it has to reach, and it arrives with a new glass batch or a tired wash bath rather than with the coating.
Hall humidity too low for the silane to hydrolyseHydrolysis of the alkoxy groups needs ambient water; below about 35 % RH the bridge does not form in the time the part spends between primer and cure.
Tin side of float glass coated instead of the air sideThe tin bearing face has a different surface chemistry and gives lower and much more variable adhesion after humidity.

Industrial CoatingsCoating lifts from glass after humidity or a wash

Fault 2 of 20 · continued

Coating lifts from glass after humidity or a wash

CHECK, IN THIS ORDER

Read the mix time on the pot and compare with the clock. Anything past the stated pot life is discarded, not used up.
Water break test on the glass as it arrives, before the primer, and again after the wash.
Check hall RH with a calibrated hygrometer at the line over the whole shift, not as a spot reading.
Identify the tin side with a short wave UV lamp: the tin face fluoresces milky.
Run ISO 2409 cross-cut at 24 h, then repeat after 24 h conditioning to ISO 6270-2 at 40 C constant condensation, or in a cabinet at 40 C and 95 % RH recorded as the house condition, on the same batch.

THEN CHANGE

Move to a two-pack silane bonding route where the application demands wet adhesion. Select the primer or bonding grade with technical service against the glass type, the cleaning route and the service condition, and qualify it on the customer's own glass with the humidity or wash cycle that applies before any production run.
Mix only what will be used inside the pot life, weigh the components, and mark the pot with the mix time.
Clean the glass properly: alkaline wash, rinse with deionised water, dry, and coat inside the shortest practical window.
Bring hall RH into 45 to 65 % at 25 C, or hold the primed parts in a conditioned area before cure.
Coat the air side, and set the glass supply so the tin side is identified at goods inward rather than on the line.

CONFIRM IT IS FIXED

Cross-cut to ISO 2409 at class 0 or 1 after 24 h conditioning to ISO 6270-2 at 40 C, tested within 30 min of removal and again after 24 h recovery, both results recorded, on the customer's own glass, plus the customer's own wash or dishwasher cycle where that is the service condition. Adhesion on glass is a property of the customer's glass, its cleaning and this coating together, so it is qualified on their material every time the glass source changes.

THE TRAP

Testing adhesion at the end of the line, passing, and shipping. A physically adhering film always passes on day one. The test that means anything is run after the interface has seen water, and it is read while the film is still wet from the cabinet as well as after it has dried back, because a re-dried film can recover a pass it did not deserve.

Industrial CoatingsCoating lifts from glass after humidity or a wash (continued)

Fault 3 of 20

Surface stays tacky, film underneath is hard

WHAT YOU SEE

A thumb twist marks the surface, parts block or fingerprint in the stack, and dust sticks to the face while the body of the film is fully hard when scraped.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Oxygen inhibition at the air interfaceAtmospheric oxygen quenches free radicals at the surface faster than they are generated, leaving the top micron uncured while the bulk cross-links normally.
LED array with no short wave content used with a coating formulated for mercuryA 385 to 405 nm array emits nothing below 380 nm, so the surface cure mechanism the coating relies on is simply absent.
Loss of short wave output from an aged mercury lamp or a devitrified quartz sleeveThe short wave line that cures the top micron goes first, while a broadband meter still reads an acceptable total.
Low peak irradiance, usually from lamp to part distance on a three dimensional partA high peak drives radicals in faster than oxygen can quench them; the same total dose delivered slowly leaves a tacky face.
Very thin film on a flat faceA thin film has a high surface to volume ratio and is proportionally harder to surface cure, which is counter-intuitive on the floor.

Industrial CoatingsSurface stays tacky, film underneath is hard

Fault 3 of 20 · continued

Surface stays tacky, film underneath is hard

CHECK, IN THIS ORDER

Confirm from the tin and the data sheet that the coating grade is specified for the lamp you are running, mercury or LED. This costs nothing and settles a large share of cases.
Lamp hours and quartz condition on the unit over the tacky face; inspect the sleeve for a cloudy or frosted appearance and the reflector for tarnish.
Measure lamp to part distance on the face that is tacky, not on the face that is convenient.
ASTM D5402 MEK double rubs on the tacky face and on a good face for comparison, run under the extraction and with the protection set out in the methods section.
Radiometer pass at running speed with a band matched head, on a fixture that puts the head where the tacky face sits, recording peak irradiance as well as dose. This is the slowest and most expensive check in the list, so it is run last, not first.

THEN CHANGE

Raise peak irradiance first: close the lamp to part distance, clean the quartz and reflector, or fit a focused reflector.
Change the lamp or sleeve if short wave output has gone; an aged lamp is cheaper than the batch.
Add a short wave mercury unit as a final station where an LED line has to run a mercury coating, or move to the correct LED grade.
Increase film weight slightly where a very thin film is the cause, within the specified window.
Where the application justifies it, inertise the cure zone with nitrogen to below about 500 ppm residual oxygen at the film surface, which removes the mechanism and typically allows 30 to 50 % less dose. Nitrogen inerting is a confined space and asphyxiation control problem, not a process tweak: it is installed only with fixed oxygen depletion monitors and alarms at the tunnel, at floor level and in any pit, with interlocked extraction, a lock-out procedure before anyone enters the enclosure, and trained entry with a personal oxygen monitor. Do not fit nitrogen to an existing line without a written risk assessment.

CONFIRM IT IS FIXED

No mark from a firm thumb twist immediately off the line, 50 or more MEK double rubs to ASTM D5402 on the worst face, and the radiometer pass inside the specified dose and peak irradiance window on that same face.

THE TRAP

Slowing the line. More time at low irradiance gives oxygen more time to quench the surface, so the tack often gets worse while the operator watches the dose reading go up.

Industrial CoatingsSurface stays tacky, film underneath is hard (continued)

Fault 4 of 20

Under-cure at the bottom of a pigmented film

WHAT YOU SEE

The surface is hard and glossy but the coating can be lifted whole with a blade or a strip of tape, and the underside is soft or smells of monomer. Common on whites, heavy colours and anti-corrosion grades on steel.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Film weight above the window for the pigment loadingPigment absorbs and scatters the radiation, so the top of the film screens the bottom; doubling the film weight more than doubles the dose the base needs.
Wrong lamp spectrum for the pigmentRutile titanium dioxide absorbs strongly below about 400 nm, so an undoped mercury lamp or a 385 nm array cures the top of a white and starves the base.
Metal substrate acting as a heat sinkA cold, thick steel or aluminium part pulls heat out of the film and slows the propagation reaction near the interface, where the cure matters most.
Dose measured on a flat test panel rather than on the partSides, radii and shadowed faces of a three dimensional part can see a third of the dose the top face sees.
Insufficient dwell between coats on a two-layer systemThe base has not cured through before the topcoat screens it from the next lamp.

Industrial CoatingsUnder-cure at the bottom of a pigmented film

Fault 4 of 20 · continued

Under-cure at the bottom of a pigmented film

CHECK, IN THIS ORDER

Weigh a coated against an uncoated panel, or measure dry film thickness to ISO 2808, on the part that is failing.
Radiometer pass at the failing face, and again on the face that passes, so you have the ratio. Use a UV-V head on a 395 or 405 nm array, because a UV-A head under-reports it.
Cross-cut to ISO 2409 and then look at the underside of the removed film for softness or odour.
Measure part temperature entering the lamp with an infra-red thermometer on the first parts of the shift and mid-shift.
Confirm which lamp type is fitted: undoped mercury, iron doped, gallium doped, or LED and at what wavelength.

THEN CHANGE

Reduce film weight to the specified window and take the opacity from two thinner coats rather than one heavy one.
Move the pigmented station to a gallium doped lamp, or to 395 or 405 nm emitters on an LED line, which puts the energy above the titanium dioxide absorption edge.
Pre-heat heavy metal parts to 35 to 45 C before the lamp so the film is not quenched at the interface.
Add a lamp or reposition lamps to cover the shadowed faces, and re-measure rather than assume.
Slow the line for pigmented work specifically, where peak irradiance is already adequate and the shortfall is total dose at depth.

CONFIRM IT IS FIXED

Cross-cut to ISO 2409 at class 0 or 1 at 24 h with no whole film lift, 50 or more MEK double rubs to ASTM D5402, no monomer odour from the underside of a deliberately removed film, and a radiometer reading inside the window on the worst face.

THE TRAP

Reaching for more film weight when the colour looks weak. Opacity and cure pull in opposite directions on a pigmented coating; the heavier film hides the colour problem and creates an adhesion failure that arrives at the customer three weeks later.

Industrial CoatingsUnder-cure at the bottom of a pigmented film (continued)

Fault 5 of 20

Topcoat peels from the basecoat

WHAT YOU SEE

The two-layer system delaminates cleanly between the layers; the basecoat is still firmly attached to the part and the topcoat comes off as a sheet or in flakes at the cross-cut.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Basecoat over-cured before the topcoat went onA fully cross-linked acrylate surface has few residual double bonds and nothing for the topcoat to react into, so the second layer relies on physical grip on a smooth face.
Slip, wax or silicone additive in the basecoatAdditives migrate to the air interface as the film levels, and the topcoat is then applied to a release surface.
Time between coats too longSurface radicals and residual unsaturation are consumed within minutes to hours, and airborne contamination settles on the waiting parts. On a conventional 2K basecoat the same fault appears as the maximum recoat window being exceeded, after which the base has to be abraded before the topcoat.
Basecoat under-cured and swelling under the topcoatThe topcoat monomer dissolves into a soft base, the two shrink differently on cure, and the stress that generates opens the interface.
Contamination between stations, typically overspray dust or air line oilThe interlayer is being coated onto whatever settled on the part between the two applicators.

Industrial CoatingsTopcoat peels from the basecoat

Fault 5 of 20 · continued

Topcoat peels from the basecoat

CHECK, IN THIS ORDER

Measure the dose the basecoat receives at its own station; on an intercoat system it is deliberately partial.
Read the basecoat data sheet for whether it is designed as a matt base or a tack cure base, and for its minimum and maximum recoat window, and confirm the grade is a system pair.
Time the interval between coats with a stopwatch on a marked part; record best and worst.
Cross-cut to ISO 2409 at 24 h and look at the failure plane under a loupe: base to part, or base to topcoat.
Check for slip or wax additives in the basecoat, including anything added at the line to fix a blocking problem.

THEN CHANGE

Set the basecoat to a partial cure that leaves the surface ready to bond, and take the final cure from the lamp after the topcoat.
Shorten and fix the interval between coats, and keep the parts covered between stations.
Abrade the basecoat before the topcoat where a 2K base has gone past its maximum recoat window, and re-qualify the abraded route rather than assuming it.
Remove slip and wax additives from the basecoat and take the slip from the topcoat instead.
Use the matched pair of grades rather than mixing a basecoat from one system with a topcoat from another.

CONFIRM IT IS FIXED

Cross-cut to ISO 2409 at class 0 or 1 at 24 h on the full two-layer build, and, where the failure plane has to be established as a reportable result, pull-off to ASTM D4541 on a flat panel of the same substrate and the same build, with the failure mode recorded as cohesive in the basecoat, adhesive at the base to topcoat interface, or adhesive at the base to part interface. ISO 2409 classifies the area of flaking and does not report a failure plane, so the loupe observation is a guide and the pull-off is the record. Accept cohesive failure in the basecoat; reject failure at the interlayer. No delamination after the humidity or thermal cycle the job specifies.

THE TRAP

Raising the basecoat dose because the base feels soft. A harder base is a worse key, so the delamination gets cleaner and more complete while every individual layer measures well.

Industrial CoatingsTopcoat peels from the basecoat (continued)

Fault 6 of 20

Sags, runs and fat edges on vertical faces

WHAT YOU SEE

Curtains and tear drops on the vertical faces of a part, a thick lower edge, and a build up at the bottom of a dipped part. The flat faces look right.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Viscosity below the window at the applicatorOnce viscosity falls, gravity moves the film faster than levelling stops it, and the film flows for as long as it stays liquid.
Film weight above the windowSag rate rises steeply with thickness; the same coating that sits still at 15 micron runs at 30 micron.
Flash or dwell too long before the lampOn a UV system the film stops moving only when the lamp fixes it, so every extra second before the lamp is another second of flow.
Coating or part temperature above the windowViscosity falls with temperature; a coating at 35 C in a pan behaves as a much thinner material than the same coating at 25 C.
Spray technique or nozzle wear putting extra material on the lower part of the fanA worn nozzle widens and softens the pattern, so material is deposited beyond the intended band and doubles up on the overlaps.

Industrial CoatingsSags, runs and fat edges on vertical faces

Fault 6 of 20 · continued

Sags, runs and fat edges on vertical faces

CHECK, IN THIS ORDER

DIN 4 mm cup to DIN 53211 at 25 +/- 1 C on a sample taken at the applicator, at start, middle and end of the run.
Wet film comb on a vertical face and on a horizontal face of the same part.
Time the part from applicator to lamp or oven with a stopwatch.
Infra-red thermometer on the coating in the pan and on the part.
Inspect nozzles and check the fan pattern on a card; compare against a new nozzle.

THEN CHANGE

Bring viscosity back to the window by temperature control rather than by adding material, then correct film weight.
Shorten the dwell before the lamp, or move the first lamp closer to the applicator, so the film is fixed before it can run.
Cool the circulating coating: a small heat exchanger holds 25 +/- 2 C against pump work and lamp infra-red.
Replace worn nozzles and re-set the fan overlap; on a dip line, slow the withdrawal rate.
Reduce film weight to the specified window and take the property from the correct grade rather than from thickness.

CONFIRM IT IS FIXED

Twenty consecutive parts with no sag visible at the inspection station, at least 1000 lux at the part with a grazing low angle light, wet film within the specified band on both vertical and horizontal faces, and viscosity logged inside the window at the start and end of the run.

THE TRAP

Thickening the coating with an additive or with a different batch to stop the run. It stops the sag and starts orange peel and poor edge coverage, and it usually pushes film weight up again, which was the real cause.

Industrial CoatingsSags, runs and fat edges on vertical faces (continued)

Fault 7 of 20

Orange peel and ribbing from the roller coater

WHAT YOU SEE

A regular dimpled texture over the whole face, or fine parallel lines running in the machine direction, most visible on a gloss finish and under a low angle light.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Viscosity above the windowThe film splits between roller and part when the nip opens and then has too little time and too little mobility to level the split pattern out.
Cold parts or a cold hallViscosity at the part is what matters, not viscosity in the pan; a cold part raises it locally at exactly the moment levelling has to happen.
Roller speed ratio or nip pressure set wrongToo much differential tears the film as it splits and leaves ribbing; too little gives insufficient transfer and a starved, textured film.
Worn, scored or hardened rubber rollerThe nip is no longer uniform across the width, so the split pattern varies and prints as lines.
Insufficient levelling time before the lampOn a 100 % solids UV system the only levelling that happens is what happens before the lamp fixes the film.

Industrial CoatingsOrange peel and ribbing from the roller coater

Fault 7 of 20 · continued

Orange peel and ribbing from the roller coater

CHECK, IN THIS ORDER

DIN 4 mm cup to DIN 53211 at 25 +/- 1 C from the pan, and an infra-red thermometer reading on the incoming part.
Measure gloss to ISO 2813 at 60 degrees across the width at five points and compare; orange peel shows as a gloss drop and as loss of image sharpness in the reflection.
Check nip impression across the full width with pressure sensitive film or an ink strip.
Inspect the roller surface and measure durometer; compare with the specification for that machine.
Time the travel from the coater to the first lamp.

THEN CHANGE

Bring viscosity to the window with temperature control, not with thinner, and hold the coating at 25 +/- 2 C.
Warm cold parts to within 5 C of the coating before the coater.
Re-set the speed ratio and nip pressure to the machine's specification, one change at a time, with a marked part after each change.
Re-grind or replace the roller once the nip cannot be made uniform across the width.
Add levelling distance or slow the line so the film has time to flow out before the lamp.

CONFIRM IT IS FIXED

Gloss to ISO 2813 at 60 degrees within 3 GU across the width and within the grade's specification, and no texture visible at the inspection station under a grazing low angle light at 1000 lux or more at the part, on twenty consecutive parts.

THE TRAP

Thinning the coating with solvent or monomer to make it level. The texture improves and film weight, hardness and chemical resistance all fall with it, so the part then fails the MEK rub test that nobody re-ran.

Industrial CoatingsOrange peel and ribbing from the roller coater (continued)

Fault 8 of 20

Pinholes and boil on wood, MDF edge and concrete

WHAT YOU SEE

Small open holes or burst bubbles over the whole coated face, worst on machined MDF edges, open grain and porous concrete, and worse on the first coat than on later coats. On a non-porous part, see solvent pop instead: the cause is different.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Air and moisture driven out of the porous substrate as the coating warmsAir in the pores expands and vapour is driven off by lamp infra-red, and it escapes through a film that is already too viscous to close behind it.
Substrate moisture content too highWater leaving a warm substrate has far more volume to move than air does, and it lifts the film as it goes.
No sealer or an under-applied sealer coatThe sealer's job is to fill and close the pore structure so the topcoat sits on a continuous surface rather than bridging thousands of open holes.
Film weight too high in a single passA heavy film resists the escaping air for long enough that the bubble bursts and cannot close.
Infra-red load from a mercury lamp arriving too quicklyA sharp temperature rise expands the trapped air faster than the film can accommodate.

Industrial CoatingsPinholes and boil on wood, MDF edge and concrete

Fault 8 of 20 · continued

Pinholes and boil on wood, MDF edge and concrete

CHECK, IN THIS ORDER

Moisture meter on the wood or board: typically 6 to 10 % for interior panel work, taken at several points including the edge.
On concrete, ASTM D4263 plastic sheet test as a first pass, then in-situ relative humidity probes to ASTM F2170, with the number and placement of holes per the standard for the floor area and each probe left to equilibrate in the slab for 24 h before it is read.
Measure sealer film weight by weighing a cut area, and check coverage on a machined edge with a torch at a low angle.
Measure part temperature entering and leaving the lamp with an infra-red thermometer.
Coat a test panel at half the film weight and compare; if the pinholes halve, the cause is film weight or escape rate.

THEN CHANGE

Apply a proper sealer coat and cure it before the topcoat, and give the edges of MDF a second sealer pass.
Hold the substrate until its moisture content is inside the window; on concrete this is often weeks, not hours, and no coating change substitutes for it.
Split a heavy build into two lighter passes with a cure between.
Reduce the infra-red load: fit a dichroic reflector, add an air gap, or move to LED where the coating allows.
Condition the parts in the coating hall overnight so the substrate, the coating and the hall are at the same temperature.

CONFIRM IT IS FIXED

No pinholes visible at the inspection station under a grazing low angle light at 1000 lux or more at the part, on twenty consecutive parts including machined edges, substrate moisture logged inside the window before coating, and specified adhesion and hardness maintained on the two-pass build.

THE TRAP

Applying a heavier topcoat to fill the pinholes. The extra weight traps more air and makes the boil worse, and the heavier film then under-cures at its base as well.

Industrial CoatingsPinholes and boil on wood, MDF edge and concrete (continued)

Fault 9 of 20

Solvent pop and craters in an oven baked film

WHAT YOU SEE

Small open craters or burst bubbles over the face of a baked part, worst where the film is thickest and on the face that meets the burner or the radiant panel first. The substrate is not porous, the film below the crater is sound, and the same job runs clean when the oven is loaded lightly.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Bake ramp too steep at the front of the ovenThe surface skins before the solvent underneath has left, so the vapour has to burst through a film that has already set. An oven that hits the part with full heat in the first minute does this reliably.
Film weight above the wet film windowThe path the solvent has to travel is longer and the skin forms sooner, so sag, pop and under-bake at depth all arrive together above the window.
Flash time too short before the ovenMost of the carrier is meant to leave in the flash zone at hall temperature; whatever is left has to fight its way out through a heating film.
Thinner blend too fast for the scheduleA fast solvent leaves the surface early and raises the skin, and the tail of slower solvent then has to escape through it.
Outgassing from castings, weldments and porous metalAir and oil held in pores and in weld porosity expand in the oven and come through the film. It looks the same as solvent pop and has a different fix.

Industrial CoatingsSolvent pop and craters in an oven baked film

Fault 9 of 20 · continued

Solvent pop and craters in an oven baked film

CHECK, IN THIS ORDER

Run an oven profile with a data logger travelling on the part, not in the air: record the ramp, the peak metal temperature and the time at temperature, and look hard at the first two minutes.
Wet film comb at the applicator, and dry film to ISO 2808 on a crater area and on a clean area of the same part.
Time the part from applicator to oven with a stopwatch, and weigh a test panel before the oven to see how much carrier is still in the film.
Coat and bake a panel at half the film weight on the same schedule; if the craters halve, it is film weight or ramp rate, not the coating.
On castings and weldments, bake an uncoated part first, then coat and bake it, and compare.

THEN CHANGE

Put a low temperature zone or a longer flash at the front of the oven so the carrier leaves before the film skins.
Bring film weight into the window and take build or opacity from two coats with a bake between.
Lengthen the flash at hall temperature and confirm it by weighing a panel rather than by eye.
Use the thinner blend specified for the schedule and the hall condition, and change it with the season rather than once a year.
Pre-bake castings and weldments to drive out entrained air and oil before coating.

CONFIRM IT IS FIXED

No craters visible at the inspection station at 1000 lux or more at the part on twenty consecutive parts, an oven profile logged on the part with peak metal temperature and time at temperature inside the grade's window on the lightest, heaviest and most shadowed positions in the load, and film weight to ISO 2808 inside specification on the same parts.

THE TRAP

Turning the oven up to drive the solvent out faster. A hotter front zone skins the film sooner, so the pop gets worse while the recorded bake looks better, and the extra heat at peak embrittles the film that did survive.

Industrial CoatingsSolvent pop and craters in an oven baked film (continued)

Fault 10 of 20

Curtain breaks or shows streaks and lines

WHAT YOU SEE

The falling curtain tears open, or runs with fine vertical lines that print as machine direction stripes on every part until the curtain is reformed.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

A particle or dried coating on the slot lipAnything on the lip disturbs the flow locally, and the disturbance runs the full height of the curtain and onto every part.
Air entrainment or draught across the curtainA curtain is a thin free liquid sheet; a door opening or an extraction fan can break it.
Flow rate or viscosity outside the windowBelow a minimum flow the curtain cannot support itself and tears at the edges first; the edge guides are where it always starts.
Edge guides dirty, wet or misalignedThe guides anchor the curtain; if the coating does not wet them cleanly the curtain pulls inwards and breaks.
Foam or entrained air in the circulation systemBubbles arriving at the slot open the curtain momentarily and leave a streak on whatever is underneath at that moment.

Industrial CoatingsCurtain breaks or shows streaks and lines

Fault 10 of 20 · continued

Curtain breaks or shows streaks and lines

CHECK, IN THIS ORDER

Stop and inspect the slot lip under a torch, wiping with a clean solvent cloth along its length.
DIN 4 mm cup to DIN 53211 at 25 +/- 1 C on a sample taken from the header tank.
Check the pump return for foam and the tank level for a vortex drawing air in.
Check the edge guides for cleanliness and alignment, and watch where the curtain first tears.
Look for a draught: a strip of tissue held near the curtain shows air movement that the operator cannot feel.

THEN CHANGE

Clean the slot lip and the edge guides at every stop, as a routine, not as a repair.
Screen the curtain from draughts and re-balance hall extraction.
Bring flow rate and viscosity back into the window together, and hold the coating at 25 +/- 2 C.
Lower the return line below the tank surface and reduce free fall to stop air entrainment; fit a de-aeration section if foaming persists, and re-aerate the material before it is stored.
Fit or clean the feed filter at the rating set for the finish, typically 50 to 100 micron for curtain work, with a coarser strainer on the return, and keep what they collect, because it tells you whether you have gel, dirt or dried coating.

CONFIRM IT IS FIXED

A stable curtain for the full length of a run with no tear, no vertical line visible on the coated face at the inspection station under a grazing low angle light at 1000 lux or more at the part, and film weight within specification across the full width measured at five points.

THE TRAP

Opening the slot or raising the flow to hold the curtain together. It covers the fault for a while, puts film weight up, and turns a streak problem into a sag and under-cure problem at the same time.

Industrial CoatingsCurtain breaks or shows streaks and lines (continued)

Fault 11 of 20

Dry spray, overspray dust and a gritty matt face

WHAT YOU SEE

A sandy or matt patch on an otherwise glossy part, usually at the edges of the fan, on the far side of the part from the gun and on the parts nearest the extraction. It feels rough under the hand and gloss reads low; the loose version lifts with a fingernail, the trapped version does not.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Gun to part distance too longThe droplet loses solvent in flight and arrives too viscous to flow into the film, so it sits on the surface as a dry particle.
Thinner blend too fast for the hall temperature and the booth air speedThe same gun distance that works in winter over-dries the droplet in a hot hall or in a high velocity booth.
Atomising pressure too high or the wrong tip fittedFiner droplets have more surface area and lose solvent faster, and they follow the booth air out of the fan instead of landing.
Booth air pattern carrying overspray back onto wet partsOverspray that has already dried in the air lands in a wet film and is trapped in it, which is the version no amount of wiping removes.
Racking too tight, so parts shadow each otherThe shadowed part sees only the edge of the fan, which is where the dry material is, and it also collects its neighbour's overspray.

Industrial CoatingsDry spray, overspray dust and a gritty matt face

Fault 11 of 20 · continued

Dry spray, overspray dust and a gritty matt face

CHECK, IN THIS ORDER

Measure gun to part distance on the failing face and compare with the booth standard; mark the gun position on the rail on a manual line.
Read hall temperature and air velocity at the part, and check the extraction balance with a tissue strip or a smoke test.
Spray a card at the standard distance and at a shorter distance and compare the two under a grazing light.
Gloss to ISO 2813 at 60 degrees on the rough patch and on a good area of the same part.
Check tip size, tip wear and atomising pressure against the setting sheet.

THEN CHANGE

Bring the gun to the standard distance and hold it there. On a manual line this is a training and a fixturing problem, not a coating problem.
Move to the slower thinner blend specified for the hall condition, with our agreement, rather than adding retarder at the gun by eye.
Drop atomising pressure to the lowest setting that still gives an even pattern, and fit the correct tip.
Re-balance the booth so overspray goes to the extraction and not across wet parts, and open up the rack spacing.
Remove dry spray before recoating rather than coating over it, because the loose material carries straight through into the next layer.

CONFIRM IT IS FIXED

No rough or matt deposit felt by hand or visible at the inspection station at 1000 lux or more at the part on twenty consecutive parts, gloss to ISO 2813 at 60 degrees within 3 GU of the retained standard at five points, and gun distance, tip and pressure recorded on the setting sheet.

THE TRAP

Adding retarder at the gun to stop the dry spray. It cures the roughness for an hour and puts sags, soft film and slow through-cure into the same batch, and nobody records how much went in.

Industrial CoatingsDry spray, overspray dust and a gritty matt face (continued)

Fault 12 of 20

Film cracks on forming, flexing or thermal cycling

WHAT YOU SEE

Fine cracks along a bend, at a swage or around a fixing hole, or a crazed pattern appearing after the part has been through hot and cold cycles. The coating itself is hard and passes every resistance test.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Over-crosslinked, brittle film from a high functionality gradeChemical resistance and hardness come from crosslink density, and so does brittleness; the two are traded deliberately, not obtained together.
Film weight above the windowBending strain at the coating surface rises with thickness, so a heavier film sees more strain for the same bend radius.
Cure shrinkage stored as interfacial stressAn acrylate film shrinks on cure while the metal or glass does not, and that locked-in stress adds to whatever the service condition applies.
Coating cured on a cold partA film cured on a cold part vitrifies early at low conversion, so the cure shrinkage is locked in as stress instead of relaxing out, and the under-converted network is less able to redistribute strain at a bend. The differential expansion term works the other way on warming, because the coating expands three to eight times as much as the steel or glass under it, so on the warm leg of a cycle the film is in compression; it is the frozen-in cure stress and the low conversion that matter here, and the cold leg of the cycle that puts the film into tension.
Part formed too soon after cureProperty development continues after the lamp; a film formed within the hour is being bent before it has finished.

Industrial CoatingsFilm cracks on forming, flexing or thermal cycling

Fault 12 of 20 · continued

Film cracks on forming, flexing or thermal cycling

CHECK, IN THIS ORDER

Mandrel bend or cupping test on a coated test panel of the same substrate and thickness, over the radius the part actually sees.
Dry film thickness to ISO 2808 at and near the bend.
Thermal cycle a sample through the service range, typically minus 20 C to 60 C for ten cycles, then inspect at 10x.
Compare a part formed at 1 h after cure with one formed at 24 h after cure.
Check part temperature at the lamp with an infra-red thermometer, and, on a conventional system, check the oven profile for an over-bake at peak, which embrittles the film in the same way.

THEN CHANGE

Move to a more flexible grade and accept the chemical resistance that comes with it; state which property the job actually needs.
Reduce film weight to the lower end of the window on parts that will be formed.
Hold parts 24 h after cure before forming, and mark the batch with the cure time.
Pre-heat parts to 35 to 45 C before the lamp so the network forms while it is still mobile, which raises conversion and lets the shrinkage stress relax out instead of freezing in. Do not pre-heat further in the belief that it matches the expansion of the part: a cured acrylate has three to eight times the expansion coefficient of steel or glass, roughly 60 to 100 ppm/K against 12 ppm/K for steel and 9 ppm/K for glass, so a film cured hot and then cooled ends up in tension. Verify any change on the mandrel bend rather than assuming it.
Increase the bend radius or the tooling radius where the design allows.

CONFIRM IT IS FIXED

No cracking visible at 10x after the mandrel bend at the service radius, after ten thermal cycles across the service range, and cross-cut to ISO 2409 at class 0 or 1 on the bent area 24 h later. Formability is a property of the coating and the customer's substrate and forming operation together, and it is qualified on their own parts.

THE TRAP

Curing harder to make the coating tougher. A harder film is a more brittle film; increasing dose on a cracking part reliably makes the cracking worse while every hardness reading improves.

Industrial CoatingsFilm cracks on forming, flexing or thermal cycling (continued)

Fault 13 of 20

Gloss and matt level drift across a shift

WHAT YOU SEE

The first parts of the day and the last parts do not match under the light box, or one rack position is consistently duller. Matt grades drift more than gloss.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Matting agent settling in the pan or the header tankSilica matting agents settle out of a coating standing between runs, so the first parts run glossy and the level moves as the tank homogenises.
Film weight driftGloss rises with film weight on a gloss grade and falls on a matt grade, so any drift in metering moves both.
Dose drift changing surface micro-roughnessMatting depends on the surface wrinkling as it cures; a change in dose or in lamp spectrum changes how much wrinkle develops.
Coating temperature drift changing flow-outA warmer, thinner coating levels more and reads higher gloss on a gloss grade.
Substrate absorbency variation between batchesA more absorbent board, wood or concrete pulls binder in and drops gloss without anything on the line changing.

Industrial CoatingsGloss and matt level drift across a shift

Fault 13 of 20 · continued

Gloss and matt level drift across a shift

CHECK, IN THIS ORDER

Gloss to ISO 2813 at 60 degrees, five points per part, on the first, middle and last parts of the run; use 20 degrees for high gloss and 85 degrees for matt.
Check the agitator is running and reaching the bottom of the tank; sample from the top and the bottom and compare gloss on drawdowns.
Weigh a coated against an uncoated panel for film weight at the same three points in the run.
Radiometer pass at start and end of the run.
Log substrate batch numbers, and drum or tote numbers, against the gloss record.

THEN CHANGE

Keep matt grades under continuous gentle agitation and re-homogenise the tank before every start.
Stir a settled drum or tote to the bottom before it is decanted, because pigment and matting agent pack hardest at the base and material taken off the top runs light in colour and gloss.
Lock film weight by fixing viscosity and temperature, and re-check after any nip or nozzle change.
Set a single reference standard part for the job, keep it at the line, and judge under the same light box every time.
Re-qualify gloss when the substrate supplier or grade changes, rather than adjusting the coating to chase it.

CONFIRM IT IS FIXED

Gloss to ISO 2813 within 3 GU of the reference standard at 60 degrees, at five points per part, on the first, middle and last parts of the run, with film weight logged inside the window at the same three points. Where 20 degrees or 85 degrees is the reporting geometry, widen the band to about 5 GU to match the method's reproducibility and state the geometry with the figure.

THE TRAP

Adjusting film weight to bring the gloss back. It corrects the reading and moves hardness, chemical resistance and cure through depth with it, so the batch that matches the standard is the batch that fails the resistance test.

Industrial CoatingsGloss and matt level drift across a shift (continued)

Fault 14 of 20

Two-pack mix loses adhesion before it is used

WHAT YOU SEE

The coating applies, flows and cures normally and the parts look correct, but cross-cut adhesion after humidity is poor on the afternoon's work and good on the morning's.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Pot life exceededThe hardener or silane has reacted in the pot rather than at the interface, so nothing reactive remains to bond to the substrate. The coating still films and cures because the base resin is unaffected, and the cup reading may not have moved at all.
Pot life read from the label without correcting for hall temperatureReaction rate roughly doubles for every 10 C, so a mix with 8 h at 25 C has about 4 h at 35 C. On a hot afternoon the material is dead well before the printed time.
Mix ratio estimated rather than weighedUnder-dosing starves the crosslink; over-dosing leaves free hardener or silane that migrates to the surface and reduces gloss and overcoat adhesion.
Fresh coating topped up into an old mixThe pot is now a blend of two ages with no usable pot life, and the clock on the label is meaningless.
Alcohol or water contamination in a system with an isocyanate hardener, or a hardener stored openHydroxyl groups react with the isocyanate and consume it before it can crosslink; the source is usually a cleaning solvent, a wet line or a tin left open in a humid store.

Industrial CoatingsTwo-pack mix loses adhesion before it is used

Fault 14 of 20 · continued

Two-pack mix loses adhesion before it is used

CHECK, IN THIS ORDER

Read the mix time written on the pot and compare with the clock, and read the hall temperature at the same time. If no mix time is written, that is the finding.
Weigh both components on a bench balance for the next mix and compare against what the operator normally does.
Check the cleaning solvent on the line: alcohol anywhere near a two-pack isocyanate system is a finding.
Check hardener tins for date opened, seal condition and any crystallisation or cloudiness.
Run ISO 2409 cross-cut at 24 h and again after 24 h conditioning to ISO 6270-2 at 40 C, on parts from the start and from the end of the pot.

THEN CHANGE

Mix only what will be used inside the pot life, in small pots, and write the mix time on the pot in marker. Discard at the stated time whether or not the material still looks usable.
Shorten the working pot life for the hall temperature: halve it for every 10 C above the temperature the data sheet quotes.
Weigh both components every time, and put a balance at the mixing station.
Never top up an old mix; discard what is left and mix fresh. Hardened material is not returned to stock.
Change line cleaning to an ester or the supplied reducer on isocyanate systems, remove alcohol from that area, reseal hardener immediately and rotate opened tins first in first out.

CONFIRM IT IS FIXED

Cross-cut to ISO 2409 at class 0 or 1 after 24 h conditioning to ISO 6270-2 at 40 C, tested within 30 min of removal and again after 24 h recovery, on parts taken from the start and from the last hour of each pot, with the mix time, the hall temperature and the weighed ratio in the shift record.

THE TRAP

Adding a little more hardener to an old pot to wake it up. The pot life has already been consumed by the reaction, and the extra hardener only adds free material that blooms to the surface and takes overcoat adhesion with it.

Industrial CoatingsTwo-pack mix loses adhesion before it is used (continued)

Fault 15 of 20

Viscosity climbing in the circulation system

WHAT YOU SEE

The cup reading rises hour by hour, film weight creeps up with it, and eventually seeds and soft gel lumps appear in the filter or as craters on the part.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Stray light reaching the coating pathA UV coating polymerises wherever photons reach it, and daylight through a roof light or a leaking lamp reflector is enough over a shift.
Temperature rise from pump work and lamp infra-redThermal initiation is slow but real, and the higher the circulating temperature the faster molecular weight builds.
Solvent flashing from an open pan on a reduced systemThe carrier leaves and the solids rise, which reads as a viscosity climb but is a different problem with a different fix.
Stabiliser headroom gone in returned material, or the oxygen it needs removedCoating that has been round the loop many times and topped up rather than replaced loses stabiliser headroom. The phenolic inhibitors used in UV coatings need dissolved oxygen to function at all, so nitrogen blanketing a tank, vacuum de-aerating the material or brim-filling a drum removes the mechanism and can gel a tank overnight.
Contamination from another chemistry in the same loopAmine, water or a residual hardener from a previous job will react with the coating and build viscosity.

Industrial CoatingsViscosity climbing in the circulation system

Fault 15 of 20 · continued

Viscosity climbing in the circulation system

CHECK, IN THIS ORDER

Look for stray light in the two safe passes set out in the controls: daylight ingress with the lamps off, then indicator cards or a probe left on the pan, hoses and tank with the lamps running normally and all guarding in place. Nobody stands in a darkened hall looking at a running lamp.
Log coating temperature at the pan hourly with a contact thermometer alongside the cup reading.
Compare a fresh sample from a sealed drum with a sample from the tank on the same cup at the same temperature.
Inspect the feed filter and the return strainer each shift and keep what they collect; gel particles look and feel different from dirt.
Check the changeover cleaning record for the loop, and check that no tank, drum or de-aeration step has removed the air headspace.

THEN CHANGE

Black out the roof lights over the coating station, shield the pan and the hoses, and fix any leaking reflector.
Fit a heat exchanger and hold the circulating coating at 25 +/- 2 C.
Cover the pan between runs and stop the practice of leaving a reduced system open.
Keep an air headspace on every tank and drum of UV material: never nitrogen blanket it, never vacuum de-aerate it for storage, and never brim-fill a drum. Where a de-aeration section is fitted for foam, re-aerate the material before it is stored.
Rotate stock through the loop rather than topping up indefinitely, set a maximum working life for material in the system, and flush and clean the loop at every chemistry changeover with the correct cleaner, recorded.

CONFIRM IT IS FIXED

Viscosity on a DIN 4 mm cup to DIN 53211 at 25 +/- 1 C within 3 s across an eight hour run, a clean filter basket and return strainer at the end of the shift, and no seeds or craters on the last twenty parts.

THE TRAP

Thinning the tank back to the target cup reading. Gel is not reversible; thinning disperses partly polymerised material through the batch, so the parts look acceptable while film weight, hardness and cure are all now wrong, and the seeds come back within the hour.

Industrial CoatingsViscosity climbing in the circulation system (continued)

Fault 16 of 20

Plastic part crazes, distorts or stress-cracks

WHAT YOU SEE

Fine surface crazing on a clear part, a warped or dished panel, or cracks appearing at mould gates, ribs and fixing bosses days after coating. Polycarbonate, acrylic and ABS are the usual sufferers.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Monomer or solvent attacking a stressed mouldingMoulded-in stress plus a solvating liquid is the classic environmental stress cracking combination; the liquid finds the stress and opens it.
Infra-red load from a mercury lampThe part softens and distorts under radiant heat well below the temperature the console shows for the lamp.
Moulded-in stress never relieved before coatingThe stress was already there; the coating only revealed it.
Cure shrinkage of a heavy film pulling a thin panelThe coating shrinks on cure and a thin moulding follows it, which shows as dishing on a single sided coat.
Wrong grade for the polymerA coating formulated for metal or glass carries monomers chosen for adhesion and hardness rather than for compatibility with polycarbonate or acrylic.

Industrial CoatingsPlastic part crazes, distorts or stress-cracks

Fault 16 of 20 · continued

Plastic part crazes, distorts or stress-cracks

CHECK, IN THIS ORDER

Anneal a sample of the uncoated part at the polymer's recommended temperature and coat it; if the crazing stops, the cause is moulded-in stress.
Measure part surface temperature immediately after the lamp or the array with an infra-red thermometer and compare with the polymer's heat distortion temperature.
Identify the polymer positively rather than by appearance, and check it against the grade's substrate list.
Coat both sides of a sample panel and compare distortion with a single sided coat.
Check wetting on the part as it arrives with wetting tension test inks applied per the ink supplier's method, or contact angle to ASTM D7334 where a number has to be defensible. ASTM D2578 is written for polyethylene and polypropylene film, so on moulded PC, PMMA and ABS quote it as the ink method rather than as a conforming test. Treatment level decides whether the coating wets and bonds, not how much coating is applied.

THEN CHANGE

Move the pigmented or clear coat to an LED array, which removes the infra-red load a mercury lamp puts into the part. A high irradiance LED head plus the exotherm of the curing film still heats the surface, so measure part surface temperature immediately after the array and compare it with the polymer's heat distortion temperature before you accept the change.
Anneal or stress relieve the mouldings before coating, and raise it with the moulder as a moulding problem.
Fit dichroic reflectors and an air gap, and add forced cooling under the lamp on a mercury line.
Reduce film weight and cure in two lighter passes so shrinkage is shared.
Change to the grade specified for that polymer. UniCure Ultra 5000 is our broadest compatibility plastics grade and the usual starting point for a compatibility trial, but compatibility is specific to the polymer, the grade, the release system and the moulded-in stress, so it is always trialled and qualified on the customer's own mouldings before a production run.

CONFIRM IT IS FIXED

No crazing visible at 10x at 48 h on twenty parts, flatness within the customer's drawing tolerance, and cross-cut to ISO 2409 at class 0 or 1 at 24 h. Compatibility is a property of the customer's moulding, its polymer grade and its moulding stress together, and it is qualified on their own parts before a production run.

THE TRAP

Blaming the coating and slowing the line under a mercury lamp to get more dose. Slower means longer under the infra-red, so the part gets hotter, the distortion gets worse, and the stress cracking arrives faster.

Industrial CoatingsPlastic part crazes, distorts or stress-cracks (continued)

Fault 17 of 20

Blisters and rust creep on steel weeks after a good run

WHAT YOU SEE

Small domed blisters, sometimes with liquid inside, and rust spreading back from a scribe, a chip or an edge far faster than the salt spray result suggested. It appears weeks or months after a batch that passed every test on the line, and the pattern follows the steel and the blast batch, not the applicator or the rack.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Soluble salts left on the steel after blasting or picklingChloride and sulfate are hygroscopic. Under the film they draw water through the coating osmotically and the cell that forms lifts the film and corrodes the steel underneath. A few tens of mg/m2 is already too much for immersion or condensing service, and none of it is visible.
Flash rust formed between preparation and coatingA blasted surface rusts within hours at high humidity, so the coating bonds to a loose oxide layer rather than to steel.
Blast grade or profile not to specificationAdhesion on steel is largely mechanical. Too shallow a profile gives the coating nothing to key into, and too coarse a profile leaves peaks standing proud of a thin film, which is exactly where rust starts.
Zinc coated steel with white rust, passivation or anti-fingerprint oilNew galvanised steel carries a passivation layer and often an anti-fingerprint oil, and stored galv grows white rust. All three stop a coating bonding to the zinc, and a dry wipe removes none of them.
Condensation on the prepared steel before coatingSteel below the dew point carries a water film that redissolves whatever salt is left and puts it straight back into the interface.

Industrial CoatingsBlisters and rust creep on steel weeks after a good run

Fault 17 of 20 · continued

Blisters and rust creep on steel weeks after a good run

CHECK, IN THIS ORDER

Bresle patch to ISO 8502-6 with conductivity to ISO 8502-9 on the prepared surface before any coating, at least three patches per batch and at least one on a weld or an internal corner where washing is worst.
Blast grade against the ISO 8501-1 pictorial standards and profile to ISO 8503 by comparator or replica tape, recorded per batch.
Surface temperature against dew point to ISO 8502-4 at preparation and again at coating.
Cut a blister open, test the liquid with pH paper and a chloride strip where the volume allows, and look at the steel underneath for pitting.
Check the abrasive: whether it is recycled, whether it is separated, and when it was last tested for salt.

THEN CHANGE

Wash the prepared steel with clean fresh water, or wet abrasive blast, and then re-test with the Bresle patch rather than assuming the wash worked.
Set a maximum time from preparation to coating, typically within the same shift and before any visible flash rust, and coat only inside the dew point margin.
Bring blast grade and profile to the specification for the system, typically Sa 2.5 to ISO 8501-1 with the profile inside the grade's stated range, and record both per batch.
On galvanised steel use the pre-treatment or primer specified for zinc. Sweep blasting, alkaline cleaning and a dedicated zinc primer are not interchangeable, and stored galv carrying white rust is cleaned back or rejected.
Agree the maximum soluble salt level in the specification with the customer before the job, because it is their steel and their service condition that set it.

CONFIRM IT IS FIXED

Bresle conductivity to ISO 8502-9 inside the agreed limit on at least three patches of the prepared surface, blast grade and profile recorded to ISO 8501-1 and ISO 8503, substrate at least 3 C above dew point at the time of coating to ISO 8502-4, and neutral salt spray to ASTM B117 rated to ASTM D1654 against the agreed hours, run on the customer's own steel with their own preparation. Corrosion performance is a property of the customer's steel, its preparation, its edges and welds and this coating together, and it is qualified on their parts.

THE TRAP

Coating over flash rust or over a salt contaminated surface because the blast booth is committed and the line is waiting. The parts look right, pass the tape test and pass at goods inward, and the cost arrives months later as a whole batch of blistered work that has to be stripped.

Industrial CoatingsBlisters and rust creep on steel weeks after a good run (continued)

Fault 18 of 20

Coating yellows or chalks in service

WHAT YOU SEE

A white or clear finish going yellow within weeks indoors or months outdoors, or an exterior part losing gloss and powdering at the surface when wiped.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Aromatic urethane acrylate chemistry used in an exterior applicationThe aromatic ring absorbs ultraviolet and forms coloured photo-oxidation products; it is a durable chemistry indoors and a yellowing one outside.
Under-cure leaving photoinitiator and its fragments in the filmResidual initiator continues to absorb light and generate coloured products long after the part has left the line.
Excess dose or excess initiator on a white, or an over-bake on a conventional systemOver-irradiating a titanium dioxide filled film yellows it directly, and so does holding a stoving white above its peak metal temperature window, which is why a white can be made worse by more lamp or more oven.
No light stabiliser package for the exposure the part seesAn exterior coating needs an absorber and a hindered amine stabiliser; an interior grade carries neither and was never meant to.
Heat in service rather than lightYellowing near a lamp housing or on an automotive under-bonnet part is thermal, and no light stabiliser will address it.

Industrial CoatingsCoating yellows or chalks in service

Fault 18 of 20 · continued

Coating yellows or chalks in service

CHECK, IN THIS ORDER

Ask what exposure the part actually sees: interior, behind glass, or full exterior, and for how many years.
Radiometer pass, or the oven profile on a conventional system, to confirm cure is inside the window rather than short or excessive, and MEK double rubs to ASTM D5402 to confirm the film is fully cured.
Accelerated weathering on the customer's own substrate with the coating in question, xenon arc or fluorescent UV to the method agreed, with a control panel alongside.
Colour measurement with a spectrophotometer, reported as delta b and delta E against the retained standard, not judged by eye.
Confirm whether the yellowing tracks light or heat by masking half a part and exposing it.

THEN CHANGE

Move exterior work to an aliphatic chemistry with a light stabiliser package, and accept its cost and its cure characteristics.
Correct cure into the window; under-cure and over-irradiation both yellow a white, from opposite directions.
Re-specify the grade for the exposure rather than adding a stabiliser at the line, which does not disperse properly into a coating at that stage.
Where the part is behind glass, put that in the specification: the glass removes the short wavelengths and changes what is needed.
Where the cause is heat, change the chemistry or the part's thermal environment, not the stabiliser.

CONFIRM IT IS FIXED

Delta b and delta E measured against the retained standard after the agreed accelerated weathering exposure on the customer's own substrate, within the limit written into the specification, with a control panel run alongside. We state the formulation and the class of stabilisation used; service life is qualified on the customer's part in their own exposure.

THE TRAP

Treating yellowing as a colour problem and re-matching the white a little bluer to cover it. The shift continues, the part goes green before it goes yellow, and the underlying cause, which is usually cure or the wrong chemistry for the exposure, is now hidden behind a special shade nobody can reproduce.

Industrial CoatingsCoating yellows or chalks in service (continued)

Fault 19 of 20

Chemical resistance short of spec a week later

WHAT YOU SEE

The part passes the solvent rub test at the line and fails at the customer after cleaning fluid, hand sanitiser, sun cream, oil or a detergent wash. The film may soften, swell, go matt or lift entirely.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Under-cureChemical resistance comes from crosslink density; an incompletely cured network lets the reagent in between the chains and swells. On a conventional system this is an under-bake: time at peak metal temperature short, not oven set point wrong.
Property tested too earlyOn a two-pack system full property development takes 5 to 7 days at 25 C, so a test run at the line is testing the wrong film.
Film weight below the windowA thin film is penetrated to the interface in less time, so a marginal chemistry fails first where the coating is thinnest, usually at edges, radii, holes and weld crowns, because surface tension pulls the film back from an edge while it is still liquid.
Wrong grade for the reagentA coating resistant to alcohol and water may have no resistance to a ketone or an alkaline cleaner; resistance is specific, not general.
Slip or flow additive at the surfaceAn additive rich surface layer has a different chemistry from the bulk film and is often the layer that goes matt.

Industrial CoatingsChemical resistance short of spec a week later

Fault 19 of 20 · continued

Chemical resistance short of spec a week later

CHECK, IN THIS ORDER

ASTM D5402 MEK double rubs at 24 h and again at 7 days, on the same batch, reported with the cloth, the load and the stroke length used, and run under the extraction and protection set out in the methods section.
Dry film thickness to ISO 2808 at the failing area and at the flat face, including an edge and a hole.
Radiometer pass at the failing face to confirm dose and peak irradiance, or an oven profile logged on the part for a conventional system.
Get the actual reagent, the concentration, the contact time and the temperature from the customer, and run a spot test with it rather than with MEK.
Compare a panel cured on the line with a panel given a laboratory full cure; if the laboratory panel passes, the problem is cure, not chemistry.

THEN CHANGE

Correct cure first: dose, peak irradiance and lamp condition, in that order, or peak metal temperature and time at temperature on an oven line.
Hold parts 7 days before release testing on two-pack systems, and state the conditioning time on the release record.
Bring film weight to the upper end of the window on the faces the reagent reaches, and add an edge stripe coat where edges and welds are the failing point.
Move to the correct grade for the reagent once cure and thickness are proven correct.
Remove slip additives from the top layer where the matting is at the surface only.

CONFIRM IT IS FIXED

50 or more MEK double rubs to ASTM D5402 at 24 h and 100 or more at 7 days for a two-pack build, plus a pass on the customer's own reagent at their own concentration, contact time and temperature. Chemical resistance is a property of this coating on the customer's substrate at their film weight, and it is qualified on their parts.

THE TRAP

Running the MEK rub at the end of the line, getting a pass and releasing. The film is still developing, and a pass at one hour tells you the surface cured, not that the network is finished.

Industrial CoatingsChemical resistance short of spec a week later (continued)

Fault 20 of 20

Concrete floor coating blisters or delaminates

WHAT YOU SEE

Domed blisters, sometimes with water underneath, or whole sheets lifting weeks after application, often in patches that follow the slab rather than the application pattern.

WHAT IT LOOKS LIKE

CAUSES, MOST LIKELY FIRST

Moisture vapour driving up through the slabWater vapour moving from the ground through the slab reaches an impermeable coating, condenses under it and lifts it. The pressure is generated by the slab, not by the coating.
Laitance and curing compound left on the surfaceThe coating bonds to a weak surface layer, and that layer separates from the concrete below under any load.
Slab too youngA new slab carries free water and high alkalinity for weeks; coating it early guarantees the first two causes at once.
No mechanical profileAdhesion on concrete is largely mechanical; a smooth power floated slab gives the coating nothing to key into.
High surface alkalinity attacking the binderAlkaline hydrolysis at the interface degrades ester linkages in the binder, which shows as a soft, soapy interface under the lifted coating.

Industrial CoatingsConcrete floor coating blisters or delaminates

Fault 20 of 20 · continued

Concrete floor coating blisters or delaminates

CHECK, IN THIS ORDER

ASTM D4263 plastic sheet test as an initial screen, then in-situ relative humidity probes to ASTM F2170 for a number to work from, with the number and placement of holes per the standard for the floor area and each probe left to equilibrate in the slab for 24 h before it is read.
Surface pH with pH paper and distilled water on the prepared slab.
Pull-off adhesion to ASTM D4541 with a portable tester on test patches; record whether failure is in the coating, at the interface or in the concrete itself.
Inspect the prepared profile against the ICRI concrete surface profile chips; typically CSP 2 to 3 for a thin film coating.
Cut a blister open and check for liquid water underneath, and test that water with pH paper.

THEN CHANGE

Do not coat until the slab moisture is inside the limit written into the specification for the system. The acceptance figure is set by the coating, not by the slab: 75 % RH is a conservative default carried over from flooring adhesives, many 2K epoxy systems are qualified to 80 or 85 % RH, and the figure used is recorded with the grade. On a new slab this is weeks, and no coating change substitutes for it.
Where the moisture cannot wait, fit a moisture mitigation membrane system, agreed in advance, and coat over that.
Prepare mechanically by shot blasting or diamond grinding to remove laitance and curing compound and to create the profile; acid etching is not equivalent.
Apply test patches on the actual slab, in the actual conditions, and pull them off before committing to the floor.
Re-specify to a more alkali resistant chemistry where the slab pH cannot be brought down.

CONFIRM IT IS FIXED

Pull-off adhesion to ASTM D4541 above 1.5 MPa, or failing in the concrete rather than at the interface, on at least three test patches on the actual slab, with the slab relative humidity recorded to ASTM F2170 against the limit in the specification before application. Floor adhesion is a property of the customer's slab, its age, its preparation and its moisture condition together with this coating, and it is qualified on their floor.

THE TRAP

Sealing over a blistered area with more coating. The vapour pressure driving the blister is still there, so the repair lifts too, and the extra film makes the next blister larger and harder to see coming.

Industrial CoatingsConcrete floor coating blisters or delaminates (continued)

At the press

One page to work from, and the methods behind every number.

Quick reference

One page at the press

Symptom	Check this first	Change this first
Craters, fisheyes, coating pulls back	Water break test on the cleaned substrate; blow the spray air onto a mirror	Fix the cleaning stage and the compressed air, not the coating
Passes tape test, fails after humidity	Mix time on the pot against the clock; the glass wash and the glass batch	Mix fresh inside pot life; clean the glass properly and re-test after conditioning
Tacky face, hard body	Is the grade specified for the lamp fitted, mercury or LED; then lamp hours and quartz	Correct the grade or the lamp; clean quartz and reflector and close the lamp to part distance
Whole film lifts off, underside soft	Dry film thickness to ISO 2808 and lamp type on that station	Reduce film weight; move white to a gallium doped or 405 nm source and read it with a UV-V head
Topcoat peels off the basecoat	Basecoat dose and the interval between coats against the recoat window	Cut basecoat dose to a partial cure; shorten the interval
Sags and runs on vertical faces	DIN 4 mm cup to DIN 53211 at 25 C at the applicator and dwell time to the lamp	Cool the coating to 25 C; shorten the dwell before the lamp
Orange peel or machine direction ribbing	Cup reading at 25 C and part temperature entering the coater	Warm the parts to within 5 C of the coating
Pinholes on wood, MDF edge or concrete	Substrate moisture meter, or ASTM F2170 probes equilibrated 24 h in a slab	Seal first and cure it; split the build into two lighter passes

Industrial CoatingsQuick reference

Quick reference

One page at the press

Symptom	Check this first	Change this first
Craters in a baked film on a non-porous part	Oven profile logged on the part, first two minutes of the ramp	Lower the front zone and lengthen the flash before the oven
Curtain tears or shows vertical streaks	Slot lip and edge guides under a torch; foam in the return	Clean the lip and guides; screen the curtain from draughts
Sandy matt patches and overspray dust	Gun to part distance and hall temperature at the booth	Bring the gun to the standard distance; move to the slower thinner blend
Cracks at a bend or after thermal cycling	Dry film thickness at the bend; time since cure	Reduce film weight and hold 24 h before forming
Gloss or matt level drifting through the shift	ISO 2813 at 60 degrees, five points, start and end of run	Agitate the matt grade continuously; lock coating temperature
Cup reading climbing hour by hour	Daylight ingress with the lamps off, then indicator cards on the pan with the lamps guarded and running	Black out and shield the loop; fit a heat exchanger at 25 C
Clear plastic part crazing or warping	Part surface temperature straight after the lamp or the array	Move to LED, or fit dichroic reflectors and cooling, and re-measure the part temperature
Blisters and rust creep on steel weeks later	Bresle patch to ISO 8502-6 and conductivity to ISO 8502-9 on the prepared steel	Wash or wet blast the prepared surface and re-test before coating

Industrial CoatingsQuick reference

Quick reference

One page at the press

Symptom	Check this first	Change this first
Resistance fails at the customer, passed here	MEK double rubs at 24 h and at 7 days on the same batch	Hold two-pack parts 7 days before release testing

Industrial CoatingsQuick reference

Test methods

The measurements this guide relies on

Measurement	Method	What it tells you, and the usual acceptance
Cross-cut adhesion	ISO 2409	A lattice cut through the film to the substrate, tape applied and removed, classified 0 to 5, with the cut spacing set by the film thickness: 1 mm for films to 60 micron on a hard substrate, 2 mm for 61 to 120 micron, 3 mm for 121 to 250 micron. The method is not valid above 250 micron, so heavy concrete and floor builds go to pull-off instead. Target class 0 or 1 at 24 h on rigid substrates, then repeat after humidity conditioning, because a physically adhering film passes at day one. Not used on very thin films over soft substrates, where the blade cuts the substrate rather than the coating.
Humidity conditioning before adhesion testing	ISO 6270-2 (or ASTM D2247 / ASTM D4585)	24 h at 40 C constant condensation is the house condition; where a cabinet is set to 40 C and 95 % RH instead, record that as the house condition rather than citing it as the standard. Test within 30 min of removal and again after 24 h recovery, and record both, because a re-dried film can recover a pass it did not deserve. Where the service condition is a wash, a boil or a dishwasher cycle, run the customer's own cycle as well.
Pull-off adhesion	ASTM D4541	A dolly bonded to the coating and pulled normal to the surface with a portable tester, reported in MPa with the failure mode stated. The house method for concrete, heavy metal coatings and any case where the failure plane has to be a reportable result, run on a flat panel or a flat area of the same substrate and build. On concrete, above 1.5 MPa, or failure inside the concrete itself, is the usual acceptance; the failure mode matters more than the number.
Solvent double rubs	ASTM D5402	Methyl ethyl ketone on a cloth under a defined load, rubbed until the coating softens or the substrate shows. Reports cure state, not adhesion. Typically 50 or more for a cured industrial clear at 24 h and 100 or more for a two-pack build at 7 days. Always report the load, the cloth and the stroke length with the number. Run it under local exhaust or in a fume cupboard, away from lamps, heaters and any other ignition source, with eye protection and butyl rubber or laminate film gloves: MEK has a flash point near minus 9 C, an eight hour exposure limit of the order of 200 ppm, and it permeates nitrile within minutes, so the line glove is not suitable for this test. Keep the solvent in a closed, labelled, earthed container and put used cloths in a lidded metal bin.
Flow cup viscosity	DIN 53211 (DIN 4 mm cup); ISO 2431 (ISO flow cup, 4 mm jet); ASTM D1200 (Ford); ASTM D4212 (Zahn)	Efflux time in seconds, with the cup and the standard quoted together. A DIN 4 second is not an ISO 2431 second, a Ford B4 second or a Zahn 2 second, and a target set on one cup is re-established on the other rather than converted. ISO 2431 specifies 23 +/- 0.5 C; the house condition of 25 +/- 1 C is recorded as an agreed deviation, because efflux time moves roughly 2 to 3 % per degree. Spray 16 to 25 s, dip and flow coat 16 to 30 s, curtain 25 to 45 s, roller coat 30 to 60 s and above. Readings taken at whatever the hall happens to be are the commonest reason a finish drifts across a shift.
Film thickness	ISO 2808	Dry film by magnetic gauge on steel and eddy current on aluminium, calibrated on the same substrate class; wet film by comb at the applicator; on board and paper by weighing a cut area of coated against uncoated stock. Measure edges, holes and weld crowns as well as the flat face, because they always carry less. Typical builds: 3 to 8 micron on glass and plastic hard coats, 8 to 25 micron on decorative metal, 2.0 to 3.5 g/m2 on board.
Specular gloss	ISO 2813	60 degrees as the standard geometry, 20 degrees for high gloss and 85 degrees for matt. Five points per part, judged against a retained reference standard kept at the line. Acceptance typically within 3 GU of the standard at 60 degrees; widen to about 5 GU at 20 degrees on a high gloss and about 5 GU at 85 degrees on a deep matt to match the method's reproducibility, and state the geometry with every acceptance figure. Gloss on a porous substrate is a property of the substrate and the coating together, so it is re-qualified when the substrate changes.

Industrial CoatingsTest methods

Test methods

The measurements this guide relies on

Measurement	Method	What it tells you, and the usual acceptance
Hardness and abrasion	ASTM D3363 pencil hardness; ASTM D4060 Taber abrasion; ASTM D1044 haze after abrasion with haze to ASTM D1003	Pencil hardness gives a quick floor check, typically 2H to 4H for a UV hard coat, and is operator sensitive, so one person runs it. Taber to ASTM D4060, with the wheel, load and cycle count stated, is reported as mass loss or wear index. On a transparent coating or a coated plastic, report abrasion as haze increase to ASTM D1044 with haze measured to ASTM D1003, stating the wheel, load and cycle count, typically CS-10F wheels, 500 g per wheel and 100 cycles.
Wetting tension and contact angle	ASTM D2578 (polyethylene and polypropylene film); ASTM D7334 contact angle	Test inks or dyne pens applied to the part about to be coated, not to a retained sample, with the system stated because pens and inks do not read alike. ASTM D2578 covers polyethylene and polypropylene film only, so on rigid mouldings and on metal it is quoted as the ink method and not as a conforming test; where a number has to go into a specification, use advancing contact angle to ASTM D7334. On glass and metal the water break test is the working check.
Dew point and substrate temperature	ISO 8502-4	Hall temperature, relative humidity and substrate surface temperature read at the line, with the dew point taken from the tables or the instrument. The rule is substrate at least 3 C above dew point at application and until the film is through the flash zone. A cold part from winter stores can sit below dew point at a hall RH of 55 %, so the reading is taken on the part, not on the air.
Soluble salts and steel preparation	ISO 8502-6 Bresle patch with ISO 8502-9 conductivity; ISO 8501-1 blast grades; ISO 8503 profile	The Bresle patch extracts salt from a known area and conductivity converts it to a surface level, at least three patches per batch with one on a weld or a corner. Blast grade is judged against the ISO 8501-1 pictorial standards, typically Sa 2.5 for an industrial build, and profile is measured with a comparator or replica tape to ISO 8503 and recorded per batch. All three are taken on the prepared surface before coating, because none of them can be recovered afterwards.
Neutral salt spray	ASTM B117 or ISO 9227 NSS, rated to ASTM D1654 and ASTM D714 or ISO 4628-2	Continuous salt fog on a scribed coated panel. B117 and ISO 9227 define how the cabinet is run and contain no rating scheme, so the exposure is rated separately: creep from the scribe to ASTM D1654, blistering to ASTM D714 or ISO 4628-2, rusting to ISO 4628-3, and the result reported as hours to the agreed rating. A comparative screen between systems, not a prediction of service life. On metal it tests the pre-treatment and the coating together, so the customer's own conversion coating and their own metal are used.
Oven cure profile	No single test standard; instrument method, recorded in house	A multi-channel data logger travels with the part, thermocouples attached to the heaviest section, the lightest section and a shadowed face. Report the ramp, the peak metal temperature and the minutes above the stated temperature for each channel, against the grade's schedule. Air temperature in the tunnel and the controller set point are not the cure; the metal temperature is. Re-profile after any change to loading, line speed, burner or extraction, and at least twice a year.
UV dose and peak irradiance	No single test standard; band matched radiometer, instrument and head band recorded	A radiometer carried through at running speed on a fixture that puts the head where the worst part face sits, reporting dose in mJ/cm2 and peak irradiance in W/cm2. Match the head band to the source: UV-A for a mercury arc or a 385 nm array, UV-V at roughly 395 to 445 nm for 395 and 405 nm arrays, since the UV-A band ends at 400 nm. Record the make, model, head band, calibration date, running speed and where on the part the head was fixtured, or the trace cannot be compared with the next one.

Industrial CoatingsTest methods

Test methods

The measurements this guide relies on

Measurement	Method	What it tells you, and the usual acceptance
Concrete moisture and alkalinity	ASTM D4263 plastic sheet; ASTM F2170 in-situ relative humidity; ASTM F1869 calcium chloride	The sheet test is a go or no-go screen taped to the slab overnight. The in-situ probes give the number to work from, with holes numbered and placed per the standard for the floor area and each probe left to equilibrate in the slab for 24 h before it is read. The acceptance limit is set by the coating system and written into the specification with the grade: 75 % RH is a conservative default carried over from flooring adhesives, and many 2K epoxy systems are qualified to 80 or 85 % RH. F1869 may be used alongside where the specification calls for it, but it measures only the top 12 to 20 mm. Surface pH is checked alongside with pH paper and distilled water.

Industrial CoatingsTest methods

When the press-floor fixes have not worked

What to send us

Twenty parts, both good and bad, from the same run, unwashed, wrapped face to back in clean paper and not in bubble wrap, with the rack position, nozzle or fixture marked on each one.
A sample of the uncoated substrate from the same batch as the failed parts, in its as-received packaging, plus the substrate supplier, grade and batch number. On steel, add the blast grade, the profile reading and the Bresle result for that batch.
One litre of the coating drawn from the pan at the time of the fault, and one unopened container from the same batch. Draw UV material into a clean opaque or amber container filled to no more than three quarters, so it keeps the air headspace its stabiliser needs, close it securely, keep it below 25 C and out of daylight, and never decant into a clear bottle. Label both with our batch number, the date and time drawn and the draw point, ship with the safety data sheet and the correct dangerous goods declaration, and keep a retained duplicate at your end.
The shift log for the run: cup readings with the cup named, times and temperatures, coating and part temperature, hall RH and dew point margin, film weight, line speed, and the mix time and weighed ratio for any two-pack material.
The radiometer traces, with the make and model of the instrument, the band of the head, its calibration date, the running speed, and a note or photograph of where on the part the head was fixtured.
For an oven line, the data logger profile taken on the part, with the channel positions in the load, the ramp, the peak metal temperature and the time at temperature.
Lamp or LED hours on each unit, the date the quartz or window was last cleaned, and the date the lamps were last changed.
The actual service condition the part failed in: the reagent and its concentration, the contact time, the temperature, the wash cycle, or the exposure and its duration, in the customer's own words rather than as a description of a test.

Technical servicesales@ufindustries.com | +91 8750051884 | E105, GNEPIP, Industrial Area Site-5, Greater Noida, Uttar Pradesh 201306, India

Industrial CoatingsTechnical service