Technical guides
Troubleshooting guide
Most print and coating faults are not ink faults, but they get treated as ink faults because ink is the easiest thing to change. This guide sets out a diagnostic order that works outward from what changed most recently, one variable at a time, across gravure printing, flexo printing, coating and varnishing, sealing, lamination and adhesion, and curing and drying.
Most print and coating faults are diagnosed in the wrong order. The ink or the coating gets changed first, because it is the easiest thing to change and the easiest thing to blame, and because a drum can be swapped in minutes while a substrate reel or a drying setting feels like a bigger decision. In practice the cause sits further out: the substrate and its surface condition, the anilox or the engraved cylinder, the doctor blade and its setting, the drying or cure capacity available at the speed you are running, or a line condition that drifted quietly over a shift. Changing the ink at that point does not fix the fault. It adds a second unknown on top of the first.
Work outward from what changed most recently. A new reel, a cleaned or replaced anilox, a new blade, a different job speed, a lamp or dryer that has been running for a long time, a viscosity that has been topped up repeatedly during a run: these are the first places to look, in the order they occurred. Then change one variable at a time, and write down what you changed, what you measured and what the print did. Two changes at once destroys the diagnosis, because a fault that improves tells you nothing about which of the two was responsible, and a fault that stays tells you nothing about either. Process figures quoted in this guide, for anilox volumes, cylinder angles, viscosities, cure energies and seal temperatures, are industry-typical ranges given for orientation. They are not Arctic measurements and not guarantees, and your own line conditions govern.
Where a fault cannot be resolved on press, send it to us. Arctic’s technical centre and laboratory in Greater Noida will examine a sample of the substrate, a sample of the ink or coating, and a sample of the faulty work together. Testing any one of them alone usually explains nothing, because the fault normally lives in the interaction between them: a surface that has aged, a coating asked to do something the substrate will not support, a cure that was adequate for one weight and not the next. Send all three, with the job conditions as you ran them, and the answer is usually visible.
Gravure printing
This section covers faults seen on a wide-web solvent gravure press printing flexible packaging film, surface and reverse print. Each entry gives the fault as a press crew would call it, the likely causes in the order they usually turn out to be the cause, one thing that can be checked at the press in minutes, and the corrective action. Work down the cause list in order and change one variable at a time, recording the viscosity, ink temperature, speed and pressures you started from, so the job can be put back if the change does not help.
Several of these faults look alike on the reel and separate only on a test, so where two causes present the same way the entry names the test that tells them apart. Read the whole action before starting, because some of the corrections pull against each other: drying harder cures one trapping fault and causes the other, slip cures blocking but can cost lamination bond, and a fuller pan cures start-up drying-in but feeds misting.
Checks are made with the press running unless the entry says otherwise. Anything that involves touching the blade, the cylinder surface or the impression roller is done with the press stopped, guarded and isolated. Static and misting are solvent-atmosphere safety items as well as print defects, so they are not worked around to keep a job running.
Process figures quoted here are industry-typical ranges given for orientation only. They are not Arctic measurements, not a specification and not a guarantee, and they do not replace your own set-up sheet for the cylinder, film and ink in use. Where a property needs to be measured rather than judged, the recognised test method is named. For ink-side questions the Arctic technical centre and laboratory in Greater Noida can work through a retained sample of the ink, the film and the printed reel together, which is usually faster than changing grades blind.
| Symptom | Likely causes, most probable first | Check first | Action |
|---|---|---|---|
| Solids look grainy or speckled, as though the engraving screen is showing through. Worst in the heavy blues and blacks, and it gets worse as the press comes up to speed. | Viscosity too high for the cell depth and the speed, so the ink cannot flow out and coalesce before it reaches the dryer; solvent blend too fast for the hall temperature, so drying starts in the cell or on the film before flow-out; electrostatic assist off, set too low, or switched on but not actually drawing current, so ink is not drawn cleanly out of the deeper cells; low or decayed film surface energy, so the transferred ink draws back into islands after it lands; blade pressure high enough to deflect the blade and over-wipe, robbing volume from the cells; cell walls polished or the screen partly plated over on a worn cylinder. | Cup viscosity taken in the pan at running temperature, with the hall temperature noted at the same time, and the ESA setting and the current actually drawn on that unit, both compared against the standing set-up sheet before anything is changed. | Bring viscosity down in small steps with the slow blend recommended for the Arcova grade, a second or two on the cup at a time, and watch the solid close. Raise ESA one step at a time and confirm the current rises with it. Back blade pressure off to the lowest setting that still wipes clean. Separate flow-out from wetting before going further: if the speckle responds to viscosity and ESA it is release and flow-out, but if the film reads low on wetting tension and the ink has drawn back into islands with clean film between them, treat it as the crater and wetting fault instead, because more solvent will not cure it. Reduce the first dryer inlet temperature only where the ink surface is visibly skinning, and re-check retained solvent and trapping before carrying that reduction through the run. If a second, newer cylinder prints the same job clean, the first cylinder is worn and needs re-engraving. |
| Highlights and vignettes break up. The fade goes ragged, the lightest dots drop out altogether, and it is worse on hard, low-compressibility film and as speed rises. | Electrostatic assist off, mis-set or arcing; viscosity too high for the smallest cells to release; impression roller and impression pressure together, the roller hard, glazed, worn, out of round or carrying flats, or the pressure set below the sheet value, so the film is not pressed into the cells evenly across the width; shallow highlight cells partly blocked with dried ink; film surface energy too low, or the surface too rough and unyielding for fine dot contact. | Confirm the ESA current actually rises on that unit when it is switched on rather than just the lamp coming up, and read impression pressure against the set-up sheet. Feel the impression roller for flats and glazing only with the press stopped, guarded and isolated, never on a running nip. | Set ESA first, then impression, in that order. Raise impression pressure in small steps and watch the tail of the solids for smear and the film for distortion. Replace impression rollers that have gone hard or glazed, a typical working band for film gravure is 70 to 85 Shore A. Reduce viscosity towards the set-up value with the recommended blend. Separate blocked cells from the rest: an ESA or impression fault affects a whole band or the full width, whereas blocked cells lose the same dots in the same part of the image on every revolution with the rest of the highlight clean. Clean those cells with the recommended solvent and a soft brush, never with anything abrasive. |
| Fine lines running in the web direction through the solids. Some wander slowly across the cylinder, others sit in the same place for the whole reel. | Hard particle trapped under the blade, usually dried ink skin, filter debris or a pigment agglomerate; nick or burr on the blade edge; blade angle or pressure wrong; blade oscillation stopped or stroke too short; contaminated or unfiltered return in the ink circuit; ink starvation at one end of the pan. | Offset the blade laterally by a few millimetres, or stop the oscillation at a different point in its stroke. A line that moves with the blade is the blade or something caught under it, a line that stays in the same place on the web is in the cylinder and belongs to the next fault. Then, with the press stopped and guarded, wipe a tissue or cotton bud along the blade edge and look for snagging, and look into the pan return for skin and debris. | A line that wanders is nearly always debris: flush and re-filter the ink through the recommended mesh, typically 100 to 200 micron for gravure, and clean the pan, hoses and blade holder. A line that stays put but moves with the blade is the blade or the holder: fit a new blade, deburr and clean the holder, check the backing blade is straight and the holder parallel, and reset to the press standard, typically a contact angle in the 55 to 65 degree band with the lowest pressure that wipes clean. Confirm oscillation is running and the stroke is at the set value. Keep the pan covered so airborne dust stays out. If the line does not move when the blade is offset, stop fitting blades and go to the cylinder score entry. |
| A hard line that will not clean off. It is in the cylinder itself, you can feel it with a fingernail, and it prints on every repeat. | Hard contamination circulating in the ink, commonly chrome flakes, dried ink lumps or grit picked up from the floor; blade pressure far too high; blade material too hard for that chrome surface; chrome layer thin or already worn through to the copper; ink pan left uncovered. | With the press stopped, guarded and isolated, pass a fingernail or a plastic scraper over the cylinder surface to separate a real score from a print defect, and confirm the line does not move when the blade is offset. Then strip the ink filter and look for chrome flakes and grit. | A scored cylinder cannot be recovered on press. Take it out and have it re-chromed or re-engraved. Before the replacement goes on, find the source: change the filter, clean the whole ink circuit including pan, pump, hoses and reservoir, cover the pan, and reduce blade pressure to the lowest that wipes clean. Running on with a scored cylinder chews through blades, deepens the score and carries debris forward to the next unit. |
| A regular pattern, tartan or a rosette, in the mid-tones where two or three colours overlap. It was not on the approved proof. | Screen angles on the separations too close together, or one angle repeated on two strong colours; the artwork itself carrying a fine repeat at a pitch close to the engraving ruling; ruling, cell geometry or stylus angle differing between cylinders engraved at different times or by different suppliers; image resampled at a ruling close to the engraving screen; register drifting far enough to shift the overlap; a unit bouncing and modulating the overlap. | Look at the affected mid-tone through a loupe to see whether the beat is between two colours or inside one separation, then pull the cylinder record sheets and compare ruling, screen angle and stylus angle across the whole set, and check register on that area. | This is corrected in pre-press, not on the press. Do not apply the offset rule of 30 degrees between colours, because electromechanical engraving offers only a limited set of angles for a given head and cell compression, commonly quoted around 30, 37, 45, 53 and 60 degrees as typical values. Give the two strongest colours the largest separation the engraving set allows, never repeat one angle on two strong colours, and where that is still not enough change the ruling or the cell compression on one separation. Keep ruling and stylus geometry consistent across the set, typically 60 to 140 lines per cm for flexible packaging depending on image detail. Where the design carries its own fine repeat, re-rasterise or re-angle that element rather than the whole separation. On press the only useful moves are to tighten register and to stop any bounce in the unit, either of which can turn a borderline pattern into a visible one. |
| It matched at the start, by the third reel the shade has crept. The blue looks dirtier and the light tints have gone strong. | Solvent evaporating from an open pan, so the solids content of the ink climbs and the same cell volume carries more pigment, which lifts strength and pulls the tints up while viscosity rises with it; the viscosity controller dosing neat fast solvent, or the crew topping up with solvent rather than ink, so strength falls away instead; ink temperature climbing through the run, which shifts viscosity and release together; pigment settling in a pan or drum that is not being agitated; contamination from wash-up residue or from pick-back at the following station; dryer temperature or line speed changed mid-run. | Viscosity and ink temperature at every unit, measured with the same cup, at the same point and in the same way as at start-up, plus confirmation that the viscosity controller is dosing the correct blend and not neat fast solvent. | Read the direction of the drift first, because it names the cause. Strength up across the set with viscosity up is solvent loss, so cover the pans and correct with the specified blend. Strength down with viscosity sitting on the set number is over-dosing of solvent, so top up with fresh ink, not solvent. One colour going dirty while the rest hold standard is contamination or pick-back from the following unit, so wash that circuit out and look at the next unit's blade. Control viscosity and temperature together, not viscosity alone, and always read in the pan at running temperature. Keep drums agitated. Use the blend specified for the Arcova grade at the running speed and move to a slower blend for long runs in a hot press hall. Re-check colour every reel under the same illuminant and viewing conditions, ISO 3664, on the same backing card, not against memory of the first reel. |
| The second colour will not lie down on the first. It beads up, looks mottled, or lifts off and comes back on the next blade. | First-down ink still too wet entering the next nip; second ink viscosity too low, so its solvent attacks the layer beneath; print sequence wrong for the coverage and strengths involved; two different ink systems or two different solvent blends running in the same set; first-down ink over-dried and glassy, or carrying slip at its surface, so nothing keys to it. | With the press at crawl, touch the web with a clean glove at the entry to the next unit, then look at that unit's blade for pick-back of the previous colour. A tacky web with pick-back on the blade is under-dried. A hard, glossy first layer with no pick-back and the second colour beading on it is over-dried or slip contaminated. | Name which of the two it is before touching a dryer, because the corrections are opposite and the wrong one doubles the fault. Under-dried: raise air volume on the first station first, then temperature, in small steps, until the film is dry to the touch at the next nip but not brittle. Over-dried: reduce heat and air on that station and move the blend slower, and take surface slip out of the first-down colour. In both cases bring the second ink's viscosity back up towards its set-up value to cut solvent attack. Review sequence so that a heavy opaque solid is not asked to trap over a large area of wet ink. Keep the whole set within one family, Arcova for wide-web gravure, on one recommended blend. Mixing systems or blends between stations is one of the commonest causes of a trapping fault. |
| Little round holes in the solid, each with a clean rim and the film visible through it. Often in patches or down one side of the web. | Film surface energy too low or treatment decayed since the reel was made; slip additive or antiblock bloom on aged film; silicone or oil contamination from the press, a sprayed lubricant or a leaking bearing; antifoam or roller lubricant carried into the ink; foam and entrained air in the pan; ink surface tension too high for that film. | Wetting tension across the width and at both ends of the reel. The shop-floor pen is the quick press-side read and is known to read high on an aged or contaminated surface, so confirm a doubtful result with the test-fluid method, ASTM D2578 or ISO 8296. At the same time wipe the unwind and the idler rollers with a clean white cloth and look for oil, and look into the pan for foam. | Print at the treatment level the film specification calls for, typically around 38 to 42 dyne per cm for surface print and higher again for reverse print and lamination, and treat in line where the reel has decayed. Adding solvent does not cure a wetting fault. Where the pan is foaming, cure the air ingress rather than the foam: raise pan level, check the pump suction line and fittings for leaks, and slow the pump so the return does not fall into the pan and whip air in. Never dose antifoam into a gravure ink to clear it. If the wetting tension reads correctly and craters persist, the cause is contamination: clean the roller train, change the ink out of the circuit rather than trying to filter it, and stop any use of silicone spray near the press. A wetting-adjusted Arcova grade can raise flow-out on difficult film, but it corrects the ink, not a dirty or untreated web. |
| Web clings to the rollers and to the operator, ink whiskers and cobwebs appear at the edges, dust is pulled into the print, and you get a crack off the frame when you touch it. | Static bars switched off, dirty, or set at the wrong distance; ungrounded rollers or a disconnected frame earth; high speed and film separating from rollers, charging the web; low humidity in the press hall; the electrostatic assist itself charging the web, with nothing downstream to take the charge off; solvent inks and dry film are both good insulators. | Read the web with a hand-held electrostatic fieldmeter at the maker's stated distance, on the last free span before the print and again after the chill roll, and confirm the bars are powered, clean and mounted at the right distance from the web, typically 20 to 50 mm. Check the frame earth and roller earths for continuity with a meter rather than assuming them. Do not test for charge by holding a hand near the web: in a solvent atmosphere a discharge is an ignition source. | Clean and power the bars and position them on the last free span before the print and again before the rewind. Earth every roller and the frame and prove continuity. Check the ESA setting is no higher than the highlight cells need, since it is itself a charging system. Bring hall humidity up into the band these halls commonly hold, roughly 45 to 60 per cent RH, which also steadies solvent flash-off, but raise it in steps and watch gloss, because a fast blend in a humid hall will blush. Treat this as a safety item as well as a print defect: static in a solvent atmosphere is an ignition risk, so do not defeat bars or earths to keep a job running. |
| The print has gone dull and milky, with a chalky haze over the solids and gloss well down. It came on with the weather or after the thinner was changed. | Solvent blend too fast for the hall conditions, so evaporative cooling drops the film below dew point and moisture condenses into the drying ink; hall humidity high and hall temperature low; viscosity corrected with neat fast thinner or with an alcohol rather than the recommended blend, so resin drops out of solution as the true solvent flashes first; too much of one component in a blend that has been topped up by hand over several shifts; poor air movement over the web so the humid boundary layer sits on the print. | Hall temperature and relative humidity at the press, and what the viscosity controller and the crew have actually been dosing, drum label and all, rather than what the set-up sheet says. Dose the correct slower blend into one unit on a short trial and see whether that colour clears while the others stay hazy. | Move to the slower blend recommended for the Arcova grade and stop any topping up with neat thinner or alcohol. Raise hall temperature and bring humidity back down to the working band. Raise dryer air volume on the affected station so the humid layer is carried away. Do not try to cure haze with more heat alone, that drives the fast solvent off faster and can deepen it. Once the blend is right the gloss should recover on the next reel, and if gloss is being argued over with the customer, measure it by ISO 2813 rather than judging it by eye under press hall lighting. |
| The reel smells strongly when it is broken open, the customer reports odour in the pack, and the smell is worst from the middle of the reel. | Dryers under air volume for the speed being run, more often than under temperature; too much slow solvent or retarder in the blend; ink film laid on too heavy for the design; exhaust restricted, filters blocked, or recirculation set so high that solvent-laden air sits over the web; reel wound hot and tight so trapped solvent cannot leave; contaminated recovered solvent or an off-specification solvent grade. | Dryer inlet and exhaust temperatures and damper positions against the set-up sheet, plus a web temperature reading at the chill roll with an infrared thermometer. Compare the smell of a freshly cut edge with the reel end, in a ventilated area and not repeatedly. | Raise air volume before raising temperature, most retention is an air-change problem rather than a heat problem. Clear blocked filters and restricted exhaust and bring recirculation back to the set value. Correct the blend towards faster solvent for the speed being run. Reduce ink film weight where the design allows, and where the dryers are simply short of capacity at that speed, slow the press rather than cook the web. Let reels cool before winding tight, and let them stand before lamination or slitting. Quantify the result by headspace gas chromatography on the printed film before declaring the job fixed. Arctic makes no compliance or food-contact claim for any finished pack, residual solvent and odour limits are the converter's own specification to meet and verify. |
| The reel picks on unwind, print lifts and transfers to the back of the film. Worst near the core and worst on reels that went into the store still warm. | Residual solvent still in the ink film; reel wound too tight and too warm; heavy coverage, especially reverse-printed white and solids; store or container temperature high; ink film too soft for the job; slip level in the ink or the film too low. | Reel surface temperature at the rewind by infrared, and the winding tension and taper settings on the recipe, against a reel that came off clean. Note whether the reel went to the store warm. The blocking test itself is not a press-side check, it is run off press on a cut sample held under load at the temperature the store actually reaches. | Cool the web properly on the chill rolls and let reels stand before stacking or shipping. Reduce winding tension and add taper towards the outside. Improve drying as for solvent retention, blocking and retention usually share one root cause, so work that first. Where the ink film itself is too soft, move to a harder Arcova grade. Add slip only after checking what happens to the reel next: an Arcglide product chosen for that film and store condition will cure blocking, but surface slip is also a cause of the lamination bond failure in the next entry, so on work that is going to be laminated take the route through drying, coverage and winding instead, and agree the ink side with the technical centre. Confirm the fix by a parallel plate blocking load test, ASTM D3354, run at store temperature rather than at laboratory ambient. |
| The laminate peels at the ink, not at the film. Clean colours hold, the heavy solids and the white let go, and it gets worse after ageing or after a heat process. | Residual solvent trapped under the ink film, starving the adhesive; ink not matched to the adhesive system for that structure; ink film too heavy, particularly the white; slip or wax additive at the ink surface acting as a release layer; low or decayed treatment on the print side; adhesive coat weight too low over a rough or heavy ink surface. | Peel a fresh sample by hand and look at where it actually fails, ink to film, ink to adhesive, or cohesively within the ink film, then read wetting tension on the print side before the coater. | Identify the failure interface first, because it names the fault. Failure at ink to film means treatment or wet-out, so correct the treatment. Failure at ink to adhesive means compatibility or slip contamination, so move to a lamination grade from the Arcova range and take surface slip additives out of the last-down colour. Cohesive failure inside the ink means residual solvent or too heavy a film, so fix drying and coverage before anything else. Raise adhesive coat weight over heavy white. Measure bond properly by T-peel, ASTM F904 or ISO 11339, on samples aged and processed the way the pack will be, not straight off the laminator. Where the structure is to be retorted or heat processed, settle the ink and adhesive choice with the Arctic technical centre and laboratory on trial laminates before the job is quoted, rather than assuming a grade will carry it. |
| After a stop, the first metres come out with missing image or a pale patch where the cylinder stood in the air, then it clears itself. | Solvent blend too fast for the hall temperature; cylinder parked out of the ink at the stop; dryer air still blowing back onto the cylinder while the press stands; shallow highlight cells drying first; viscosity already at the high end before the stop. | Where the cylinder comes to rest relative to the pan, and whether the dryer fans keep running on a stop. Both are visible in under a minute. | Set the press so the cylinders stay wetted at a stop. Jogging is the first choice. Raising pan level also works but feeds misting at speed, so if you raise it, put it back before the run and check the unit for spray. Shut dryer air down to the cylinder when the press stands. Move the blend slower to suit the hall temperature and the length of the stops you are actually getting. Do not correct it on restart by flooding the pan with solvent, that drops strength and pulls the colour off standard for the rest of the reel. |
| Fine ink spray around the unit, speckles on the web outside the image area and on the frame. It starts above a certain speed and stops when you slow down. | Viscosity too low; ESA voltage set too high, throwing ink out of the cells; pan level too high so the cylinder carries excess ink; blade set too far back from the nip so wiped ink is flung off; speed above the working window for a low-tack ink on that engraving. | ESA setting and pan level first, then cup viscosity. Drop the ESA one step and watch whether the misting stops, that separates an ESA fault from a viscosity fault in a minute. | Bring viscosity back up to the set-up value, then lower ESA to the minimum that still fills the highlight cells cleanly. Set pan level so the cylinder runs through a shallow, steady bath rather than a deep one. Move the blade closer to the nip within the press standard. Keep unit covers closed, keep extraction running and clean up promptly, misted solvent ink raises the solvent load in the hall and is a safety matter as much as a print defect. |
| A general haze, tone or scum over the areas that should be clean film, outside the image and in the margins. It wipes off the print but it comes back every revolution. | Blade pressure too low, or the blade worn to a flat land so it no longer wipes; blade angle out, or the blade holder not parallel across the width; blade lifted locally by dried ink build-up in the holder, or by a bent or dirty backing blade; viscosity too high, so the land areas are not wiped clean at speed; the cylinder land itself roughened, corroded or worn through the chrome, so it holds ink; ink drying on the land between units on a long run. | Look at the clean margin outside the image with a loupe and see whether the haze is even across the width or in bands. Even across the width points at pressure, angle or viscosity, banded points at the holder, the backing blade or the cylinder surface. With the press stopped and guarded, wipe the land with a clean cloth and see whether the ink comes away or is held in the surface. | Raise blade pressure in small steps to the lowest setting that gives a clean land, and no further, because over-pressure buys the streaking and scoring faults above. Fit a fresh blade if the edge has worn to a flat. Strip and clean the holder, check the backing blade for straightness and set the holder parallel, then reset the contact angle to the press standard, typically 55 to 65 degrees. Bring viscosity back to the set-up value. If the land holds ink after cleaning and the chrome is dull, rough or worn through, the cylinder is at the end of its chrome and needs re-chroming, no blade setting will cure that. |
| Bars or a repeating light and dark pattern across the web, evenly spaced, the same pitch every time, and worse at particular speeds. | Worn bearing or a flat on an idler or impression roller setting up vibration; blade holder loose or blade extension too long, so the blade chatters; impression roller out of round or with a hard spot; web tension oscillating between units; unbalanced cylinder or shaft. | Measure the pitch of the bars along the web and convert it to a circumference, that identifies which rotating part is at fault in a couple of minutes. Then, with the press stopped and guarded, take hold of the blade holder and try to move it by hand. | Match the band pitch to the circumference of the cylinder, the impression roller or an idler and service that part, do not chase it with ink changes. Tighten the blade holder and shorten the blade extension beyond the backing blade to the press standard, an industry-typical figure is roughly 1 to 3 mm. Stabilise web tension across the affected units. Changing speed may move the resonance out of the running window, but it does not fix the worn part and the fault will return. |
| Register will not hold in the web direction. It creeps out over a reel, or it hunts either side of the mark, and the control system chases it without settling. | Unwind tension varying, particularly across a splice or at the end of a reel; a dirty, dragging or seized idler adding drag into one span; film stretching with heat and tension through the dryers, worse on thin gauge; register control gain or set-up wrong for the speed and the repeat; drive, gear or shaft wear on one unit; an out-of-round or slipping roller giving a cyclic error. | Watch the register marks at speed and note the shape of the error: a slow one-way creep, a step that coincides with a splice, or a cyclic swing. Time the cyclic swing against a roller revolution. Then run a hand over each accessible idler with the press stopped and guarded to find one that does not spin freely. | Treat the three patterns differently. A step at the splice is unwind tension and splice preparation, so correct tension control and splice quality. A cyclic error whose period matches a roller is mechanical, so match the period to the circumference and service that roller, shaft or drive as for banding. A slow creep that worsens as the dryers come up is film stretch, so reduce web tension and dryer temperature towards the lowest that still dries the ink, and let the film stabilise before judging. Only when the web is running steady is it worth re-tuning register control gain. Do not dial in a standing register offset to hide a mechanical fault, it will move again at the next speed change. |
Flexo printing
Flexography lays down a very thin ink film from a raised image, so almost every fault on a flexo press traces back to three things: how much ink the anilox delivers, how hard the plate is pressed into the substrate, and how fast the ink dries in relation to press speed. Work through those three in that order before changing the ink. A start-of-run record of viscosity, anilox volume and screen, impression setting and solid density costs two minutes and turns most of the entries below into a comparison rather than a guess.
The checks here are press-floor checks: a loupe, a flow cup, a dyne pen, a densitometer, a plate thickness gauge, a steel rule and a torch. Where a figure appears it is an industry-typical range given for orientation only. It is not an Arctic measurement, a specification or a guarantee, and the working window for your own substrate, plate and press must be confirmed on press. Published standards are named only as the method for a property.
Three entries below look the same at the densitometer and are treated differently. Colour that will not come up from the first reel is a transfer or ink-choice problem. Colour that falls away within a run is usually dilution from viscosity top-ups. Colour that falls run after run on a roll that still looks clean is usually a plugged anilox. Read all three before cleaning a roll or re-inking a duct. In the same way, a plate that fills in, a plate that halos and a plate that has swollen in solvent all show as coarse, heavy print, so check plate thickness against the mounting record before blaming impression.
Two safety points before the entries: never draw a finger or a fingernail along a doctor blade edge, and stand the chamber off and isolate the station before touching a blade, an end seal or an anilox. When you need ink-side help, quote the family in use, Arcoflex for wide-web solvent flexo, Arcray for narrow-web UV and LED-UV, Arcfibre for paper, board and corrugated, Arcweb for label and shrink sleeve, together with the batch number on the container, to the technical centre in Greater Noida.
| Symptom | Likely causes, most probable first | Check first | Action |
|---|---|---|---|
| Print goes dirty after twenty to thirty minutes. Fine type and shadow dots fill in, the plate shoulders carry a ring of dried ink and the crew has to stop and wipe the plate every few reels. | Viscosity has crept up through solvent loss and the ink is no longer releasing cleanly, the most common cause; ink drying on the plate between impressions because the solvent blend is too fast or dryer return air is blowing across the plate; anilox volume too high for the plate screen so ink floods the dot shoulders; over-impression pushing ink off the dot; substrate dust, fibre or slip additive building on the plate; plate surface swollen or attacked by an aggressive wash solvent or an ester rich blend, which fills in the same way, see the plate swelling entry. | Take the station out of impression and look at the plate under a torch before anyone wipes it. Dried ridges standing around the dots mean the ink is drying on the plate. Wet ink flooding the shoulders means too much volume or too much impression. Then take a flow cup reading and compare it with the start-of-run figure, and at the next plate change put a plate gauge on the plate and compare it with the thickness recorded at mounting, so a swollen plate is ruled out before impression is touched. | Where viscosity has risen, bring it back to the start-of-run figure with the recommended thinner blend. Use retarder only where the evidence on the plate is drying in, and record every addition, because a duct corrected with retarder reel after reel ends up over-retarded and carries solvent into the reel, which shows up later as blocking and odour. Baffle or redirect any dryer air reaching the plate. Reset impression from first contact to a kiss. If the plate still fills, move to a lower volume anilox for that station and clean the anilox before restarting. Use only the wash solvent recommended for the plate type. |
| Dots print with a dark ring round them, small type has a fringe, and solids show a heavy outline at the edges. The job looks sharp on the proof and coarse on the press. | Over-impression squeezing ink out beyond the dot; anilox volume too high for the plate screen; plate and mounting tape combination too soft, so the dot flattens under pressure; ink over-reduced and too fluid to stay on the dot; worn or rounded dot shoulders on a used plate. | While running, back the impression off in small steps and watch a 50 per cent patch and a four point character through a loupe. If the halo shrinks as impression comes off, it is pressure, not ink. If the halo is there at the lightest setting that still prints, look at anilox volume and at the plate itself. A halo that arrives hours into a run on a job that started clean is more likely a swollen plate than a setting change, so check plate thickness against the mounting record. | Reset impression from zero contact upward until the image just prints, then stop. Fit a harder cushion tape for line and type work. Move the station to a finer, lower volume anilox. Typical practice is an anilox screen roughly four to six times the plate screen ruling, quoted here as common industry practice for orientation rather than as a fixed rule, and the combination must be confirmed on press. Return the ink to supplied strength and viscosity. |
| Solids on film show tiny clear holes like needle pricks, often in the same band across the web reel after reel. | A band in the same lateral position points to a localised source first: a dead or dirty segment in the corona electrode, a contaminated roller laying a stripe of oil or release, or oil and water carried in from the compressed air line. After those, general causes: slip additive or silicone migrating to the film surface; surface treatment low or decayed across the whole reel on stored film; ink surface tension too high for the film because the solvent blend is wrong; static on the web disturbing the lay. | Run a dyne pen across the web at the unwind and again just before the print station, at both edges and the centre, and mark the positions so a low lane is obvious. Treatment levels for polyolefin films are typically specified in the 38 to 42 dyn/cm band as general industry practice, not as an Arctic figure, a specification or a guarantee. ASTM D2578 is the reference method for wetting tension. Then wipe a clean cloth along each idler in the pinhole lane and look at it under a torch for oil. | Clean or repair the treater segment and the contaminated roller before changing anything else. Drain and filter the air line and check for oil mist near the unwind. Where treatment is genuinely low across the web, raise in-line corona only up to the film supplier's stated limit, since over-treating builds a weak oxidised surface layer that lifts with the ink and can block the reel, and bring in fresh, correctly treated film rather than drawing from old reels. Adjust the solvent blend toward a slower, better wetting component from the recommended list. Fit or clean the static bars and check the web earthing. Where treatment cannot be held on the film in use, send a sample of the film with the ink family and batch number to the technical centre in Greater Noida before the run rather than forcing the ink on press. |
| Solids look cloudy, grainy or like orange peel. The colour is there on the densitometer but the panel is not even, and it gets worse as speed rises. | Anilox screen too coarse, so the cell pattern is visible in the laid film; anilox volume too low for the ink strength and the film is starved; ink too fluid to level; solid printed from a plain unscreened plate surface; first dryer too hot so the film sets before it levels; absorbency variation in board or paper. | Look at the solid through a loupe at a low angle. A regular cell or screen pattern points to the anilox. Random cloudiness that follows the substrate points to absorbency or wetting. Check viscosity against the start-of-run figure and check the anilox for partial plugging with a clear tape lift. | Move to a finer anilox screen, but set the volume with the supplier rather than carrying the old volume across. A finer screen holding the same or a higher volume has cells that are deep relative to their opening, and those cells release worse and plug faster, so the mottle stays and a second fault arrives with it. Ask for a screen and volume combination that keeps the cell opening wide enough to empty cleanly at your speed. Use a surface screened plate for large solids and keep line and solid on separate stations where the layout allows. Reduce the first dryer setting so the ink can level and move the heat to the later dryers instead of raising it overall. On absorbent board, lay a base coat or a heavier first hit, and confirm with the technical centre that the Arcfibre grade in use suits the board. |
| A faint image of the heavy type or a bar appears further round in a solid on the same colour, repeating at a fixed distance every revolution. | Anilox starvation, where cells emptied by the heavy element have not refilled before the solid reaches the nip; ink feed rate or chamber pressure too low; ink too thick or too thixotropic to refill cells at speed; plate layout with a heavy element sitting directly in line with the solid; on the drying side, ink setting unevenly because of dryer air pattern. | Measure on the printed web from the ghost back to the element that caused it. Compare that distance with the print repeat from the job docket and with the anilox circumference taken from the roll specification or the end label, since you cannot get a tape round an anilox in a closed chamber while the job runs. A spacing that matches the anilox circumference is anilox starvation. A spacing that matches the print repeat is the plate layout or the drying. | For anilox starvation, raise the ink feed and chamber pressure so the chamber stays full, and bring viscosity down within the working range so the cells refill at speed. A finer screen anilox helps because the emptied band is less visible and the film is more even, not because the cells refill faster, so treat it as a second step after feed and viscosity. For a plate repeat, re-lay the plate so the heavy element and the solid are staggered across the cylinder rather than in line, or move the solid to another station. |
| Bands or bars run across the web at a regular pitch, strongest where a solid runs into open plate. The station rattles and the noise changes with speed. | Over-impression, the usual cause; plate and sticky-back combination with too little damping for the job; uneven plate relief, a bubble under the tape or a lifted plate edge; gap in the image round the cylinder giving the plate an impact every revolution; worn gears, bearings or a cylinder with runout. | Drop impression to the lightest setting that still prints and hold the speed. If the bars soften or go, it is bounce. Then change speed by a third in each direction: bounce moves and often disappears at a different speed, a drive fault does not. If the pitch matches the gear tooth pitch and stays regardless of impression and speed, it is the drive. | Reset to kiss impression. For bounce, move the tape to a softer, better damping grade. Do not fit a harder tape here: the hard tape rule belongs to dot gain and fine line work, and in a bouncing station it transmits the impact instead of absorbing it and makes the bars worse. Accept that a softer tape will add some dot gain, so reset impression and read the tone scale after the change, and compensate in the plate curve if the gain is repeatable. Re-mount any plate with a bubble or a lifted edge. Add a balancing bar or dummy relief outside the print area so the plate is never running with a bare gap. If the pitch follows the gears, take the station out and have the bearings and runout checked. |
| Fine lines run the length of the web in one colour only, always in the same place across the web, and they do not wash out. | A nick or damaged edge on the doctor blade; a hard particle trapped under the blade, usually a dried ink skin, a pigment agglomerate or substrate dust; worn chamber end seals letting debris in; blade pressure set too high and wearing the blade quickly; a chip or score in the anilox ceramic itself. | Stop the press, stand the chamber off and isolate the station. Wipe the blade and run the edge of a plastic card or a cotton bud along it to find the nick. Never draw a finger or a fingernail along a doctor blade edge, the blade is razor sharp and a damaged one more so. Then look along the anilox under a torch with magnification at the same lateral position as the line. Blade damage is found in seconds. Anilox damage stays after the blade is changed. | Change the blade and set pressure back to the lightest that wipes cleanly. Never raise blade pressure to bury a line, it wears the blade and the ceramic and the line comes back worse. Fit fresh end seals, strain the returning ink through a mesh filter, and keep a lid on open trays. If the score is in the ceramic, take the roll out and send it for repair or re-engraving, and log it so it is not reissued. |
| Density on that station drops slowly run after run. The anilox looks clean and bright but it will not give the colour it used to, and cleaning only partly restores it. | Dried ink packed in the cell base, the most common cause; UV ink polymerised in the cells through stray light or a delayed clean; over-aggressive or wrong chemistry cleaning leaving residue; heavily loaded white or metallic packing the cell floor; genuine wear on an old roll with the cell walls rounded off. | Press a strip of clear tape firmly onto the cleaned, dry anilox, lift it and look at it against a light. Plugged areas show as dull patches with no clean cell pattern. Compare the current solid density to the approved start-of-run reading on the same patch. Before stripping the roll, rule out the two faults that read the same on the densitometer: pull a draw-down of duct ink beside a draw-down of unopened ink, since a weak duct is dilution or contamination and not the roll, and check whether the shortfall was there from the first reel, which points to the entry on colour that will not reach standard. | Deep clean by the method approved for that ceramic and roll construction, then re-check with the tape lift. Shield the chamber from UV and daylight on narrow web stations running Arcray, and wash the anilox before any stop of more than a few minutes. Dedicate one anilox to white and one to metallics. Keep a volume and cleaning log per roll, have the volume measured when density drifts, and retire rolls on measured volume rather than on appearance. |
| The colour will not come up to standard from the first reel. Viscosity is at the bottom of the range, impression is as heavy as the crew dare and density is still short. | Anilox volume too low, worn or plugged; ink over-reduced with solvent or water during make-ready; wrong ink series for the substrate, so the film is soaking in rather than sitting on top; closed or over-calendered substrate surface; on UV, ink film too thin for the pigment loading. | Take a flow cup reading against the batch start figure using the same cup and the same method, ISO 2431 being the reference for flow time by flow cup, then pull a hand draw-down of the duct ink beside a draw-down of unopened ink. If the draw-downs match, the ink is right and the transfer is the problem, so go to the anilox. If the duct draw-down is weaker, the ink has been over-reduced. Note that this entry is for colour that was never there. Colour that starts right and falls away within a run is the ink strength entry, and colour that falls run after run is the anilox plugging entry. | Change to a higher volume anilox rather than chasing density with impression, which adds gain and halo without adding much density in flexo. Return the ink to supplied strength and keep top-ups as fresh ink. Confirm the family suits the job, Arcoflex for wide-web solvent work, Arcray for narrow-web UV and LED-UV, Arcfibre for paper and board. On absorbent stock, seal first or lay a base white. |
| The tonal scale prints heavy. Mid-tones close up, the vignette jumps rather than fades and highlight dots disappear within two steps of the tail. | Over-impression, the first thing to rule out; anilox volume too high for the plate screen, flooding the small dots; too much cushion in the mounting tape; ink too fluid and spreading on the dot; plate relief and dot shoulder geometry not matched to the process; on absorbent stock, spread into the surface after transfer. | Read the control strip with a loupe and a densitometer, then back impression off in small steps while watching the mid-tone patch and the highlight tail. Tone value is read from the control strip with a densitometer or spectrodensitometer and calculated by the method in ISO 12647-1, while ISO 12647-6 is the process control standard for flexographic printing and sets the tone value parameters. Two minutes of impression steps will separate pressure gain from ink gain. | Reset kiss impression from first contact. Match anilox to plate. Typical industry practice is an anilox screen of roughly four to six times the plate screen ruling, stated as common practice for orientation rather than as a rule, and confirmed on press. Fit a harder tape, unless the station is also bouncing, in which case fix the bounce first and take the gain out in the plate curve. Where gain is consistent and repeatable, have it compensated in the plate curve rather than fighting it on every run. |
| After a few hours the plate looks matt and swollen, the edges lift, the image has grown and the print goes heavy and feathery. The plate measures thicker than it did at mounting. | Ester or ketone rich solvent blend attacking the plate compound; wrong wash solvent used at the last clean; plate not fully post-exposed and detacked, so the surface is still soft; long continuous run with no rotation of plates; hot press environment accelerating the attack. | At the next stop take the plate off, or take the spare from the same set, and read it with a plate thickness gauge against the thickness recorded at mounting. A dial gauge on a plate still on the cylinder will not give a usable figure. Then check what has actually gone into the duct and onto the plate: the drum in use, the top-up record for the shift and the wash solvent on the station, all against the list recommended for that ink and plate compound. Blend composition itself cannot be measured on the press floor, so the record is the check. | Move to the alcohol and ester balance recommended for the ink and the plate compound and stop topping up with whatever is nearest. Use only the specified wash. For long solvent runs, fit a solvent resistant plate compound and rotate plates between runs so each one dries out. Re-expose and detack plates properly in the platemaking room, since a soft plate will always swell faster. A swollen plate also shows as dirty print, halo and register drift, so check the thickness before chasing any of those. |
| Register creeps in the web direction as the run goes on, worst after a splice or a speed change, and the crew is correcting every few thousand metres. | Web tension changing as the unwind roll diameter falls, or a tension step at the splice; substrate stretching in the dryers, especially on thin film and shrink sleeve; plate or tape creeping on the cylinder or a sleeve not fully seated; a plate swelling in solvent, which grows the image and moves register while the tension readouts stay steady, see the plate swelling entry; dancer or load cell calibration drifting; registration camera reading a contaminated or low contrast mark. | Watch the unwind, infeed and outfeed tension readouts through the moment the drift happens, and put a pencil line across the plate edge onto the cylinder so any plate movement shows up at the next stop. If the readouts are steady and the drift is in one colour only, measure that plate rather than rebalancing tension zones. | Rebalance the tension zones and set the unwind taper so tension holds as the roll runs down. Drop dryer temperatures to the lowest that still dries, and add air volume instead of heat. Re-seat the sleeve and use a tape with enough shear resistance for the run length. Clean the register marks and confirm the camera is locking onto them. On shrink sleeve with Arcweb, keep infeed tension and dryer heat low, since the film will take a set. |
| Foam builds up in the chamber and the tray on a water-based job. Print shows small round voids and speckle, and the level in the reservoir keeps climbing. | Air drawn in at a loose pump fitting or a cracked suction hose; ink level too low so the pump is pulling air; return line discharging above the ink surface and splashing; pump speed too high for the chamber; viscosity too low so the ink whips easily; contamination with wash water, alkali or a cleaner left in the line. | Look at the return line first. It should discharge below the ink surface. Then check the ink level and run a hand along the suction hose and fittings looking for a stream of bubbles in the ink. | Extend the return line below the surface, top the reservoir up so the pump is always flooded, and tighten or replace the suction fittings. Slow the pump to the lowest rate that keeps the chamber full. Correct viscosity with the recommended reducer rather than with plain water, since water alone drops the pH and the ink foams more readily as it does. Add defoamer only in small measured doses, recorded, since too much will cause craters and adhesion loss. Flush the system properly after cleaning so no detergent is left behind. |
| After a short stop the station prints patchy and light and only comes back after a full anilox clean. On a water-based job the blade has a dried bead on it within minutes. | Solvent blend too fast, or the water-based ink drying at the blade edge; dryer air blowing across the open chamber; ink not circulating at stops; worn or missing chamber end seals letting air in; low humidity and high ambient temperature in the press hall; heavily pigmented or fast setting ink left standing. | At the next stop, hold a hand near the chamber and feel for dryer draught, and look at the blade edge and the exposed anilox one minute after stopping. A skin forming that fast tells you the answer without any further test. | Keep ink circulating at every stop and stand the anilox off and wash it before any stop longer than a few minutes. Fit correct end seals. Baffle the dryer exhaust away from the chamber. Move the solvent blend slower with the recommended retarder, or on water-based work add the recommended retarder rather than more water. Keep a note of every retarder addition, since a duct that has been slowed repeatedly will retain solvent and can block in the reel. Where stops are frequent, plan a wash-off point into the run. |
| Colour strength falls away through a long run. Density drifts down over several hours, top-ups do not bring it back and the shade moves slightly as well. | Repeated solvent or water additions to hold viscosity, which dilute the pigment step by step, the most common cause by far; anilox gradually plugging; return ink contaminated with wash solvent, with another colour or with cleaner from the line; pigment settling in a reservoir that is not stirred; on UV, ink warming in the chamber so viscosity falls and transfer changes. | Compare the current flow cup time and the current solid density with the start-of-run record on the same patch, then draw down a sample of duct ink beside a sample of the unopened batch. If the duct draw-down is weaker at the same film thickness, the ink has been diluted or contaminated and the anilox is not the problem. If the draw-downs match and the print is still short, go to the anilox plugging entry before stripping any roll. | Top up with fresh ink, not with solvent, and keep a reserve of the batch for that purpose. Log viscosity and density every reel so drift is visible before it becomes a reject. Clean the anilox at a planned mid-run break on long jobs. Keep stirring or circulation running on whites and heavily pigmented colours. Isolate the wash line so no cleaner can reach the ink system while printing. |
| On narrow web UV or LED-UV the print rubs off at the rewind, the surface feels tacky or smells of unreacted ink, even though the lamps are on. | Lamp or LED output down through age, dirty quartz, dirty reflectors or a partly blocked cooling path; press speed raised without raising the dose; ink film too thick, especially opaque white and dense blacks, so the lower film never cures; pigment or substrate absorbing in the curing band; shutters or focus out of position; over-thick lay from a high volume anilox; wrong grade for the lamp type, since LED-UV and conventional UV curing bands differ. | Clean the quartz and the reflector, then assess cure itself with a solvent rub, ASTM D5402 being the usual practice for solvent rub resistance, or a thumb twist on the solid, at full speed and again at half speed. Do not use a tape pull to judge cure. A tape pull measures adhesion, ASTM D3359, and an under-cured film can pass it while still transferring at the rewind. If the solid cures at half speed and not at full, the dose is short. If it fails at both, look at film thickness and at the colour rather than at the lamps. | Log lamp hours and replace on measured output, not on appearance. Clean optics every shift. To get the job off now, drop the line speed until the rub test passes, and treat that as containment rather than a fix. To hold cure at production speed, add lamp capacity, reduce the laid film or split the lay. For dense whites, split across two stations rather than forcing one thick film. Match the Arcray grade to the lamp type in use and confirm the cure on the actual substrate before the run. |
| Ink lifts on a tape test or scuffs off at the slitter on treated film, even though the print looked clean and the solids read the right density. | Surface energy at the moment of printing too low, from decayed treatment on stored film, a dead segment in the corona electrode, or the untreated side of the web running to the print stations; slip or antiblock additive bloomed to the film surface; solvent retained in the ink film, from an over-retarded blend or too little dryer, so the film is still soft; over-treated old stock carrying a weak oxidised surface layer that lifts with the ink; wrong ink family for the film chemistry; oil from the air line or hand cream picked up at the unwind. | Dyne pen the printing side at the unwind and again at the print station, and confirm the treated side with the pen rather than the reel label. Then run a cross-hatch or X-cut tape test, ASTM D3359, on a print taken straight off the press and again on the same print after twenty-four hours. Adhesion that is poor on press and good the next day is retained solvent, not treatment. Adhesion that is poor on both is treatment, contamination or the wrong family. | Print on film treated within the film supplier's stated window and stop drawing from old reels. Where treatment is genuinely low, raise corona only to the film supplier's limit, since over-treating makes adhesion worse rather than better. Bring the solvent blend back to the recommended balance and give the ink enough dryer to leave the web dry at the rewind. Drain and filter the air line. Where the film chemistry is the issue, send a sample of the film with the ink family and batch number to the technical centre in Greater Noida rather than correcting on press. |
| Density is short on one side of the web and correct on the other, on one station, across the whole reel. Moving the job across the web moves the problem with the position, not with the image. | Impression not parallel, from an uneven setting on the two sides or a worn bearer; blade loading uneven across the chamber, or a blade worn at one end; anilox worn in the middle band where most work runs; chamber not filled evenly, usually a starved feed at one end or an air lock; plate or tape thickness varying across the cylinder; substrate thickness or treatment varying across the web. | Read solid density at both edges and the centre of the same patch and write the three figures down. Take a plate thickness reading at the same three positions at the next stop. If the layout allows, turn the plate end for end on a short trial. If the light side follows the plate, it is the plate or the tape. If the light side stays on the same side of the press, it is impression, blade loading or feed. | Reset impression parallel from first contact on both sides. Even the blade loading and change the blade if it is worn at one end. Feed the chamber from both ends or raise the flow so it stays full, and bleed any air lock. Re-mount the plate where thickness varies. Where the anilox is worn in a band, have the volume measured and retire the roll rather than lifting impression on one side to cover it, since extra impression there will bring halo and gain with it. |
| Fine hairs or whiskers trail from the edges of solids and type, and a light mist of ink settles on the guards and on the web either side of the image. It gets worse as speed rises. | Viscosity too high, so the ink strings rather than splitting cleanly at the nip; over-impression forcing a bead of ink to the image edge; static on the web and the plate drawing filaments out of the split; solvent blend too fast, so the ink bodies up between anilox and plate; anilox volume too high for the image; worn plate edges holding a bead of ink. | Drop the speed by a third and look at the image edges through a loupe. Whiskering that clears at low speed and returns at speed is an ink split problem, not a plate problem. Take a flow cup reading against the start-of-run figure, and check that the static bars are powered, clean and at the right distance from the web. | Bring viscosity back into the working range with the recommended thinner blend and hold it reel by reel. Reset kiss impression. Clean and re-earth the static bars and check the web earthing at the unwind. Move the blend slower with the recommended retarder where the ink is bodying up between anilox and plate. Drop to a lower volume anilox only after viscosity, impression and static have been ruled out, since less volume will cost density and can start a starvation fault instead. |
| The reel blocks. On unwinding, ink picks off onto the reverse of the web, solids look glossy on one side and dull in patches, and the effect is worse in the centre of the reel and after a warm night. | Ink not fully dry into the reel, from too much ink laid, too little dryer or too fast a rewind; a duct that has been repeatedly corrected with retarder, so the film is still releasing solvent at the rewind; rewind tension too high so the layers are pressed together; reel wound warm and stacked before it could cool; solvent or water retained in an absorbent stock; no slip or release in the top film; high ambient temperature and humidity in the store. | Feel the reel through the wrap an hour after winding. A reel still warm in the core has gone away with heat and solvent still in it. Unwind a metre from the outside and look for pick-off, then compare it with a sample from the same job taken directly off the press and left loose. Check the retarder additions on the shift record before changing dryer settings. | Slow the line or add dryer capacity until the web is dry and cool at the rewind, and lower rewind tension. Bring the blend back from any accumulated retarder additions. Let reels stand and cool on end before stacking or shrink wrapping. Reduce ink film where coverage allows. Where blocking persists, finish with an Arclear overprint varnish or an Arcglide slip grade chosen for the stock, and confirm on a stacked sample overnight before releasing the job. |
Coating and varnishing
This section covers faults that appear when an overprint varnish or a functional coating is applied, whether in line or off line, from an offset coater, a flexo coating unit, a roll coater or a screen unit. Each fault is written as a press crew would describe it, and the causes under it are ordered by how often they turn out to be responsible, not alphabetically. Work down that order rather than starting with the cause that is easiest to change.
Several of these faults look the same on the sheet but part in opposite directions once you know which one you have: craters from contamination against pinholes from foam, low gloss from a starved anilox against low gloss from absorption into the stock, a coating that is chalking because it never formed a film against a coating that is hard but never keyed down. Where that is the case the entry says what to look at first to tell them apart, because the corrective actions pull against each other and the wrong one makes the sheet worse.
“Check first” is always something that can be done at the press in a few minutes, with the sheet, the reel or the chamber in front of you, before anything goes to a laboratory. Where a figure appears it is an industry-typical range given for orientation only, not an Arctic measurement and not a guarantee: confirm the working window for the grade in front of you against its technical data sheet and your own trial. Where a standard is named, it is named as the test method for that property, and the method has to suit the substrate, since a film method and a paper method will not give the same number on the same job. Product families are referenced where the corrective action points to a particular type of coating, for example Arclear overprint varnishes, Arcglide for slip, Arcguard for mechanical resistance, Arcresist where chemical and scuff resistance are both required, Arcfibre on paper, board and corrugated, Arcprime under digital and Arcvista for decorative effects.
| Symptom | Likely causes, most probable first | Check first | Action |
|---|---|---|---|
| Orange peel: the cured or dried film has a fine dimpled texture like the skin of an orange, worst in solids and most visible under raking light | Coating too viscous or too cold at the applicator, so it stops flowing before it has levelled; the film set by drying or curing too close to the coater, before it had time to level; anilox screen too coarse for the coating weight, so the cell pattern is never fully lost; rubber roller or blanket hardened and no longer laying a smooth film; low surface energy under the coating, so the film beads slightly before it flows out, which normally shows craters as well and should be treated as the crater fault instead | Flow cup time and the temperature of the coating at the pan or chamber against the technical data sheet (flow cup to ISO 2431 is the method), then a coated sheet under raking light: a dimple pattern that repeats at the anilox screen angle is the engraving, a random dimple is flow and set | Bring the coating to its running temperature before the run rather than thinning to compensate, because a batch thinned while cold will be short of solids once it warms up. A circulating temperature in the typical 20 to 25 C region suits most water-based and UV coatings and the grade's data sheet decides. Then correct viscosity in small steps. Buy levelling time with what is on the console: switch off or turn down the first lamp in the bank and cure on the later lamps, or cut the temperature in the first air zone. Do not plan on moving a lamp that is fixed in the frame. If the pattern follows the anilox, move to a finer engraving at the nearest practical volume and hold the coating weight with solids, rather than asking for the same volume at a much higher line count: cells that deep relative to their opening release badly and plug early, and you trade orange peel for streaks and low gloss. Replace a hardened roller or blanket. |
| Ribbing or roller pattern: regular fine ridges running in the direction of travel, evenly spaced across the width, felt with a fingertip more easily than they are seen | Film split instability where the film leaves the last nip, with more coating in the nip than the speed will carry smoothly; on a roll coater, a roller speed ratio and gap sitting in the unstable region; viscosity too high, so the split fingers instead of drawing out; nip too light or not parallel across the width, which is the first lever on a roll coater and a minor one on a chambered anilox unit; roller rubber hardened, worn or out of round; on a chambered unit, engraving too coarse for the volume being laid | Change press speed by a clear step and watch the ribs: a film split pattern changes spacing or disappears, a mechanical pattern from a worn or out of round roller does not move. Then take a nip impression strip across the full width and confirm the band is even end to end | Reduce the applied volume first, since that is the quickest way out of the unstable region. On a roll coater, increase nip pressure one step to close the gap, reset parallelism if the impression band is uneven, and move the speed ratio away from the point where ribbing starts. Do not open the nip: a wider gap carries a thicker film and ribs harder. Lower viscosity slightly, or warm the coating so it flows after the split. On a chambered unit, leave the nip alone and move to a finer engraving at the nearest practical volume, and check the plate or blanket. Replace hardened, worn or out of round rollers. |
| Craters and fisheyes: round holes in the coating, often with a raised rim, the coating pulled back from a central point, sometimes only over certain colours | Silicone or oil contamination, from compressed air, chain or bearing lubricant, aerosol sprays, hand cleaners, or a slip additive carried over from a previous job; surface energy too low on the substrate or on the ink beneath, either untreated film or treatment that has decayed in storage; defoamer or an additive overdosed or incompatible, which is a common own goal after a foam problem the shift before; foam breaking at the surface and leaving a hole behind, which is the foam fault rather than this one; a particle or a skin in the coating holding the film off | A dyne pen or dyne solution across the area to see whether the surface wets (wetting tension on polyethylene and polypropylene film is measured to ASTM D2578; on coated board or over an ink film a dyne pen is an indication, not that method). Compare a printed area against a bare margin: holes only over the ink point to the ink surface or to slip in the ink, holes everywhere point to contamination or to the coating itself. Then look into the tank and change pump speed: a contamination crater has a raised rim and repeats in the same place on every sheet, a foam pinhole is clean edged, moves about and comes and goes with froth and pump speed. Look at the air line filter and drain, and ask what has been sprayed near the press | Remove the silicone source before anything else, because additive changes will not hold against a live contamination source. Re-treat or replace film that has lost surface energy, with surface-printed film typically worked in the 38 to 42 mN/m region. Add a wetting and flow additive in small controlled steps, typically a fraction of one percent, and stop as soon as the craters close, since more additive past that point brings foam. Filter the coating on the return line, clean the chamber, pan and pump, and skim any skin from the tank. |
| Poor holdout, the coating sinking into the fibre: gloss is dull and patchy on uncoated or absorbent stock, the varnish looks as if it has disappeared into the sheet, and unprinted areas look different from printed ones | Absorbent stock, uncoated board, kraft or recycled liner taking the coating into the fibre before it can form a film; coating weight too low for the substrate; solids too low, often from over-thinning with water; too long a path between the coater and the first dryer, giving the coating time to soak in; no sealing or priming layer under the varnish; ink films holding out while bare stock does not, which shows as a gloss difference between printed and unprinted rather than a loss of gloss overall | Put a drop of water on a bare part of the sheet and watch it: a drop that flattens and disappears within a few seconds is a sheet that will drink the coating, one that still stands after half a minute is not. Then compare gloss on a solid against gloss on the bare board | Seal the surface before varnishing rather than trying to bury the fibre with more gloss coating: a primer or sealer coat first, an Arcfibre grade built for paper, board and corrugated, or an Arcprime primer where the substrate is being prepared for digital. Raise the coating weight and the solids, apply a second coat wet on dry where the unit is available, and set the film as early as the press allows by bringing the first zone forward or raising it. Note that this is the opposite of what a levelling fault wants, so confirm with the water drop test that the stock really is absorbent before moving the dryer. |
| Low gloss: the coating is on the sheet at the right weight, but gloss reads below the standard sheet and the finish looks flat | Coating weight too low, usually because the anilox is plugged or worn rather than because the setting has changed; absorption into the stock, which reads the same on the sheet and wants the opposite dryer change, see poor holdout; drying or curing too fast for the film to level; matting agent carried over from a previous matt coating in the pan, pump or lines; ink beneath chalking or heavily pigmented, giving a rough surface for the varnish to sit on; the pile going away warm so the surface is disturbed before it hardens, which is normally the start of a blocking problem | Gloss at 60 degrees against a retained standard sheet (ISO 2813 is the method for the coating film; on uncoated paper and board, 75 degree gloss to TAPPI T 480 separates better at the low end). Then the anilox under a loupe for plugged cells, the coating consumption for the run against the normal figure for the job, and a water drop on a bare part of the sheet to settle whether the stock is drinking the coating | Clean or replace the anilox and restore the volume before changing anything in the coating. Then let the water drop test decide which way the dryer goes: on a sealed or coated stock, cut the temperature or the air speed in the first zone so the film has time to level; on an absorbent stock do the opposite, because slowing the set there only gives the coating longer to sink. Raise solids or coating weight, and keep gloss and matt on separate pans, pumps and filters, since a small carry-over of matting agent will pull gloss down across a whole run. If the ink beneath is chalking, fix that first, because no varnish will gloss over a powdery surface. |
| Streaks and scoring: narrow lines in the direction of travel where the coating is missing, light or heavy, often in the same lane all run | A nick or a wear point in the doctor blade; dried coating or a particle trapped under the blade; blade loaded too hard or set at the wrong angle, scoring the anilox; a damaged or worn band on the anilox; chamber end seal worn and bleeding coating into the lane; a damaged edge on the coating plate or blanket | Stop and isolate the unit, then move or reseat the blade: if the streak moves with the blade it is the blade, if it stays in place it is the anilox or the plate. Run a light along the blade edge to find the nick | Reseat or replace the blade and set it to the lightest pressure that still gives a clean wipe, since a heavy blade cures the streak today and scores the anilox for every job after it. Fit a return line filter to keep skins and dried particles out of the chamber, a mesh in the typical 100 to 200 micron region for a plain overprint varnish. Check the grade's data sheet before filtering a coating that carries wax or a matting agent, because a fine mesh will strip the dispersion out and hand you a slip or matt fault instead. Use blade oscillation if the unit has it, replace worn end seals, and clean the anilox properly rather than wiping the surface. |
| Edge build and picture framing: a thick glossy border around the coated area or at the sheet edge, the stack stands proud at the edges, the edge stays tacky when the centre is dry | The coating plate or blanket cut-out edge dragging a bead of coating and leaving it at the boundary; applied volume too high, so the surplus flows to the free edge; plate edge not chamfered, or the relief too shallow; nip set too heavy, squeezing the film outward; drying air weaker at the edge of the sheet or web, so the heavy film there stays wet; trailing edge squeegee effect at high speed | Feel the coated edge against the centre between finger and thumb on a fresh sheet, then look at the cut edge of the plate or blanket for a ridge of dried coating and check whether the edge has been bevelled | Chamfer or bevel the plate edge and re-cut the coating area with a small set-back from the trim so the boundary falls where it can be trimmed off. Reduce the applied volume, take one step of pressure off the nip, and raise the edge air in the dryer so the heavy edge is not the only wet part going into the pile. Where the pattern repeats every job, cut a relief channel into the coating plate just inside the boundary. |
| Blocking and set-off in the stack or the reel: sheets stick face to back and tear or lift the coating on separation, the reel shows a pattern transferred to the back of the web, or the pile picks on the feeder at the next operation | Stack or reel going away warm, so residual heat plus pile pressure softens a film that would otherwise hold; drying or cure incomplete, with water or solvent still retained in the film; coating weight too high; wax or slip additive from the ink beneath migrating into the coating; pile too high or reel wound too hard; high ambient humidity softening a water-based film overnight; a soft, flexible grade specified for creasing being asked to do a blocking job as well | Put a hand into the middle of the pile or against the reel core end and note whether it is still warm, then twist a thumb hard on a fresh sheet: if the finish marks, it is not hard yet and the pile will block | Cool the delivery and cut pile height and reel hardness before changing product, because most blocking is heat and pressure rather than chemistry. Where more drying or cure is genuinely needed, put the heat into the early zones and take it out at the end, and chill the sheet or web before it reaches the pile or the reel: more heat with the same delivery temperature blocks a stack faster than it hardens the film. Reduce coating weight and allow dwell before converting. Where the job needs more resistance, move to a harder Arclear overprint varnish grade, or add slip with an Arcglide grade so the surfaces release. Blocking resistance is assessed by an ASTM D3003 type method under load, temperature and time. |
| Poor slip: packs drag, sheets will not shingle or jog, cartons stall on the filling line, coefficient of friction reads above the target band | Slip additive depleted in the circulating system as the run goes on, or never added to this batch; wax settled in the pan or stripped out by too fine a filter; too little dwell for the additive to reach the surface, so an early sheet tests differently from the same sheet the next morning; a matt or textured finish gripping more than the gloss it replaced; coating weight too low for a complete slip surface; rough or uncoated stock gripping through a thin film | Coefficient of friction on a current sheet against the retained standard, with the method chosen for the substrate: ISO 8295 or ASTM D1894 on film, TAPPI T 815 or TAPPI T 549 slide angle on paper and board. At the press, slide two sheets face to back on a tilt board and note the angle at which they move, against a retained sheet from a run that behaved | Restore slip with an Arcglide grade or the matched slip addition for the varnish in use, and keep the tank stirred and recirculating so wax stays dispersed. Check the filter mesh, since a fine filter will strip a wax dispersion out of the coating within an hour and leave the pan looking full of a coating that no longer slips. Allow the specified dwell before judging the result, and do not mix batches at different slip levels part way through a run. |
| Too much slip: packs slide on the pallet, sheets stream in the delivery and will not jog square, the reel telescopes, glue joints fail or peel | Slip or silicone additive overdosed, often added twice by two shifts; a low friction grade specified for machine speed being run on a job that has to stack; coating weight high, so the slip surface is complete and unbroken; slip additive migrating up out of the ink beneath; coating running over the glue tab where it was meant to be kept clear; residual silicone in the pan or on the rollers from cleaning materials, which normally shows craters before it shows slip, so if the film is clean look at the additive dosing first | Coefficient of friction against the retained standard by the method for the substrate, then look at the glue tabs on a finished blank and at the plate cut-out to see whether the coating is running into the glue area | Change to the matched grade without added slip, or blend slipped and unslipped coating in a controlled ratio to bring friction back into the band, rather than trying to correct by dropping coating weight, which will cost gloss and rub resistance as well. Re-cut the plate so the coating keeps clear of glue tabs, and stop any silicone containing cleaner or spray being used near the coater. |
| Chalking and poor rub resistance: the coating rubs away as a fine powder, colour transfers to a thumb or to the next sheet, scuffing appears on the pack after transit | Water-based coating dried before the resin coalesced, from cold air, cold stock or flash drying without heat, so the binder never formed a continuous film; under-cure of a UV or LED-UV coating, from aged lamps, dirty reflectors, speed too high, or a photoinitiator that does not match the LED output; coating weight too low to give a continuous film; the ink beneath chalking and taking the varnish with it; matting agent concentrated at the surface; wax heavy ink below preventing anchorage, which usually shows as a hard film lifting rather than a powdery one, see poor adhesion | Thumbnail scratch and a hard rub on a fresh sheet as it comes off, against a retained sheet from the start of the run. For UV, lamp hours and a dose measured with the radiometer run through in the sheet or web path at production speed, since a static reading held under the lamp flatters the press. A thin overprint typically wants somewhere in the 80 to 150 mJ/cm2 region depending on the grade and the chemistry, and the grade's data sheet decides. Note whether the fault worsened towards the end of the run, which points to lamps or to a depleted batch | For a water-based coating, raise the dryer air temperature so the web or sheet stays above the coating's minimum film formation temperature: air volume alone will not form a film, and more cold air makes it worse. For UV, clean the reflectors, replace aged lamps and cut speed until the rub passes. Raise the coating weight, and where the job is genuinely abrasive move to an Arcguard grade for mechanical resistance, or an Arcresist grade where chemical and rub resistance are both required. Rub resistance is measured by a TAPPI T 830 or ASTM D5264 type rub test. |
| Poor adhesion: the coating is hard and unmarked but lifts cleanly off the ink or the film when tape is pulled, sometimes taking the ink with it, and it peels back from the cut edge after slitting or die cutting | Surface energy too low under the coating, either untreated film or treatment that has decayed in storage, so the film cured as a separate skin instead of wetting out; wax or slip additive in the ink beneath sitting between the ink and the coating; ink not dry when the coating went on, so the anchorage is to a soft film that fails later; the wrong coating for the substrate, for example a paper and board grade run on film, or no primer where the substrate needs one; a fully cured first pass being coated over on a second pass, with no free surface left to key to | Tape pull on a fresh sheet and on one that has stood, and note where the break is: coating off ink is an ink surface problem, coating and ink together off the substrate is a treatment or substrate problem. Then a dyne pen on a bare margin of film. Where a graded result is wanted, cross-cut to ISO 2409 or a tape test to ASTM D3359 is the method | Correct the surface first: re-treat film in line, or reject stock that has lost its treatment, with surface-printed film typically worked in the 38 to 42 mN/m region. Take wax out of the ink beneath rather than trying to make a coating stick to it. Give the ink enough dwell or interdeck drying before the coater. Where the substrate needs a keying layer, use a primer made for it, an Arcprime primer where the substrate is being prepared for digital, an Arcfibre grade on paper, board and corrugated. Do not raise cure to cure an adhesion fault, since a harder, fully cross-linked film keys worse, not better. |
| Foam and micro-bubbles in a water-based coating: froth on the pan or in the chamber, the tank level climbing, and pinholes or small craters in the dried film | Return line discharging above the liquid level and beating air into the batch; pump running faster than the chamber needs, so coating is churned and back pressure rises; air drawn in through a leaking suction fitting, or a tank level low enough to vortex; defoamer depleted through the run, or the wrong type for the coating; the batch over-thinned with water, which dilutes the defoamer and drops viscosity so entrained air is carried instead of released, and which can bring detergent in as well if wash water has been used; anilox cells carrying entrained air into the nip | Look into the tank: is the return pipe below the surface, and is the level high enough not to vortex. Then catch the coating coming off the anilox on a clean strip and see whether the bubbles are already there before the nip. Drop the pump a step and watch whether the holes in the film follow it, which separates foam pinholes from contamination craters | Submerge the return line and raise the tank level, then drop the pump to the lowest speed that keeps the chamber flooded, and check the suction side for leaks. Add defoamer in small controlled steps and stop as soon as the foam collapses, since overdosing defoamer causes craters and is the fault that usually follows this one. Let a foamed batch stand to deaerate before running it again rather than trying to run through the foam, and correct viscosity with fresh coating rather than more water. |
| Drip-off effect fails to separate cleanly: the matt and gloss contrast is weak, the gloss areas are dull, the matt areas have picked up gloss, or the boundary between them is fuzzy rather than sharp | The matt primer has dried or set before the gloss coating reached it, so there is nothing left to repel: usually an interdeck dryer or IR left switched on between the units; primer film weight too low, giving an incomplete repellent layer, or so high that it floods the boundary; gloss coating wetting the primer instead of beading, because a standard overprint varnish has been used in place of the matched drip-off partner; gloss coating weight too low to draw up into a bead; absorbent stock pulling the primer in; press speed changed mid run, which changes the interval between primer and coating | Confirm at the console that every drier, IR bank and air knife between the primer unit and the coater is switched off. Then throw the coating unit off, run a few sheets and touch the primer at the delivery: if it has already set there, it had set before the coater and there was nothing left to repel the gloss | Switch off all drying between the primer and the coater, since the effect depends on the primer still being wet. Raise the primer film weight in small steps until the matt area is fully covered, and raise the gloss coating weight so it has enough film to bead. Use the matched drip-off pair, an Arclear overprint varnish with its intended primer, or an Arcvista grade where the effect is decorative, rather than pairing whatever is in the pan. Where the stock is pulling the primer in, seal it first. Hold press speed steady through the run. |
| Mottle and uneven gloss over the printed areas: the coating is even but the gloss is blotchy over solids, worst on heavy ink coverage | Heavy spray powder holding the coating off the ink, or a coarse powder grade; ink film not dry or still setting when the coating went on, so the two mix or the coating is absorbed unevenly; different absorbency between printed and unprinted areas; anilox volume marginal, so coverage is only just complete and any variation shows; two-sided or uneven stock | Run a few sheets with the spray powder off and compare, then check the powder grade and the setting at the hopper | Cut the spray powder to the least that still stops set-off and change to a finer grade, since powder is the most common cause and the quickest to test. Watch the pile as you do it, because cutting powder too far brings set-off back. Raise the anilox volume so coverage is not marginal, give the ink more dwell or more interdeck drying before the coater, and where the stock is absorbent put a primer under the varnish. |
| Coating cracks, crazes or flakes on the crease or fold, showing a white line along the score | Creasing rule and matrix wrong for the calliper, so the board itself is cracking and the coating is only showing it; coating weight too high, so a thick film cannot follow the fold; a hard, high-gloss film with little flexibility; folding against the grain; material and shop floor too cold at the folding stage; too long between printing and converting for a UV film that goes on cross-linking; a UV coating taken well past the cure it needs into embrittlement | Fold a sheet by hand over the crease as it comes off and again after standing, and look at the back of the crease: if the board fibre is broken, the fault is the crease and not the coating, and no change of coating will hide it | Correct the crease rule and matrix first wherever the crease test shows broken board fibre. Otherwise reduce the coating weight, which is the safest single lever, and specify a flexible grade for jobs that crease. Bring board and shop floor to a normal converting temperature, typically in the 18 to 22 C region, and run creases with the grain where the layout allows. Trim UV cure only after all of that, and only down to the lowest dose that still passes the rub test, because an under-cured film will chalk, mark and block, and you will have traded one fault for three. |
| Coating weight uneven across the width: one side heavy and glossy, the other light or starved, consistent from sheet to sheet or metre to metre | Nip not parallel, or packing uneven across the width; blade loading uneven, tighter at one end than the other; anilox worn in a band from long narrow-width running; chamber starved at the end furthest from the feed port because pump flow is too low; substrate calliper profile varying across the web or the sheet | Take a nip impression strip across the full width and compare the band at both ends and the centre. Read gloss at three points across a coated sheet, and where no gloss meter is to hand cut equal strips from each side and the centre and weigh them against uncoated strips from the same pallet on the same scale | Reset nip parallelism and correct the packing before touching the coating. Balance the blade loading end to end, raise the pump flow so the chamber floods across the full width and returns freely, and plan narrow-width work across different bands of the anilox so wear does not concentrate. Replace an anilox with a worn band, since no setting will recover volume that is no longer in the cells. |
| Curl after coating: sheets bow towards or away from the coated side in the pile, board will not feed or jog square at the next operation, and the curl grows or relaxes overnight | Water-based coating applied to one side only, wetting and then drying that side so it shrinks against the uncoated back; drying too hard, taking the sheet below the moisture content it came in at; coating weight too high for the calliper; stock conditioned differently from the pressroom, or unwrapped too early; grain across the sheet rather than along it; a one-sided board whose liner and back respond differently | Lay a fresh sheet on a flat bench and note which way it lifts, then lay one that has stood overnight beside it: curl that relaxes overnight is moisture coming back into the sheet, curl that stays is set into the fibre. Check pressroom humidity and whether the stock was left unwrapped at the press | Cut the drying back to the least that still hardens the film and reduce the coating weight, since most coating curl is moisture driven out of one side. Move to a higher solids grade so less water goes onto the sheet, or coat or wet the reverse to balance it where the job allows. Keep stock wrapped until it is needed and let it reach pressroom conditions before printing, and lay off in smaller piles so sheets can equalise. Where the job will not tolerate curl at all, a UV coating puts no water into the sheet. |
| Matt coating glossing up: the matt finish shines where it has been handled, along the fold, on the running edges through the folder gluer, or in tracks across the pile | Pressure and rubbing flattening the matting layer, in the pile, at the folder or on the filling line; coating weight too low, so the matting layer is thin and easily burnished; film not fully hard when the pile went away, from a warm stack or a short cure; matting agent settled in the pan or the lines, so the end of the run carries less than the start; too much slip in the coating, so surfaces move against each other under load instead of sitting still | Rub the matt area firmly with a thumb on a fresh sheet and on one that has stood, then read gloss at 60 degrees on a rubbed and an unrubbed area of the same sheet against the retained standard | Let the film harden and cool before it goes under load: cut pile height, cool the delivery and allow dwell before folding and gluing. Raise the coating weight so there is enough matting layer to take handling, and keep the tank stirred and recirculating so the matting agent stays dispersed through the run. Where the pack is handled hard, move to a grade built for mechanical resistance, an Arcguard grade, or an Arcresist grade where scuff and chemical resistance are both wanted, rather than trying to hold a burnish-free matt with a soft film. A controlled slip addition from an Arcglide grade reduces drag marks, but too much of it lets packs slide and stream, so move in small steps and check friction against the standard. |
Sealing, lamination and adhesion
Sealing and lamination faults are often blamed on the ink or the coating when the cause sits in the machine settings, the film surface or the filling line. This section is ordered so the most probable and cheapest cause is tested first. Every fault gives one check a crew can do at the press, the laminator or the filler in a few minutes, before anyone cuts samples for the laboratory. Arctic’s sealing coatings (Arcseal), foil lacquers (Arcfoil) and barrier coatings (Arcbar) sit inside these structures, and the same diagnostic order applies whether the failure is in a coating, in an adhesive or in the sealant film itself.
Three habits make the rest of the section work. First, record the failure mode and not only the number: a bond that splits down the middle of the adhesive and a bond that lifts cleanly off the film are different faults with different fixes. Second, remember that several of these faults present identically from opposite causes. A seal can fail because it is too cold or because it is too hot, and a hot tack failure looks the same on either side of the hot tack peak, so where the entry says to step a setting, step it in both directions and read the trend before deciding which way to move. Third, treat every figure quoted here as an industry-typical starting range rather than a setting for your job, and confirm the real window on your own film and your own machine by a stepped trial. Where a property has a published method, the method is named (for example ASTM F88 for seal strength, ASTM F1921 for hot tack, ASTM D1876 or DIN 53357 for bond peel, ASTM D2578 for wetting tension, ASTM F1929 for dye penetration leak detection) so results can be compared between shifts and between suppliers.
One rule runs through the whole section: temperature is the last thing you raise and the first thing people reach for. Heat is quick, it flatters a weak seal for one shift, and it costs window at the other end in lock-up, burn-through and lost hot tack. Closure, pressure and dwell come first.
| Symptom | Likely causes, most probable first | Check first | Action |
|---|---|---|---|
| Seal strength is weak or varies across the web width: some packs pull hard, others open in the hand, from the same reel and the same settings | One jaw zone running cold from a failed heater element, a drifting or loose thermocouple, or a loose heater band; dwell too short for the speed the line has been pushed to; jaw pressure uneven because the jaws are worn, misaligned, or the toggle or spring pack is tired; PTFE tape worn through, torn, or built up with baked residue; sealant layer thinner than specified across part of the width, or a reel of the wrong sealant grade; web tension drawing the sealant thin at one edge. A machine cause and a film cause look the same in the reject bin, so use the pattern to separate them: a weak band that repeats in the same web position on seal after seal is the jaws, while weak seals scattered at random with no relation to jaw position are the film surface, that is contamination, bloom or a thin sealant layer | Run a set of seals right across the jaw width, number them in web order and pull them by hand, then repeat on a second set and see whether the weak positions repeat. Then stop the machine, open and isolate the jaws, and read each zone at the jaw face with a surface probe or a temperature indicating strip, comparing the reading with the controller display | Set every zone by the measured jaw face temperature and not by the controller display, and replace any heater or thermocouple that will not hold. Re-shim or re-align the jaws until a carbon impression or pressure indicating film shows even contact end to end, and replace worn or built-up PTFE. Where more seal strength is genuinely needed, add dwell by reducing speed or fitting a longer seal bar before adding temperature, because extra temperature costs window at the other end: lock-up on a peel seal, burn-through on a light gauge, and loss of hot tack on the line |
| Seal will not peel: the peel seal has locked up and the film tears or delaminates instead of opening at the seal | A weld grade sealant loaded in place of the peelable grade, which is first on this list because it costs nothing to rule out and because it is the usual answer when the fault starts at a reel or job change; jaw temperature above the top of the peel window, so the peel layer has fused; dwell too long, which acts on the seal in the same way as extra temperature; jaw pressure too high; on a peelable laminate, ink, varnish or adhesive carried into the seal band, which locks the structure so the failure moves into the laminate instead of the peel interface | Read the grade and batch on the roll label against the job specification before touching a setting. Then seal a short series of samples with the setpoint stepped down in small increments, mark each with its temperature and peel them by hand at the machine. A peel face that opens cleanly once the setpoint comes down is heat and dwell; a face that is still smooth, glossy and fused at the bottom of the peel window is the grade, and no setting will recover it | Confirm the grade first. If the grade is right, drop the setpoint in steps until the seal peels cleanly and repeatably, record the temperature at which lock-up begins as the top of your window, and set the job in the middle of the window and not at the top. Reduce dwell and pressure before touching temperature again. Peel peel-seal samples at room temperature and not straight off the jaws: a seal pulled while still warm reads low and stringy, which is a testing error and not a fault, and it has sent more than one crew after a problem that was not there |
| Seal peels too easily, or packs open in transit and handling although the seal looks continuous | The reel threaded or wound the wrong way round, so the sealant is not face to face, which gives almost no strength and cannot be corrected by any setting; jaws not closing fully because of a linkage, an air supply fault or an interlock, or pressure set too low; dwell too short at the speed being run; jaw temperature below the seal initiation range for that sealant; slip and anti-block additive bloomed onto the sealant face, or dust and powder on it; sealant surface aged or oxidised in long storage | Confirm the sealant is face to face: unwind a short length, check reel orientation and thread against the job sheet, then fold a piece of the film sealant to sealant and seal it on the jaws with the machine empty. If that sample is strong, the film is capable and the fault is closure, pressure or dwell. Read jaw closure with pressure indicating film or a carbon impression before changing any setting. A sample sealed on the laboratory bench sealer at a known temperature, pressure and dwell is a good second step, but it is not the first one | Correct the thread or reel orientation if it is wrong. Restore full jaw closure and set pressure, then add dwell by reducing speed, and only then raise the setpoint in small increments until strength plateaus. Temperature raised first hides a closure fault and brings its own faults: burn-through, distorted seals and loss of hot tack once closure is later restored. To separate bloom or dust from the settings, wipe a sample of the sealant face with isopropanol, let it flash off and seal it again. If that seal is strong, the surface is the problem, and the reel or the way it is stored has to change. Wiping proves the diagnosis, it does not fix a running reel |
| Channel leakers: the seal looks complete but packs leak, and a leak test shows a fine channel running through the seal band | Product, powder or liquid trapped in the seal at the moment of closing; a wrinkle or pleat folded into the seal, most often at a gusset or at the fin seal crossover; jaw face nicked, scored or carrying a lump of cured residue; knurl or seal profile too coarse for the film gauge, so the channels are never closed out; jaws misaligned so pressure falls away at one end of the bar | Hold suspect seals up to a strong light and sight along the seal band for a bright line running through it, then squeeze a few filled packs under water at the machine and watch for a stream of bubbles from the seal band. Dye penetration is the laboratory confirmation, ASTM F1929, not the press-side check | Clean and dress the jaw faces and remove any burr or baked residue. Correct film tracking and the forming shoulder so no pleat can enter the seal, and re-set the crossover so the fin seal is not doubled under the cross seal. To close out knurl channels, add dwell first, then change to a finer profile or a flat land where the film gauge is light, and raise pressure last: a coarse knurl at high pressure on a thin film cuts or thins it, and a channel leaker becomes a burst seal |
| Seals survive the line but fail after retort or pasteurisation: seals open, blister or lose most of their strength | Sealant or adhesive not intended for the process, where industry-typical conditions are around 121 C for retort and around 70 to 95 C for pasteurisation; seal made at the bottom of the window, leaving no margin once the structure softens in the vessel; laminate processed before the adhesive has fully cured; moisture or product trapped in the seal, which steams under heat; pack overfilled or over-pressurised so the seal carries load while hot; poor bond in the laminate that only shows once wet and hot | Pull seals from the same reel before and after processing and compare the failure mode on each, not only the number, then check on the reel ticket how many hours elapsed between lamination and the vessel against the adhesive's stated cure time | Move the seal setting to the middle of the window and re-qualify after processing, not before. Hold laminate for the full stated cure at the stated store temperature before it goes anywhere near a vessel. Check the supplier's technical data for the sealant, the adhesive and any coating in the structure, including Arcseal, against the process actually being run, and qualify the finished structure by your own processed-pack trial rather than on a data sheet alone. Reduce fill volume or headspace so the seal is not loaded while hot, and do not trade cure time against a production date |
| Delamination: the laminate separates in the reel, at the slitter or in the finished pack, and bond values come back low | Incomplete cure from a wrong mix ratio, a tired or out of date hardener, or a cold store; solvent retained in the adhesive or in the ink beneath it; treated film below the wetting tension needed at the moment of coating; adhesive coat weight too low, or laid down unevenly across the width; heavy ink coverage, or a wax or silicone bearing ink, sitting at the lamination interface; nip temperature or pressure too low, so the adhesive never wets out | Peel a strip by hand from the reel edge and again from the centre, and look at which surface the adhesive has stayed on. Then check the mix ratio actually recorded at the mixer, by weight, the hardener batch and date, and the pot life clock on the batch in use | Work in this order. Verify the mix ratio by weight and not by volume, verify the hardener batch and date, and discard any batch past pot life. Clear retained solvent by raising drier air volume and exhaust and by reducing line speed rather than by raising drier temperature, because extra heat skins the adhesive and traps solvent underneath. Dyne test the film immediately before the coating head, and hold reels at the stated cure temperature for the full stated time before slitting. Only when cure, solvent and wetting are all confirmed, raise coat weight one small step and re-measure bond by ASTM D1876 or DIN 53357. Coat weight raised first flatters the number for a day, adds retained solvent, curl and blocking, and hides the real cause |
| Tunnelling: long bubbles or ribs running down the web in the laminate, worst near the core of the reel | Tension mismatch at the nip, most often one web stretched while the other runs slack, so the stretched web recovers after winding and the other buckles; two webs of different modulus or gauge combined and then held under tension on the rewind; reel wound too hard, so the inner turns cannot relieve; adhesive with too little green tack at the nip to hold the webs while they relax; a change in store temperature or humidity after winding | Draw a line across both webs with a marker just before the nip, run a few metres, then stop and look at the line on the laminate: the web that has moved is the one carrying the wrong tension. Feel or measure reel hardness from core to outside | Reduce tension on the stretched web to the minimum that still tracks, and taper rewind tension down towards the outside of the reel. Do not reach for nip temperature as the first move: extra nip heat softens the lighter web and lets it stretch further, which is the root cause you are treating. If green tack really is short after tension is right, address it with the adhesive specification and a slower rewind rather than by heating the nip. Let reels condition in the same room as the laminator before slitting, and avoid pairing a heavy web with a very light one at full tension |
| Curl: the laminate cups or rolls up on itself, so it will not lie flat on the former, the filler or the labelling head | Tension imbalance between the two webs, the same root as tunnelling, where the web that was stretched at the nip recovers and goes short; unequal shrinkage after the drier, with the hotter or thinner web pulling short; solvent retained in one web, plasticising it and letting it relax later; adhesive coat weight high on a thin structure; paper or board taking up or losing moisture after lamination | Cut a 100 mm square from the reel, lay it on the bench and note which way it cups. The laminate curls towards the web that has gone short. Then separate tension from heat: if the square curls more after a few minutes in a warm place, heat shrink is involved; if it was curled straight off the reel and does not change, look at tension | Reduce tension on the web the laminate curls towards, because that is the web that was stretched and has recovered short, and raise tension on the other web only if the line will not track otherwise. Where the cause is heat, reduce drier temperature on the side that is shrinking instead of moving tension. Clear retained solvent by slowing the line or raising drier air volume rather than raising temperature. Condition paper and board webs to the room before laminating, keep finished reels wrapped, and reduce coat weight where a thin structure is being overloaded with adhesive |
| Poor adhesion on treated film: ink, varnish or adhesive lifts cleanly off the film with a tape pull, and wetting looks patchy on press | The reel treated on one side only and the untreated side running, or the web threaded so the treated face is on the wrong side, which is free to check and common at a job change; the in-line treater off, tripping, or with a worn electrode or a dirty back-up roller; treat level set for a different film than the one running; corona treatment decayed in storage, which happens faster in a warm store and with a long stock time; slip and anti-block additive bloomed to the surface, masking the treatment; back-side transfer of treatment in a tightly wound reel | Dyne test the intended side with wetting tension solutions or pens at both edges and the centre, and at both ends of the reel, immediately before the coating or printing head, and pull a tape test on a printed or coated sample at the same time. Industry-typical practice on polyolefin films is to want the surface at roughly 38 to 42 dynes/cm before printing or laminating, and the method is ASTM D2578. Read the two together: a good dyne reading with a failed tape pull means the surface is bloomed or you are treating the wrong face, because corona will oxidise migrated additive and give a reading that the bond does not honour | Run in-line corona ahead of the head and set it by dyne test rather than by amperage, checking the reading again after any speed change. Use the oldest reels first, store film below the film maker's stated temperature and off warm floors. If the low reading, or the good reading with a failed tape pull, is caused by additive bloom, change the reel: boosting treatment over a bloomed surface gives a reading that fades within hours and a bond that was never there |
| Coating or adhesive will not key to aluminium foil: it beads, crawls or lifts cleanly, often in patches or from one end of the reel | Residual rolling oil left on the foil because annealing was incomplete; annealing uneven along the reel, so one end is clean and the other is not; foil contaminated in store by an oily atmosphere, an oil mist from nearby machinery, or bare-handed handling; reel very old or unwrapped in store | Water-break test at the press: take a foil sample, flood it with clean water and watch. A clean surface holds a continuous film for several seconds, an oily surface breaks into droplets almost at once. Back it up by wiping with a clean white cloth and isopropanol and looking for a stain | Quarantine the reel, water-break test the other reels of the same batch before they are loaded, and reject or return what fails. Record the water-break result on the goods-in check for every foil delivery. Do not try to wipe a running web clean: wiping a sample proves the diagnosis, it does not fix the reel. Specify fully annealed foil with low residual lubricant for lacquer and adhesive work, including where Arcfoil lacquers are being applied, and keep foil wrapped, off the floor and away from oil mist |
| A bond or a seal has failed, but nobody can say whether the fault is the adhesive, the coating or the surface underneath it | Adhesive split down its own middle, leaving material on both faces, points to cure: wrong ratio, tired hardener, short cure time, cold store, retained solvent, or a coat weight heavy enough to hold solvent in; adhesive all on one web with the other web clean points to that web's surface: low wetting tension, rolling oil, slip bloom, or a wax or silicone bearing ink; ink lifting with the adhesive points to ink-to-film adhesion and not to the adhesive at all; the film itself tearing or splitting means the bond is stronger than the substrate, which is a pass and not a fault | Peel a strip slowly by hand and look at both faces in a raking light, then touch each face with a fingertip for tack. Wet, tacky material on both sides is cohesive; a dry, clean face on one side is adhesive. Peel slowly, because a fast pull can tear a laminate that would have peeled and disguise the mode | For cohesive failure, work on cure: verify the mix ratio by weight, the hardener batch and date, the cure temperature and time, clear retained solvent, and check that coat weight is not so high that it cannot dry through. For adhesive failure, work on the surface: dyne reading at the head, cleanliness, foil oil, and the ink or varnish at that interface. Record the failure mode alongside every bond and seal test result so the trend, and not just the last number, is visible to the next shift |
| Seal contamination at the filling line: product, powder or oil in the seal band, giving weak seals, leakers and rejects that only appear once the filler is running | Filler dribbling or stringing as the nozzle lifts; fill level too high for the seal position, so product splashes into the seal band; dosing too fast for the pack, throwing product up the wall; powder dust carried into the seal area by static or by the air displaced as the pack fills; no jaw wipe, deflector or air blow fitted on a dusty or oily product; oily product creeping along the sealant face | Stop the filler and run twenty empty packs at the same speed and the same settings. If those seals are clean and strong, the fault is the filler and not the sealing station | Fit or re-aim the seal-area deflector or air blow, and add suck-back or a drip tray to the nozzle to stop stringing. Slow the dose, lower the fill level, and ionise the forming tube to stop dust being pulled into the seal. Put jaw cleaning on a timed schedule rather than on demand. Where the product is unavoidably oily, specify a sealant that seals through contamination, but treat that as a margin and not as a substitute for a clean fill |
| Hot tack failure: the seal is sound once cold, but it opens or creeps while still hot as the product weight lands on it | Machine speed raised so the product now lands before the seal has cooled; cooling jaws switched off, blocked, scaled, or set too warm; seal temperature below the hot tack initiation temperature for that sealant, so the seal is barely formed when it is loaded; seal temperature above the hot tack peak, so the seal is still molten when it is loaded; dwell too short, so the seal is not formed through the full sealant thickness; a sealant grade with a narrow hot tack window run at a speed it was not chosen for. Too cold and too hot fail in exactly the same way on the line, and the reject alone will not tell you which it is, so the temperature has to be stepped in both directions before anything is concluded | Run a stepped temperature ladder at the jaws, five or six settings spanning the current setpoint both up and down, and load each seal the way the machine loads it: open the jaws and within about a second lay a light weight on the seal or let the pack drop. Mark each sample with its setting. If creep gets worse as the setting goes up, the job is above the peak; if it improves as the setting goes up, the job was below the window and dropping the setpoint will make it worse | Set from the ladder, just below the peak found on the machine, and confirm in the laboratory by ASTM F1921 where the job justifies it. Do not drop the setpoint on a hot tack complaint until the ladder has shown which side of the peak you are on. Restore and clean the cooling jaws and check chilled water flow and temperature. Add dwell where the seal is not formed through, delay the product drop, or reduce speed so the seal has time to set, and give the pack a supported path so the seal does not take the full weight while hot |
| Seal strength falls away exactly where print or varnish runs into the seal band | Ink or varnish acting as a barrier between the two sealant faces; ink not fully dry, so retained solvent is trapped in the seal; heavy white or metallic ink with high pigment loading in the seal area; overprint varnish overlapping the seal band; register drift carrying a design that was never meant to reach the seal into it | Pull seals from a printed area and from an unprinted area of the same web, side by side, and compare. A clear difference points at the print and not at the jaws | Keep print and varnish out of the seal band at artwork stage, with enough tolerance for normal register drift, and mark that clear band on the proof. Where ink must cross the seal, choose a sealable grade and qualify it by a stepped seal trial. Increase drier capacity or reduce speed so no solvent remains, and agree register drift limits with the press crew before a long run |
| Blocking: the coated or laminated web sticks to itself in the reel, tears or lifts coating when unwound, and shows shiny transfer marks on the reverse | Reel wound before the coating has cooled or cured, so heat is stored in the reel; coat weight higher than needed; slip additive not yet migrated to the surface on a fresh reel; rewind tension too high, holding the layers in intimate contact; reel stored warm, stacked, or stood against a hot wall or in direct sun | Unwind a metre by hand at the machine and listen. A tearing or tacky release sound, with a shiny mark transferred to the back of the web, is blocking. A crackle with the web clinging and no transfer mark is static, which is a different fault with a different fix | Cool the web on chill rolls before the rewind and check that they are actually cold. Reduce rewind tension and taper it towards the outside of the reel, and reduce coat weight to the minimum that gives the required property. Store reels standing in a conditioned store, not stacked and not against a warm surface, and allow fresh reels to stand before slitting where a slip coating such as Arcglide needs time to develop |
| Burn-through and distortion: holes, thin spots or a wavy, shrunken seal band, film sticking to the jaws or stringing as they open, often appearing a shift after somebody raised the temperature to cure a weak seal | Jaw temperature above the top of the sealing window for that structure; dwell too long at low speed, or a seal left in the jaws at a stop; PTFE tape missing, torn or bedded in, so hot metal touches the film; jaw pressure too high for a light gauge film, particularly with a coarse knurl; a thin or unbalanced structure with too little heat resistant outer layer for the setting being run; jaws not opening cleanly, so the film is dragged while still soft | Sight along the seal band against a strong light for thin patches and pinholes, and look at the jaw faces for melted deposit while they are cool. Then step the setpoint down in small increments and pull seals at each step. If seal strength holds while the distortion disappears, the job was above the window and never short of heat | Bring the setpoint down to the middle of the window found by the stepped trial, and recover any lost strength with dwell and even pressure rather than with heat. Renew PTFE tape and clean baked deposit off cool jaws. Reduce pressure on light gauge films and fit a finer profile or a flat land. If the required strength cannot be reached without distortion, the structure or the sealant grade is wrong for the speed the line is being run at, and that is a specification question and not a setting |
| Bond or seal strength is acceptable at the laminator but falls in storage once the pack is filled, and the loss is worst at the product line and least in the headspace | Aggressive fill attacking the bond: surfactants, essential and spice oils, solvents, acidic or alcoholic products; adhesive not chosen for that fill or for that shelf life; retained solvent or amine migrating to the interface and softening the bond; heavy ink coverage at the lamination interface, which is the weakest plane an aggressive fill can find; warm storage, which speeds up all of the above; the original number taken before full cure, so the value recorded at the laminator was never the final value | Hold filled packs and empty packs from the same reel side by side in the same store, then peel both. If empty packs hold and filled packs fall, the fill is involved and the laminator settings probably are not. Note where on the pack the bond has gone: loss at the product line and not at the headspace carries the same message | Take the failure to the adhesive specification with the fill, the fill temperature and the shelf life wanted, and re-qualify with a filled pack trial rather than with empty laminate. Reduce ink coverage at the lamination interface where the design allows it, clear retained solvent by air volume and line speed, and store filled packs at the temperature the shelf life was set at. Record cure time and test time with every bond result, so an early number is never read as a pass |
| The cross seals are sound but the back seal leaks: a lap seal that opens, or a fin seal that lets air through at the crossover | On a lap seal, sealant being asked to bond to the reverse face, which is print, varnish or an untreated outer layer rather than sealant, so the pairing can never reach the strength of a fin seal; the back seal bar set independently of the cross jaws and never re-set when line speed changed, so temperature, pressure or contact length is short; forming shoulder and tube alignment letting the overlap wander, so overlap width varies down the pack; the fin doubling under the cross jaw, so the cross seal has four thicknesses at the crossover and is under-heated exactly there; back seal bar worn, bowed, or not parallel to the tube | Cut a back seal open along its length and look at overlap width and at the crossover, then run a strip of pressure indicating film or take a carbon impression under the back seal bar and look for contact along its full length. Both can be done in a few minutes with the machine stopped | Set the back seal station in its own right, temperature, pressure and contact length, and re-set it whenever line speed changes. Correct the shoulder and tube so overlap width stays constant. On a lap seal, confirm the outer surface at the overlap is one the sealant can bond to and keep print and varnish out of that band at artwork stage. At the crossover, add dwell or fit a relieved jaw profile so the extra thicknesses get the heat they need, rather than raising the temperature for the whole seal and losing the window everywhere else |
Curing and drying
Curing and drying faults are usually faults of energy: how much reaches the ink, where in the film it reaches, and how long the film has to give up solvent or water before the next nip. Most of what follows is diagnosis rather than reformulation. Before changing an ink, establish whether the film is short at the surface, short at depth, or short overall, because the three have different causes and different cures.
Two faults in this section present identically at the press and have opposite cures. Ink that lifts on a tape pull may be undercured at the base, or fully cured and poorly stuck down. Look at what comes away: soft, smeared or tacky ink on the tape is undercure, and dose is the answer; a hard film lifting cleanly in whole squares, over print that passes a rub test, is an adhesion fault, where more dose usually makes it worse. Settle that question before touching the lamps.
Where a number appears below it is an industry-typical range quoted to give a starting point, not an Arctic measurement and not a guarantee: every figure has to be proved on your press, with your laid-on weight and your substrate. Record the dose, air temperature and speed that passed the job, and treat that record as the setting, not the power step on the console. Test methods named here (ISO, ASTM) are the recognised methods for the property in question; the acceptance level is yours to agree with your customer.
| Symptom | Likely causes, most probable first | Check first | Action |
|---|---|---|---|
| The cure is wrong but nobody agrees where. One operator says the surface is tacky, another says the whole ink film lifts off the substrate in a sheet. | A tacky surface over a hard base is oxygen inhibition or too little short-wavelength energy at the top of the film; a hard skin over a soft base is too little long-wavelength penetration, from heavy laid-on weight or a pigment that absorbs and scatters at the cure wavelength; soft throughout means total dose is short, not spectrum; a film that is hard right through and still lifts is not a cure fault at all but an adhesion fault, and dose will not fix it; on solvent and water systems the same skin over soft base pattern means a surface skin has formed and sealed the film before the carrier could leave. | Two checks at the press, side by side on the same print. On UV and LED-UV work a solvent-wetted cotton bud rubbed lightly for ten double rubs (ASTM D5402 style) reads the surface: it dulls or smears if the top is short. Do not use a solvent rub on solvent-based or water-based ink, because a thermoplastic film redissolves whether it is dry or not: use a dry fingernail or cloth rub and a sealed jar smell test instead. A cross-cut and tape pull (ISO 2409, or ASTM D3359 on flexible film) reads the base. Then look at what comes away on the tape, because this is what separates undercure from an adhesion fault: soft, smeared or tacky ink is undercure at depth; a hard film lifting cleanly in whole squares, over print that passes the rub, is an adhesion fault and belongs with the treatment and primer entry below. | Treat a surface fault as irradiance and oxygen: clean quartz or LED windows, restore the working distance, divert product and web air off the cure zone. Treat a depth fault as penetration: reduce laid-on weight, split heavy colours over two stations with a cure between, or move the job to a cure with more long-wavelength output. Treat an all-over fault as dose: raise power steps or lamps rather than dropping press speed, and re-record the setting that passes. Do not add dose to a film that is already hard, because that is an adhesion fault and more cure usually makes it worse. |
| Print feels tacky or greasy on the surface straight after cure, marks with a fingerprint, but scrapes off hard and the base is sound. | Low irradiance at the ink surface, from a fouled quartz sleeve or LED window, wrong lamp height or a low power step; product or web air blowing across the cure zone from a cooling fan, air knife or badly set extraction hood, pulling oxygen into the film surface; thin films of overprint varnish where the oxygen-affected layer is a large fraction of the whole coat; press hall well below the 20 to 25 C most UV inks are formulated to run at; a photoinitiator package short on the surface-curing component. | At the next stop, with the lamp shuttered, look at the quartz or LED window against a white card and feel for air movement over the cure zone with the back of a hand at a safe distance. Then divert or switch off only the product or web cooling air blowing across the print for one pass and rub test again: if the tack clears, it is oxygen, not dose. Never shut the lamp cooling air or the extraction to make this test: a mercury lamp run without its cooling air overheats, the quartz devitrifies and the lamp can fail. | Clean the window and reflector, restore the focus distance and step the lamp up. Redirect or duct away product air crossing the cure zone, leaving lamp cooling and extraction at the machine specification; where the machine allows it, an inert hood or nitrogen blanket removes the fault outright. On Arclear overprint varnishes, a small increase in laid-on weight reduces the proportion of oxygen-affected film. Bring the press hall and the ink up to the working temperature before running. |
| Opaque white lays down well and the top surface cures, but the ink under it stays soft and can be pushed off the substrate with a thumbnail. | Titanium dioxide scatters and absorbs strongly in the band the cure system works in, so very little energy reaches the base of a heavy white; laid-on weight of the white above what the cure can reach through; a mercury lamp with its long-wave output down, or a narrow-band LED at a wavelength the white transmits poorly; several whites laid wet on wet before a cure station; heavy solids of deep black or blue behave the same way. | Take a printed and cured sample and run it through the same cure a second time at the same speed, which doubles the dose without changing the laid-on weight. If the base hardens, the cure is short on dose. If it stays soft after a second and a third pass, the energy is not reaching the base at all and the fault is depth, so no amount of extra power on the full weight will fix it. Where a lighter anilox or sleeve is available at the stop, a half-weight strip cured at the normal setting gives the same answer. | Split the white over two stations with a cure between them, and use the heaviest weight only where opacity genuinely demands it. Put the white on a station served by a lamp with good long-wave output. Ask for the Arcray grade formulated for through-cure rather than for surface speed where the job is a narrow-web opaque white. Slower running adds exposure time but not penetration, so treat a lower speed on the white station as a small last resort, not as the fix. |
| An ink that cured on the mercury press will not cure on the LED press, or only cures at a crawl. Whites and deep blacks are the worst. | The ink is a conventional UV grade, whose photoinitiator absorbs in the short wavelengths an LED array does not emit; the pigment absorbs at the LED wavelength, carbon black and some organic blues in particular; laid-on weight too heavy for a narrow-band system; a mixed set on the press, part LED grade and part conventional, after a colour was drawn from the wrong shelf. | Read the cure system printed on the ink label and compare it with the array wavelength on the machine plate. This is a two-minute check and it settles the question: a conventional grade on an LED array is a specification fault, not a press setting. If the label does match the array, the fault is weight, dose or pigment and belongs with the through-cure entry above. | Run the grade supplied for the cure on the machine. Arcray is supplied in grades matched to LED-UV as well as to conventional UV. Do not try to make up a spectral gap with dose or with slower running: the film may pass a rub test and still carry uncured monomer, which comes back later as odour, blocking or adhesion loss. Where a heavy black or white must run on LED, reduce laid-on weight and split it across stations. |
| The same job at the same speed and the same power setting cured fine a month ago and now needs the press slowed to pass a rub test. | Mercury lamp near end of life, where output falls well before the lamp fails visibly; quartz sleeve fogged, etched or solarised; reflector dulled or coated with ink mist and varnish; power supply not delivering the set output; extraction pulling so much cold air past the lamp that the envelope runs below its working temperature. | Read the lamp hours against the changeover figure and look at the quartz and reflector at the next stop, holding a white card behind them. A greyed sleeve or a matt reflector is visible to the eye. If a radiometer puck is available, run one pass at a fixed speed and compare with the figure logged when the lamp was new. A loss that tracks lamp hours is ageing; a loss that tracks the time of day or the weather is an ambient fault and belongs with the drift entry below. | Change lamps on measured dose, not on hours: run a puck at a fixed reference speed weekly and log it per lamp position. Clean quartz and reflectors on a schedule rather than on a complaint. Where the airflow past the lamp is suspect, have it set back to the machine specification rather than throttled by hand, because under-cooling fails lamps. Do not let a permanently reduced press speed become the compensation for an ageing lamp, because the same drift will reach the next job unannounced. |
| One band down the web, usually in the same lane on every job, cures softer than the rest. | A light or heavy lane in printing rather than in curing, from a worn band on the anilox, uneven impression or a marked doctor blade, which fails a rub test in the same pattern; a failed or derated diode bank in the LED array; coolant flow low or coolant warm at one segment, so junction temperature rises and output falls; the window fouled with ink mist or varnish over that section; web flutter or a slack lane changing the working distance; the array shimmed unevenly, so the air gap is larger on one side. | Take a full-width strip of cured print and rub test it at five points across the web, and at the same time compare density or laid-on weight at those five points, because a lane that is printing heavy will fail the rub without anything being wrong with the cure. If the weight is even and the rub is not, sight along the LED window for a dull or dirty section and put a hand on the coolant lines to feel for a warm segment. | Rule out the anilox, impression and blade first. Then clean the window and confirm coolant temperature and flow against the machine specification. Log cure across the width weekly so drift is caught before it reaches a job. Have the array output mapped and the bank replaced if the lane is still short after cleaning. Note that output falls away quickly as the working distance grows: typical narrow-web LED arrays are quoted around 8 to 20 W/cm2 peak irradiance at 385 or 395 nm, and a few millimetres of extra gap is enough to lose a rub test. |
| Cure passes every check during make-ready and fails as soon as the press reaches running speed. | Dose is irradiance multiplied by exposure time, so doubling the speed halves the dose and make-ready proves nothing about the run; lamps on a lower power step than the job record; only part of the lamp bank switched in; a shutter partly closed or an interlock holding a lamp at standby. | Do the rub and tape tests at running speed, taking the sample on the fly and marking it with the speed it was taken at, and read the power step and the number of lamps on at the console against the job record before you do anything else. | Set and prove the cure at the intended run speed, then record the dose that held it as part of the job setting. Add a lamp or a power step rather than accepting a permanently slower press. Where the press cannot make dose at speed, move the heaviest laid-on colour to a station that has a cure immediately after it. |
| White and clear coats take a yellow cast after cure, worse at the reel core and worse on heavy coat weights. | Two opposite mechanisms produce the same cast and have to be separated. Undercure yellowing: residual photoinitiator, its breakdown products and amine synergist left in the film, worst in a heavy coat or under a deep colour, and it fades as dose rises. Overdose and heat yellowing: too much short-wavelength energy on a thin varnish, plus infrared from mercury lamps held in a tightly wound reel, and it deepens as dose rises. A third source is the substrate, where an optical brightener knocked back by UV exposure reads as yellowing of the print. | Cure a step wedge of the same varnish at three power settings and read the three against unprinted stock within a few minutes, then read the same wedge again at twenty-four hours, because the cast moves over the first day. Read the direction of the trend, not the absolute: cast falling as dose rises is residual initiator and amine, that is undercure, and the answer is more dose; cast rising as dose rises is over-irradiation and heat, and the answer is less dose and less infrared. If the unprinted stock has yellowed alongside the print, the brightener in the substrate is at least part of it. | Cure to the lowest dose that passes rub, adhesion and the twenty-four hour retest, and never below it: dropping dose to chase colour leaves monomer in the film and buys odour, blocking and adhesion loss instead. Where the trend shows undercure, raise dose and reduce coat weight rather than changing the ink. Ask for a low-yellowing grade of Arclear where the work is a clear over white or a white on a bright stock. Judge the cast by measurement, CIELAB b* or yellowness index (ASTM E313) against a retained standard, rather than by eye under press lighting, and cool reels before they go into store. |
| The reel smells of ink or varnish when it is unwound, and the odour carries into what is packed. | Uncured monomer and free photoinitiator left in the film, which is undercure under another name; thin colours cured while a heavy white or varnish underneath was left short; retained solvent from an earlier station rather than the cure; the web wound warm, so volatiles are trapped between the laps; ink returned to the drum after standing open. | Unwind a metre, cut a piece and seal it in a clean glass jar, conditioned warm before it is opened, industry-typical practice being 40 to 60 C for a set time, then smell the headspace against a jar of unprinted substrate conditioned the same way. A jar left cold for ten minutes is too weak a screen and will pass reels the customer rejects. This separates the print from the reel and from the press hall. | Treat odour as a cure fault first: measure the dose, raise it, re-run the same job and re-test. Cool the web before the winder and do not stack reels warm. Do not release or reject on a sniff test alone: send retained samples for headspace gas chromatography and agree an acceptance level with the customer for the pack in question. Arctic makes no claim of food-contact status, migration performance or third-party approval for any product, and that assessment sits with the converter and the brand owner. |
| Print is hard and passes a rub test, but tape lifts it cleanly off the film, sometimes only the day after printing. | Substrate surface energy too low, or in-line treatment decayed during storage, which is the most common of these by a wide margin; corona applied to the wrong side of the web; no primer where the film needs one; cure so fast that the ink had no time to wet the surface; overcure, where the film has cross-linked hard and shrunk away from a low energy surface, which is the one to suspect when everything else checks out and the dose is at the top of the range. | Take a sample from the reel at a stop and run dyne test inks or pens across the web before the first print unit (ASTM D2578 style), reading it straight away, since handled and stored film reads lower than it prints. Then take a cross-cut tape test (ISO 2409, or ASTM D3359 on flexible film) at the press and repeat it on the same sample an hour later and again at twenty-four hours. Confirm the ink is hard: if it rubs soft, this is undercure at depth and not this fault at all. | Treat in line to the level the film needs rather than relying on treatment applied at the film plant, and confirm the treater is working on the printed side. Use an Arcprime primer where adhesion depends on it. Where a step wedge shows adhesion improving as dose comes down, run the lowest dose that still passes rub and the twenty-four hour retest, because more cure is not more adhesion and past a point it is the reverse. Always re-test adhesion after twenty-four hours, since cure shrinkage faults commonly take that long to appear. |
| Reel smells of solvent at the winder, or a laminate made from it shows fine bubbles and a weak bond a day later. | Dryer air temperature or volume short for the laid-on weight and the speed; a slow solvent or retarder added to cure a drying-in fault upstream and never taken back out; speed raised without raising the dryer; deep cells and heavy solids laying on more solvent than the dryer was set for; exhaust restricted or the supply and exhaust balance wrong, so the dryer air is already saturated; cold substrate entering the dryer. | At the last dryer before the winder, read the supply air temperature and take a web temperature straight after it, using a contact probe or a temperature-indicating strip on the web. An infrared thermometer is not reliable here: on thin transparent film it reads through the web and on metallised stock it reads the surroundings, so if one is all that is available, read it on a heavy printed solid with the emissivity set for it and treat the figure as indicative. Then seal a freshly wound sample in a jar and smell it against blank film, and check the solvent addition log for the shift while you are there. | Restore the dryer to the setting for this speed and coverage: typical practice on solvent gravure over filmic substrate is final dryer air in the 40 to 70 C band, with the web brought up towards that temperature but held below the substrate's distortion or shrink onset point before the last nip, and the dryer running slightly negative so vapour is carried away rather than recirculated. Take retarder back out once the real drying-in fault is fixed. Prove the result by headspace gas chromatography and set a house limit with the customer, trade figures commonly quoted in the low single-digit mg/m2 range for total retained solvent, which is a trade figure and not a specification of any Arctic product. |
| Water-based ink prints cleanly but is still wet at the next unit or at the rewind, and the press has to be slowed to hold it. | Water takes roughly four to six times the energy per kilogram to evaporate compared with common press solvents, depending on which solvent the comparison is against, so a dryer set for a solvent job is short by a wide margin; dryer air already humid, from a humid press hall or from recirculated exhaust; laid-on weight too heavy because the anilox volume is coarser than the colour needs; a non-absorbent substrate, film or coated board, so the whole water load must leave through the air; viscosity corrected with plain water rather than the supplied reducer, adding water that then has to be removed; air velocity at the web too low, which matters more than air temperature. | Read the air temperature at the dryer nearest the fault and feel the exhaust, then check the anilox volume against the job record and take a hygrometer or dew point reading in the press hall. All three take minutes. | Raise air velocity at the web before raising temperature, because stripping the saturated boundary layer does more than heat does. Confirm the exhaust is vented outside and not recirculating. Drop to the finest anilox that still holds the colour. Correct viscosity with the reducer supplied for the Arcoflex or Arcfibre grade. On absorbent board an infrared boost ahead of the air dryer helps; on heat-sensitive film keep infrared low and add air instead. |
| A skin forms on the ink in the duct or tray during a stop, and dried crust appears on the anilox lands and around the chamber blades. | UV ink exposed to stray light in the press hall, from the cure station, an inspection lamp or daylight through an open shutter, starting polymerisation in the tray; solvent or water ink drying at the free surface in moving air from a dryer or an extraction hood; a long stop with the anilox standing and the chamber under air; oxidative skinning in a varnish left in an open duct; ink returned to the drum already carrying gel specks. | Put a piece of the skin or a crust speck in the matching solvent or in water and work it with a spatula. A dried skin softens and breaks back down; gelled UV ink stays as hard rubbery specks and will not redissolve, which tells you it is stray light and not air. Then, at the next stop, look at where light falls along the ink path and hold a hand in the air stream over the duct and the chamber, and look at returned ink in the drum for gel specks before it goes back on the press. | Shield the whole ink path from UV, inspection lamps included, and use UV-filtered glazing in halls running UV work. Keep chambers closed and circulate ink through stops. Cover open ducts and trays. Wash the anilox on any stop beyond a few minutes rather than letting it dry. Filter returned ink and never put ink carrying skin or gel back into a working drum. |
| The reel picks or blocks on unwind, print pulls off onto the reverse of the web, worst at the core and on heavy solids. | Incomplete cure, so the film is still thermoplastic and gives under pressure; the reel wound warm, since heat plus nip pressure is the blocking test itself; winding tension too high, particularly at the core; slip not yet developed because the print underneath it never cured, so an Arcglide product cannot do its job; a coating or treatment on the reverse of the film interacting with the print. | Press two pieces of freshly cured print face to back under firm thumb pressure for a few seconds and try to part them, then take a core temperature on the reel as it comes off the machine with a contact probe, since an infrared reading on a wound film reel is unreliable. | Fix the cure first and the winding second: cool the web ahead of the winder, taper tension from core to outside and let reels reach room temperature before stacking. Confirm the fix with a blocking test under defined load, temperature and time, ASTM D1146 or ASTM D3003 style, or a house method written down and kept, rather than with a thumb test alone. |
| Thin polyethylene, shrink film or light PET ripples, necks in or goes out of register at the cure station, and it is worse at low speed and on stops. | Infrared load from a mercury lamp reaching the web; the web standing still under a lamp or dryer during a stop because a shutter or interlock did not act; the chill roller not in contact, not cold, or absent under the cure; heavy laid-on weight forcing a high dose onto the stock least able to take it; dryer air set for a different substrate and never changed back. | Take a web temperature immediately after the cure station at run speed with a contact probe or a temperature-indicating strip stuck to the web, not with an infrared thermometer, which reads through thin clear film and gives a low and false number. Then watch what the reading does through a one-minute stop. On shrink film compare it with the shrink onset temperature on the film data sheet. | Move heat-sensitive work to LED-UV, which puts far less infrared into the web, or to a lamp with a dichroic reflector over a chilled drum in contact with the web. Prove that shutters close and lamps step down on a stop by making a stop and watching. Cure at the lowest dose that passes rub and adhesion. On shrink sleeve work with Arcweb, set the cure so peak web temperature stays below the film's shrink onset with a working margin, and check it again whenever the speed changes. |
| Print passes rub and tape flat on the press, then cracks along a crease, a fold or a score, or flakes at the die cut and on the slitter. | Overcure and high cross-link density, so the film is hard and brittle, which is the usual result of chasing every other fault in this section with more dose; total laid-on weight of ink plus varnish too high over the fold; a hard scuff-resistant grade used where a flexible one is needed; board run dry, or creased against the grain, or a worn creasing rule, so the substrate gives no support to the film; on film, a print cured hard over a substrate that still stretches under the knife. | Fold a cured sample by hand and over a straight edge, and look at the fold with a loupe, then repeat at twenty-four hours, which is when a brittle film is at its worst. Cure a second strip of the same job one or two power steps down and fold that: if the crack closes up at the lower dose, it is cure and not the substrate or the rule. | Cure to the lowest dose that passes rub, adhesion and blocking, not the highest the press can deliver. Reduce total coat weight over the folded area, or keep varnish off the crease where the design allows. Ask for a flexible grade of Arclear where the work is creased or folded, or Arcguard where the job needs scuff resistance and flex together. Check the creasing rule and the board grain direction before blaming the ink, and do not answer a fold crack with more dose. |
| Cure and drying hold on the first reels and go off through the shift, or settings that held all winter will not hold in the monsoon. | On water-based and solvent work, hall temperature and humidity rising through the day, so the dryer draws in air already carrying moisture and the drying rate falls; ink warming in circulation, so viscosity falls and laid-on weight rises; substrate coming out of a cold store in the morning and out of a warm hall in the afternoon; reducer or retarder added through the shift and never recorded; on UV work, extraction and hall air changing as other machines start and stop. This is not lamp ageing, which moves over weeks and not over hours. | Log hall temperature and relative humidity, ink temperature and viscosity, dryer and cure settings and the rub or jar result at the start, middle and end of a shift for a week. A fault that tracks the clock or the weather is ambient; a fault that tracks lamp hours or a puck reading is ageing; a fault that arrived with one reel is a substrate fault. The log answers all three and takes minutes a shift. | Fix the air first: vent exhaust outside, stop recirculation, and draw dryer supply air from a controlled source where the building allows it. Control ink temperature rather than correcting viscosity all shift, and record every addition. Condition substrate in the hall before printing. Set the dryer and cure for the worst case of the day and hold that setting, rather than chasing it reel by reel. |
| Water-based work on board or corrugated dries at the surface but the sheet or web curls, cockles or loses flatness, and register wanders on the later colours. | Water taken into the fibre and driven out unevenly, so the two sides of the board finish at different moisture contents; one dryer driven hard instead of moderate drying spread over the run; heavy laid-on weight from an anilox coarser than the colour needs; board arriving dry from store, so it takes water up fast and gives it back fast; infrared set high on an absorbent stock; flute and liner already at different moisture content on the incoming corrugated. | Read board moisture with a handheld meter at the unwind or infeed and again after the last dryer, and note which way the curl goes: curl towards the printed side usually means that side has been dried hardest, curl away from it usually means water has gone in and not come back out evenly. | Put less water on: use the finest anilox that holds the colour and correct viscosity with the reducer supplied for the Arcfibre grade rather than with plain water. Spread the drying across the dryers available instead of driving one, and keep infrared low on stock that is already dry. Condition and store board flat in the hall before the run, and let printed stock stabilise before die cutting or gluing. |
How the product is meant to be applied
Many of the faults above start as a mismatch between the product and the process. The application guide sets out what governs film weight on each process and what we need to know to specify correctly.
Send us the fault
Where a fault cannot be resolved on press, send a sample of the substrate, the ink or coating and the faulty work together. Testing any one of them alone usually explains nothing.