A roll comes off the oven. You cut a metre and weigh five points across the width. The three in the middle sit close together and the two edges drift. On the next roll the drift moves to the other side. The first instinct is that the machine is not accurate enough. Work out the direction first, and half the search disappears.
Variation runs in two directions. Across the width it is called cross-web. Along the length it is called down-web. The causes barely overlap. A wide spread across the width usually sits in the rolls or in the metering gap. A spread along the length points at the pump, the tension, or the stability of the slurry. Without splitting the two, you end up feeling your way around the machine.
The most common cross-web pattern is thick in the middle and thin at the edges. A roll has some give in it, and load bends it. The gap in the middle opens up wider than at the ends, so more slurry passes through the centre. Two fixes are used. One is to grind a crown into the roll face, slightly larger in the middle, to pre-load the bend back out. The other is internal counter-support, or splitting the loading into several zones so the roll stays straight under load. How much crown is ground depends on your width and your slurry viscosity. Change the product and the thickness profile changes with it. When you compare coating machine manufacturers, the crown value is one of the numbers you can ask to have written into the technical agreement.
Thick at the edges and thin in the middle sends you to the two ends of the gap first. If the ends of a comma blade or a metering roll do not agree, the delivery is not flat. A nick or a flat spot along the blade leaves a fixed band across the width. This kind of deviation is stable, and it sits in the same place tomorrow as it does today. That is how you separate it from a slurry problem. How many millimetres of edge get trimmed is another item a coating machine manufacturer can put in writing.
Surface tension is at work at the edges as well. Slurry tends to pull toward the centre, so a ridge builds up along each edge. That ridge runs well above the middle, and it is normally trimmed off as waste. Most lines set that trim width in the low tens of millimetres, and the exact figure follows the surface tension of the slurry and the surface energy of the substrate. Include that band in your thickness readings and the cross-web number inflates several times over.
Parallelism matters too. If the roll faces are not parallel, thickness across the width walks off as a slope, thick at one side and thin at the other. If roll runout is high, thickness rises and falls once per revolution. Both values have a factory figure, and you can ask for them to be measured in front of you at acceptance. Rolls are the hardest thing on the machine to fix on site. Regrinding means shipping them out and back, with your line down in between. When an equipment supplier hands the machine over, whether those two sheets exist tells you something on its own.
The metering method decides how sensitive the machine is to all of the above. A comma blade meters through a gap. Wet film comes out at roughly half to sixty percent of that gap, and the ratio moves with viscosity and speed. The gap is the only control, so the process is sensitive to roll flatness, slurry rheology, and temperature. A slot die is pre-metered. Output is set by pump speed, and the gap only keeps the bead connected. It does not set thickness. Put the same slurry through a slot die and the cross-web profile usually improves a lot, at the price of more sensitivity to pump pulsation. When an equipment supplier configures the coating head, those two effects are calculated together.
Down-web problems are quickest to chase with a period. Weigh a continuous stretch along the direction of travel and note the distance between peaks. If the period matches the circumference of a particular roll, the problem is on that roll. If it matches the pump rotation, the problem is in the feed. If there is no fixed period and the drift is random, the cause is usually in the slurry or in bubbles. Inside the feed system, a gear pump meters accurately but pulses, and the pulse shows up directly as a periodic thickness wave. A single screw pump delivers continuously and shears the material less, which suits shear-thinning slurry, at the cost of some metering accuracy. This is a decision that should be made for you when the feed system is specified. Ask a machine builder which pump they use and why. A vague answer usually means rework later.
The same pump behaves differently depending on the pipe run. Long, narrow piping loses pressure and the flow starts to hunt. Trapped air makes the coating come in waves. A supplier who works in this field can usually quote both routes, and the two numbers rarely land in the same place.
Slurry changes on its own. Solids content drifts a little from batch to batch. A few degrees of temperature takes viscosity down a step. Settling leaves the top of the tank leaner than the bottom. The gap never moved, and the thickness walks anyway. The scene that repeats in every plant is the first bucket of the morning disagreeing with the last bucket of the afternoon, with nothing changed in between. Stirring the tank, lagging the pipework, and checking temperature and solids before feeding are all cheap, and they remove a whole class of drift. When a coating line supplier works up your proposal, whether they ask about settling and temperature swings is a good sign for what comes later.
The substrate carries its own tolerance. Cross-web thickness tolerance on film and foil is not always small enough to ignore. Coat weight is a difference, so a thick spot in the substrate reads as a thick spot in the total. Weigh a blank sample of the substrate first. Surface energy and tension govern how the slurry spreads, and where surface energy is low the slurry pulls back and the coating runs thin. Uneven corona treatment leaves bands that line up with the electrode positions. Give the web width and the thickness tolerance at specification stage, and the proposal a coating line supplier returns will be far more specific.
Drying magnifies whatever is already there. If one side of the web dries faster, the coating on that side shrinks earlier, and the finished film ends up thinner on one side than the other. The test is simple. Measure the same set of points before the oven and after it. If the spread is already there before the oven, the problem is in the coating. If it is flat before and uneven after, the problem is in the drying. A coating line manufacturer usually holds the original drying curves, and the cross-web air velocity distribution is written into them.
How you state the tolerance matters more than the number you state. A percentage on its own guarantees an argument. Write the measurement method in with it. How many points, where they sit, how many millimetres of edge are trimmed, wet film or dry film, weighing or a gauge. Spell those out and acceptance day loses most of its disputes. On the same machine, two different methods can report spreads that differ several times over. Some manufacturers hand over a single number and no method, and the negotiation goes nowhere. Send the same slurry to two coating line suppliers and the tolerance they promise often differs. On this point, a coating line manufacturer who has run projects in Europe or North America will raise the method before you ask. One who has never been asked tends to wait for you.
If you want to make this easy, ask for four numbers at quotation stage. The committed runout and face parallelism, how cross-web thickness tolerance is held, which pump feeds the head, and whether there is any cross-web thickness feedback. Put those four questions to a coating machine manufacturer and the depth of the answers tells you a lot. Ask the same four of two different suppliers and they come back at very different depths. One OEM worth copying runs a roll of your slurry before dispatch, takes nine points across the width, and attaches the data to the acceptance report. Not many machine builders do it. The ones who do tend not to need chasing afterwards. Whether a coating line manufacturer will put those two commitments into the contract is visible before you sign.
The really stubborn cases are rarely one broken part. They are several small quantities stacking up. A roll with a little runout, slurry temperature drifting a little, substrate tolerance off a little. Each one sits inside its own limit, and together they put a pattern on the surface. Freeze the variables one at a time and you usually find the one being magnified.
The last roll comes out of the oven. You cut a sample and weigh five points across it. When those numbers sit close together, every sheet you measured earlier starts to earn its keep.
Published by Innater Intelligent Equipment Co., Ltd, a maker of non-standard coating machines and automated lines. Email 13316631159@163.com | Website https://www.getinnater.com/ | Tel +8613316631159
