A customer sent the same batch of coating to three coating line suppliers last year. Three pails, one recipe. Two weeks later he came to me with three reports, all of them pretty, all of them quoting tolerances tighter than he had asked for.
He wanted to know which one to trust.
I told him to ignore the tolerance row and look at the line speed row.
Two of the three had left it blank. One had filled it in. And the one that filled it in also quoted the widest tolerance of the three. Guess which one I told him to buy.
A trial is the only stage before signing where a claim can die. Any supplier can print a spec sheet. Material running across a head does not lie. But a report is selective, and what gets left out matters more than what gets written down.
Pin the conditions down before anyone runs
Most people open with, can you trial this. Sure, but trial what. Send the same conditions to every equipment supplier you are comparing, otherwise the reports that come back cannot be set side by side. Pin these down before you ask anyone to run anything.
Whose material. Which batch of substrate, what width and thickness, any surface treatment. Target wet and dry thickness, and the tolerance you can live with. Line speed, the one that matters most and the one people dodge most. Drying temperatures, how many oven zones, what exhaust volume. And what gets tested, thickness, adhesion, appearance, or all of it.
Leave any of these open and the reports that come back cannot be compared. Same pail, run at twenty metres a minute and at sixty, those are different drying regimes and different films. Any coating machine manufacturer will ask for these first, because without them he cannot quote honestly. The one who agrees with everything you say is the one to watch.
What a report has to contain
My own list, habit rather than standard.
Actual line speed, so you can check the drying yourself. Measured wet and dry thickness, both of them, because dry alone cannot be walked back to a process window. Solids content, since it decides whether the wet to dry conversion was honest. Oven zone temperatures and air flows, at least as a range. How adhesion was tested, cross hatch or tape, and what grade. Rewind condition, edges even, no blocking, no wrinkles. How many metres were run, and whether anything broke or spliced on the way. Whether thickness was measured offline or online, how many points, and the spread, not just the mean.
That last one deserves a paragraph. The mean is the easiest number in the world to make look good. Run a kilometre, test the best fifty metres, and the average comes back beautiful. What you need is the spread over the whole run, ideally the curve off an online gauge. An experienced machine builder will not refuse to hand that curve over. The one who does is usually hiding a curve that looks bad.
How the numbers get massaged
A few I have watched happen. Not an exposé, just so you know what you are reading.
Running slow. Trial at twenty metres a minute and the data looks lovely, but your line has to run at sixty. Drying time differs by a factor of three, and the film is not the same animal.
Short runs. Two or three hundred metres, a sample off each end, test the best piece in the middle. Nothing about stability shows up in that. Some OEMs run trials on a single zone machine, so the length they can run is limited anyway, and that is worth asking about up front.
House material. Not your pail, the supplier's own mix, better levelling, fewer bubbles. If the report does not say where the material came from, you have no idea.
Results without conditions. Dry thickness eighteen microns, plus or minus half a micron, reads beautifully. Speed, temperatures, solids, not one line. That report was not written for you, it was written for your boss. Every manufacturer will tell you his numbers are fine. The question is whether he wrote the conditions down.
And one that is not quite cheating. The gauge samples once a metre, so periodic barring never shows up in the data. Nothing fake, just resolution too coarse to catch it. Ask the coating line supplier to state the sampling rate too, and you will know how far to trust the numbers.
How we run trials, and you can ask anyone for this
Continuous, never under a kilometre. At your target production speed, or twenty percent above it if you want headroom. Thickness measured online, the whole run stored, curve attached to the report. Three metre samples off the head, the middle, and the tail, labelled and posted back so you can test them yourself. Drying curve printed for every zone, and the coating line manufacturer hands over the raw data, not just the summary. The customer is welcome on site or on video. We do not run trials with only one party in the room.
A draft lands the same day, the typed report within two days. Plenty of coating line suppliers offer that, but few record the whole run end to end. One OEM I know sends the customer the entire video, waste rolls and all, and I respect that. While you are making demands, ask the machine builder to post the retained samples back to you as well, not just a sheet of paper.
Try asking once and watch. The ones who say yes and the ones who start hedging separate themselves in about ten seconds.
Between a passing trial and a running line
This is where projects flip.
A trial machine usually has one or two oven zones. Your production line has six or eight. Change the length and the drying curve moves with it, so material that dried fine on the trial can fail on the line, and the other way round. Trial speed does not scale linearly to production speed. Drying time, shear rate, and levelling time all move at once.
A trial passing does not mean production runs. A responsible coating line manufacturer writes that into the report and suggests a pilot run before you commit to a full line. A few coating machine manufacturers go further and book the pilot slot for you. I have met very few who do that, and when you meet one, hold on to him. The one who writes that sentence is usually more trustworthy than the one handing you a tidy tolerance table.
So is a trial still worth it. Yes. What it proves is not that production will work. It proves two things, that this material works with this method, and that this equipment supplier is willing to tell you the truth.
Three questions that sort them out
Back to the customer with the three reports. What I gave him was three questions.
Was line speed written down. If yes, he is letting you recheck his maths. Did he give a spread, or only a mean, because a mean alone is half the data hidden. And will he let you stand there, or at least record the whole thing.
Two passed. The third never did produce a speed number. He bought from the one who had written it down. Production tolerances came out a shade wider than the trial, and nothing fell over.
A trial is not a contest over whose numbers look best. It is a contest over who will spread them out in front of you. Pretty reports anyone can write. Data laid bare only comes from someone who actually ran it. I have sat on the coating machine manufacturer side of this table for years, and the customers who scare me are not the picky ones. They are the ones who only read the tolerance row.
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
