The sample sat flat on the table. You ran a hand across it and moved on. Panels from this container do not do that. Lay one down and the middle stands off the table, or the ends curl up when the center is pressed. That is a different panel from the one you approved, and it is visible before a single screw goes into a wall.

Stop packing the rest. Take one panel out of this container and photograph two things. First, a straight edge laid across the face at mid-length, so the gap under the edge is in the frame. Second, the same panel stood on edge, shot down the long side, so the curve shows as a line that is not straight. Those two pictures are the hold. A moisture theory, a loading diagram, and a gap number typed into chat do not release the rest.

The product page does not show this curve

Indoor decorative wall materials are the range you ordered: fluted and slat panels, WPC and PVC, in the finishes the catalog shows. The page names the product and the look. It does not lay a panel from this container on a table, and it does not show a straight edge bridging a hollow in the middle. Use the category page to confirm what you bought. Use the two photos to decide whether the remaining bundles close.

The commercial WPC guide tells a buyer to judge makers on thickness, moisture performance, and to start with physical samples before the order. That is the right advice at buying time. It does not photograph a bowed panel from a shipment that already left the sample stage. If the factory answers the bow by quoting the same guide back at you, they are repeating the promise instead of showing the panel. The sample already kept that promise. This container has to match it.

Bow is a curve along the face

Wood and wood-composite sheets move, and the trade has names for the ways they move. A general reference on wood warping describes bow as a warp along the length of the face of the wood. That definition is useful for one sentence in your note: the defect you are naming has a name, and it runs along the face, not through the thickness. The same page has no packing step, no photo list, and no number for a wall panel. Do not let a dictionary entry become the answer.

Keep bow apart from the other movements. A panel can twist, so one corner sits high while the opposite corner sits low. It can cup, so the edges curl across the width. It can bow, so the middle of the face rises along the length. Each one shows differently in a photo. If you write only “warped,” the factory will inspect whatever is easiest to fix, pack the rest, and the defect you saw will still be in the next crate. Look at the panel again before you write, and name the movement: the middle is high along the length, or the corners do not touch, or the edges curl across the width.

Also keep the panel apart from the surface. A scratch, a dull patch of grain, or a color step between bundles is a finish question with its own photos. It does not make a bowed panel acceptable, and a perfect surface does not flatten the middle. If the factory sends close-ups of the grain and no straight-edge shot, they have answered a different complaint.

Photograph the straight edge on a panel from this lot

Open a bundle from the shipment that bows. If you have already seen the bow on panels from one bundle, do not let them bring a panel from a different bundle for the photo. Lay the panel face up on a table you trust, or on the floor of the warehouse if the table is too short. Lay a straight edge across the face at the middle of the length, and a second time along the length if the first photo does not show the gap well enough. Shoot square to the panel, from a low enough angle that a reader can see daylight between the edge and the face.

Write the bundle number on the panel with a marker, or hold a card with the number in the frame. If the order has more than one color or code, write the code too. A gap photo with no panel identity can be answered with any panel from anywhere. The card is what ties the picture to this container.

Then repeat the same shot on the approval sample if you still have it, same edge, same distance, same light. Two photos side by side say more than one photo and a paragraph. If the sample gap is nothing and the container gap is visible, the reader does not need to be told which lot they are looking at.

Shoot the same panel down the edge

The second photo is the same panel stood on edge, camera close to the end of the panel, looking down the long side. A bow that is hard to judge across a wide face becomes obvious as a curve in a line. The eye is good at telling a straight line from a bent one. If the long edge runs true, shoot the short edge the same way, because bow can sit in either direction and the face shot alone may hide one of them.

The picture above is only the kind of product this check is about. It is not your bundle, and it does not show a straight edge anywhere. Your two photos have to show your panel, your edge, and your card. A catalog picture of the profile, a render of an installed facade, or a showroom photo does not carry the identity of this shipment.

If the panel is too long for the warehouse floor shot, shoot the edge in two halves and say so in the note. Do not accept a photo of only one end of the panel. A bow is strongest at the middle, which is exactly the part a single-end photo leaves out.

A moisture story does not straighten this lot

The first answer you will get is moisture. The factory will say the panels took on humidity in the container, or that the warehouse air changed after the sample was cut, and that the same thing happens to every lot. Moisture is a real force on wood-composite panels. It is also a story about causes, and you were not sent a cause. You were sent panels that do not sit flat. If they believe the bow grew in transit, they still have to show a panel from this lot, because that is the panel you have. A theory about the voyage does not lay the middle back down.

Do not accept a storage demo as the release either. Photos of panels stacked flat, off the ground, in a dry room, show how the next lot could be kept. They do not describe the bundles already strapped. If they want to prove their storage is right, fine, but the release for this shipment is still the straight edge on a panel from this container.

The same logic covers an acclimation instruction. Letting panels rest in the room before install is site work that happens after you accept the goods. You can acclimate a bowed panel and it will still be bowed. If the reply says “it will settle after installation,” the answer is the same two photos, taken now, on a table, before one more bundle is closed.

A loading plan is not the panel

The second answer you will get is packing. They will send a diagram of how panels sit in the container, vertical or flat, on bearers, with bracing. A drawing like that can be a real improvement, and it is worth asking for when the conversation reaches fixes. It is still a plan for cartons that have not been packed. The panels in front of you were packed some way already, and they bowed. Whether the fix is a new orientation, more bearers, or shorter bundles, the proof is the same: after they pack one test bundle the new way, a panel out of that bundle goes on the table with the straight edge, and the two photos are taken again.

Until that photo exists, a diagram is a promise with lines on it. Do not let the conversation end at the drawing. Do not let it end at a test stack either, if the test stack is never opened. One bundle loaded the new way, then opened and photographed, is the shortest path back to packing with evidence.

Keep the two subjects separate in your note: the panels that are already made, and the way the next ones will be packed. The first is decided by the straight-edge photos. The second is decided by the diagram and the test bundle. Mixing them lets the factory discuss packing for a week while the cartons that already bow are waiting for a decision.

A typed gap number is not the photo

Someone will offer a measurement. They will type a gap figure into chat, or send a feeler gauge reading with no panel in the frame. A number can be read on the sample, on a good day, on a table they chose. You cannot see any of that in digits. You can see it in the photo, because the photo carries the edge, the panel, and the card together.

Do not turn the hold into a debate about tolerance either. The moment you accept the frame “how many millimeters is acceptable,” the conversation becomes negotiation, and the panels sit packed while it runs. You approved a sample that sat flat. The standard for this shipment is that sample, in the same setup. If the factory wants an agreed number for future orders, that is a useful separate talk with the sample on the table, not a substitute for the photos on this lot.

If they insist on measuring, let them, but the ruler or gauge has to be in the frame with the gap it measures, on the panel with the bundle card. A close-up of a gauge on a bench measures nothing you can use.

Write the hold the same day

Write to the person who can stop the line. Name the order. Say the sample sat flat on the table and panels from this container bow in the middle. State the hold in one sentence: no more bundles packed until you have the two photos from a panel taken out of this shipment. Name them. First, the straight edge across the face at mid-length, with the gap visible and the bundle number on a card in the frame. Second, the same panel on edge, shot down the long side, so the curve is a visible line. A release is your note that the panel sits and reads straight like the sample. A moisture story, a loading diagram, a storage demo, or a typed gap figure does not release the hold.

Mark the bundles already closed. Do not mix them with bundles packed after the photos agree. A crew that installs one flat panel and one bowed panel will blame the wall, the adhesive, and the installer before anyone opens a spare bundle. Opening one bundle now shows the curve. Packing the rest, and explaining the bow after the wall is covered, hides it until the room is finished.

Keep the sample photos and these two in the order folder. The next lot will drift. You will see the drift only if the flat sample and the bowed panel are in the same place when someone asks whether to keep packing. A verbal “this lot is the same as last time” is how the curve ships twice.