Panel Warpage: 5 Causes Before the Reflow Oven
Panel warpage is the loss of board flatness across a PCB array, and it is measured as the maximum distance between the panel and a flat reference. It matters long before the finished product, because a panel that is not flat does not sit correctly in the printer, the placement machine or the reflow oven.
The cause is usually a stack of small effects rather than one mistake. Copper distribution, lamination pressure, storage conditions and the way the panel is handled after fabrication all push the same way, and the order of operations matters as much as any single parameter.

What Panel Warpage Does to Placement
When a panel is not flat, the stencil does not contact the board evenly. One end of the array prints a full deposit and the other end prints a skimming, and the placement machine compensates for height rather than for a curved surface.
The oven adds its own problem. A panel that is flat when cold can bow when it is hot, and a bowing panel can lift a lead out of a molten joint. The defects that follow look like placement faults or profile faults, and the real cause is the flatness of the base material.
Where Board Flatness Is Lost in Fabrication
Lamination is the first place to look. Copper that is not balanced between the two sides of a core builds internal stress, and the panel relieves it by bowing when it cools. Our balanced stackup notes explain how an odd layer count makes that harder to control.
Machining and baking come next. A panel that is dried too quickly, or a core that is baked at the wrong temperature, will move when it is pressed. The panel that leaves the press flat can still move later if the internal stress was built in rather than relieved.

Measuring the Panel on a Surface Plate
The reference method is simple: the panel is placed concave side down on a surface plate and the largest gap is measured with a feeler gauge or a height gauge. The measurement should be taken at room temperature after the panel has settled, not immediately after an oven or a cleaning line.
The result is recorded as a distance and as a direction, because the direction tells you which way the layers are pulling. A panel that bows along the short axis is usually reacting to a different imbalance from one that twists, and the fix depends on distinguishing the two. Our warpage control notes cover the process responses.
Reflow, Heat and the Second Warp
The reflow oven is where flatness is tested hardest. The board softens as it approaches the glass transition temperature, the copper and the laminate expand at different rates, and a heavy component adds a load that the hot board can no longer carry without changing shape.
Board support and pallets are the usual countermeasures. A support pin placed under a large component keeps the surface where it was set, and a pallet with a flat base holds the outline. Both should be designed with the panel rather than chosen at the machine.
Strain Gauge Data at Depaneling
Depaneling puts the highest mechanical strain on a board during the whole build. A router that is pushed too fast, a shear that is not parallel or a snap that is done by hand all bend the array, and the damage that results is invisible until a joint fails in the field.
A strain gauge mounted near the stress concentration during a trial run gives a number that can be compared with a limit. The measurement is the only way to show that a depaneling method is acceptable, and our strain measurement notes describe the setup.
Storage, Humidity and Flatness
Laminate takes up moisture from the air, and a panel stored in a humid room will move as it absorbs. Baking before assembly dries the panel and also relieves stress, which is why a bake is sometimes used to rescue a warped lot rather than to protect against popcorning.
Panels should be stored flat and supported along their length, not leaning in a rack or stacked under weight. A stack of heavy panels left in a warm room will take a set, and the set is visible the moment the stack is broken.
Accept or Reject: Setting the Limit
The acceptance limit should be written as a percentage of the panel diagonal or as an absolute figure for the product, and it should be agreed with the assembly line rather than with fabrication alone. A limit that satisfies the fabricator but not the printer is a limit that produces scrap at the next step.
Where the limit is close to the process capability, the decision should be made on the measured distribution rather than on a single panel. The out of tolerance panel is easy to see; the drift toward the limit is what needs to be managed.
Records for a Warp Claim
A warpage claim needs the measurement, the method, the temperature and the sample size. Without those four items the discussion becomes a comparison of two gauges and two opinions, and the customer is left without an answer.
Keeping the measurement with the panel lot also allows a correction to be traced. If a lamination change is followed by a rise in warp, the record shows it immediately, and our depaneling method notes describe the later steps where the same data is used again.
Pressing Parameters and Their Influence
Press cycle, temperature ramp and the cooling rate all act on the stress that ends up locked in the panel. A fast cool leaves more stress than a slow one, and a press that is opened while the stack is still hot will release a panel that bows as it reaches the table.
The parameters belong in the process recipe for the stackup, not in a general instruction, because a thin four layer panel and a thick twenty layer panel do not respond in the same way to the same cycle. A change of supplier for the prepreg is also a change that can move the result.
How Panel Size Changes the Answer
A long panel accumulates more movement than a short one, so the same imbalance in the laminate produces a larger gap on a bigger array. That is why a limit written as a percentage of the diagonal is more useful than a fixed figure in millimetres.
Panel design also sets how much of the array is unsupported. A rail that carries the panel through the line adds stiffness along two edges, while a fine web between units adds almost nothing, and the middle of a large array is where the warp shows up first. Our panelization notes cover those geometry choices.
FAQ
Can a warped panel be flattened? Sometimes. A controlled bake under a flat weight can relax some of the stress, and the result should be measured again before the lot is released. Where the warp comes from unbalanced copper the effect will return during reflow.
Does a thick board warp less? Thickness helps, but it is not a cure. A thick panel with badly balanced copper can still bow, and it takes more force and more heat to bring it back, which makes the correction harder once it has happened.
How many panels should be measured? Enough to describe the lot, which in practice means at least five spread through the batch, including the first and last. A single reading describes one panel and says nothing about the run. The IPC materials standards define the reference method.



