Cup Bottom Curl: The Joint That Leaks Before the Seam Does
The bottom of a paper cup is a curl that is folded around the bottom disc and sealed into the wall, and it fails differently from the side seam: it weeps slowly along a circumference rather than beading at one point. Specify the curl radius and the bottom diameter with tolerances, the seal pressure and dwell, and the adhesive weight if adhesive is used, then test with a hot fill held for thirty minutes on a dry surface. A cup that passes a five minute check and marks the table at twenty is showing a curl that is compressed but not bonded, and no visual inspection of the base will find it.
Scenario
A converter supplies 12 ounce double-wall cups for a takeaway chain. The chain runs its own acceptance test: fill with water at 90 degrees, stand on a paper towel, inspect after five minutes. The cups pass. In service, staff report that a small share of cups leave a wet ring on the counter after ten or fifteen minutes, and one store reports a cup that collapsed at the base when a customer picked it up.
The returned cups are cut open. The bottom curl shows a bright, glossy band around about a third of the circumference, which under magnification is a compressed but unbonded zone: the curl was pressed against the wall, the coating flowed, but there was no adhesion. The remaining two thirds are properly bonded. On the failed cups the bottom diameter measures 0.6 millimetres smaller than on the passing cups from an earlier lot, which means the curl was drawn tighter and the sealing pressure was consumed by the geometry rather than applied to the interface.
The five minute test did not catch it because the leak path is thin and vertical, and the water needed twenty minutes to travel it.
What the Bottom Joint Is
A paper cup base is a disc with a curled edge. The curl is formed, coated side out, and the wall is wrapped around it and pressed so that the wall's inner coating and the curl's coating fuse. The joint is therefore a ring, and its quality depends on three things: the geometry of the curl, the temperature and pressure at the press, and the time the interface spends at that temperature and pressure.
Two seal styles exist. A heat seal fuses the two coatings directly and needs the coatings to be compatible and the interface to reach a temperature above the softening point. An adhesive seal uses hot melt applied to the curl and is more tolerant of coating variation but adds a weight to control. A converter that changes from one to the other needs a different set of numbers, and a specification written for the first does not describe the second.
Why the Curl Geometry Matters
The curl is where the pressure goes. A press applies force to the wall and the curl transmits it to the interface; if the curl is tight, part of that force is absorbed in deforming the paper rather than in pressing the two surfaces together. The result is a joint that looks formed and feels firm and is not bonded.
Three geometric values decide this. The curl radius, typically 1.5 to 2.5 millimetres, which sets how the paper flows around the corner. The bottom diameter relative to the wall, which sets the interference. And the curl height, which sets how much of the wall is engaged.
| Parameter | Typical value | Effect if out of range |
|---|---|---|
| Curl radius | 1.5 to 2.5 mm | Too tight cracks the coating at the fold |
| Bottom diameter | per drawing, plus or minus 0.3 mm | Too small starves the seal of pressure |
| Curl height | 3.5 to 5.0 mm | Too short reduces the bonded band |
| Seal temperature | 180 to 220 Celsius | Too low gives a compressed, unbonded ring |
| Seal pressure | 0.4 to 0.6 MPa | Too low leaves channels |
| Dwell | 0.5 to 1.0 second | Too short does not fuse the coating |
| Adhesive weight, if used | 0.05 to 0.10 g per cup | Too low leaves dry patches |
| Board coating | 15 to 25 gsm inner | Too thin cannot bridge the fold |
Why the Failures Look Alike and Are Not
A weeping base can come from four causes, and they present the same way to a customer.
An unbonded ring from low pressure or low temperature. A cracked coating at the curl, where the fold was too tight for the board and the coating fractured along the crease. A wall that is out of round, so the curl is engaged on one side and starved on the other. And a bottom disc that is too small, which leaves the curl loose in the wall and the joint relying on friction.
Each has a different fix, and the diagnostic is the same: cut the cup, look at the ring at magnification, and measure the bottom diameter. A bright band is pressure or temperature. A line of white or cracked coating at the fold is a curl radius. A ring that is bonded on one side and open on the other is roundness. A curl that sits loose with no compression mark is the diameter.
Testing That Reproduces the Field
Five minutes at 90 degrees is a useful quick check and an incomplete one, because the bottom leak path is long and slow. Three additions close the gap.
Extend the hold to thirty minutes on a dry paper towel, and photograph the towel. The mark is the evidence, and it is objective: a supplier and a buyer looking at the same mark argue about the cause rather than about whether there was a leak.
Fill at the temperature the product will actually reach, not at 90 for a product served at 95, and not at 90 for a cold drink that will be tested at 5. The joint's failure mode changes with temperature because the coating softens and the board swells.
Add a load. A cup is pressed down when it is placed in a holder and squeezed when it is lifted, and both apply a bending moment at the base. A 20 newton load applied for ten seconds and then released, followed by another thirty minute hold, catches a joint that survived the static test.
Roundness and the Starved Side
The wall is a wrapped blank, and its roundness depends on the blank's edge quality, the forming mandrel and the seam. If the wall is oval by even 0.3 millimetres, the curl engages fully on the long axis and barely touches on the short one, and the seal is starved on two opposite sides.
Measure the diameter at two perpendicular directions on a sample per shift. A difference over 0.3 millimetres is worth investigating before it becomes a bottom leak, and it is a measurement that costs a caliper and a few seconds.
The same measurement catches a mandrel that has worn and a blank that has been cut at the wrong width. Both are invisible in the finished cup and both produce a base failure that appears as a pressure problem.
Coating Continuity at the Fold
The fold at the curl is the tightest radius in the cup, and it is where the coating is most highly strained. A coating that is too thin or too brittle fractures along the crease, and the fracture is a continuous path from the inside to the outside.
Look at the cut section at the fold with magnification and a light from behind. A continuous line of light through the coating is a fracture, regardless of how good the seal around it looks. The fix is on the board: a heavier coating, a more ductile grade, or a curl radius that respects the board's elongation.
This is the failure that confuses a converter most, because the seal is perfect and the cup still leaks. It is also the one that a five minute test at low temperature will never find.
Machine Settings and the Log
Temperature, pressure and dwell, logged hourly against the bottom diameter and the towel test result. A drift in the press heater shows as a falling bond quality with a stable diameter; a drift in the curl former shows as a changing diameter with stable settings.
Plot the two together, and the two causes separate. A log with only one variable tells you that something changed and not what.
What Triggers a Stop
A bright band over more than five percent of the circumference, any mark on the towel at thirty minutes, a bottom diameter change over 0.3 millimetres, or a roundness difference over 0.3 millimetres. Each of these is a number, and each has a sample frequency on the shift sheet so that the decision is a comparison rather than a judgement.
What Belongs in the Specification
The curl radius, the curl height, the bottom diameter and the roundness limit, all with tolerances; the seal style with temperature, pressure and dwell ranges; the adhesive weight if adhesive is used; the coating weight and grade; the hot fill temperature and hold time; the load test; and a section photograph of a bonded ring at first article.
A cup drawing that shows the rim and the height has described the opening. The base is where a cup fails in a customer's hand, and the curl is the part that decides it.
Bottom Line
A bottom leak is a ring that was compressed and not bonded, or a coating that fractured at the fold, and both are found by cutting the cup and looking rather than by filling it for five minutes. Specify the curl geometry and the seal settings with tolerances, measure roundness as well as diameter, hold the hot fill for thirty minutes with a load, and keep the towel photograph with the lot. A cup that weeps at the base has usually passed every check the factory ran.