Bottom sealing looks like a decision made once when the machine is purchased and never revisited. On an automatic case erector it is in fact the largest consumable line item you control, and it decides whether your cases survive an eight-pallet stack, a 30-day sea container, or a chilled distribution centre held at 4 °C.

Most comparisons of hot-melt glue against tape stop at a slogan: glue is stronger, tape is cheaper and easier. That framing is misleading in both directions. Hot-melt is not only the stronger bond — on realistic volumes it is also three to four times cheaper per case in consumables. Tape is not merely the convenient fallback — below a specific annual volume it is the correct engineering choice, because the capital cost of a glue module has nothing to amortise against.

This article gives the two numbers that actually drive the decision: cost per case and measured bond strength. It also covers the board grades, temperature windows and PLC fault codes that quietly move both of them.

The Short Answer

Take a standard FEFCO 0201 box, 400 x 300 x 250 mm, in single-wall B-flute board. Sealing the bottom consumes roughly 0.9 g of EVA hot-melt, or about 0.75 m of 48 mm BOPP tape. At mainstream bulk prices — around USD 3.50 per kg for hot-melt and USD 1.20 per 66 m tape roll — that is USD 0.0032 against USD 0.0137 per case. The consumable gap is close to four-fold, and it compounds with every additional case you ship.

Bond strength runs the opposite way. On virgin kraft board a correctly set hot-melt seal tears 85 to 98 per cent of the fibres out of the liner in a peel test, which means the adhesive has become stronger than the board itself. Tape on the same board typically reaches 40 to 75 per cent, and that figure collapses further on dusty or heavily recycled grades.

Two practical consequences follow. Above roughly 400,000 cases per year, a glue module fitted to a machine such as our Single-Piece Carton Erector pays for itself in consumables within 12 to 18 months. Below that volume the question is not which method is better in absolute terms, but which is correct for this line — and tape often wins, because it removes the nozzle cleaning, tank top-up and glue-degradation routines that dominate maintenance hours on a hot-melt machine.

How Bond Strength Is Actually Measured

Supplier claims about “strong seals” are worth very little. Ask for two tests instead, and insist both are run on the board you actually buy rather than on a demonstration sample.

The fibre tear test

The standard method is ASTM D2860 / TAPPI T-812. A technician peels the sealed flap back at 180 degrees and reads the proportion of board fibres that came away stuck to the adhesive rather than separating cleanly at the bond line. That single reading tells you whether the adhesive is doing its job or merely sitting on the surface.

A fibre tear above 90 per cent means the adhesive is stronger than the board — the correct outcome, and nothing needs changing. Between 70 and 90 per cent is acceptable for most corrugated work, though it is worth re-testing on double-wall grades where the inner liner behaves differently. A result between 40 and 70 per cent points to a surface problem: board dust, insufficient compression time, or glue applied below its working temperature. Anything below 40 per cent will open in transit, and the right response is to stop the line and recalibrate rather than to increase glue volume.

Conditions matter as much as method. Run the test first at 23 °C and 50 per cent relative humidity, then repeat it at the lowest temperature your cases will genuinely see in the supply chain.

What temperature does to the seal

A standard EVA hot-melt is a thermoplastic: it softens as it warms and turns brittle as it cools. Held at −20 °C for 24 hours, a standard grade falls below 30 per cent fibre tear and the seal becomes a fracture point in the case. A cold-grade pressure-sensitive formulation designed for frozen distribution holds 75 to 90 per cent under identical conditions. Acrylic BOPP tape also survives the freezer, sitting between 60 and 80 per cent, which is why frozen food lines frequently stay on tape even at high volume.

In chilled distribution at 4 °C the picture closes. Hot-melt recovers to 80 to 90 per cent and cold-grade adhesive to above 90, while tape reaches 70 to 85 per cent and starts to be limited by surface dust rather than by adhesive chemistry.

At ambient temperature the ranking swings back to glue. Above 40 °C sustained for 30 days — a closed container on a tropical route — standard hot-melt shows slight creep but holds 85 to 90 per cent fibre tear, and tape lands in the same band with a higher risk of edge lifting if the board is warped.

The variable that decides most real projects is board composition. Moving from virgin kraft to board with 60 per cent recycled content costs hot-melt 15 to 25 percentage points of fibre tear, but it costs tape 25 to 40 points. Tape adhesion is a surface property, so dust and clay coating defeat it directly; glue penetrates the liner and binds fibres. If your purchasing team optimises board price aggressively, that single difference will eventually decide your sealing method for you.

Board Grade Changes the Answer

Sealing performance is a function of flute profile, wall count and recycled content. The values below are the working envelope we size nozzle pressure, glue grammage and compression dwell against when configuring a machine.

Three field notes matter more than the numbers themselves. Below a 110 g/m² liner, hot-melt strikes through thin E-flute and leaves a visible bead on the printed face; the fix is to reduce grammage and lengthen compression dwell, not to raise nozzle temperature. Board containing 70 per cent or more recycled content raises airborne dust, and blowing the flap path off at 0.5 to 1.0 bar is far cheaper than specifying a more aggressive adhesive. Double-wall boxes need longer compression, not more glue: excess adhesive squeezed onto the machine frame during the transfer stroke is the defect we see most often on BC and AB board.

What the Seal Really Costs Over a Year

Consumable cost is easy to model once you know the grammage and the length of tape per case. Use this arithmetic directly:

Sealing cost per case = (adhesive grams per case ÷ 1000) × glue price per kg
Tape cost per case    = tape metres per case × tape price per metre

Annual saving = cases per year × (tape cost per case − glue cost per case)
Payback in months = glue module price ÷ (annual saving ÷ 12)

Applied to the 400 x 300 x 250 mm box above, the annual gap scales linearly with volume and looks like this.

A glue module typically adds USD 12,000 to 18,000 to machine cost, which places payback at 14 to 20 months at one million cases a year — and that is before counting the labour released by not changing tape rolls. Those hidden costs are easy to underestimate. A 66 m roll is exhausted every 60 to 70 metres of case seam, and every roll change costs the line 25 to 40 seconds of output. A misthreaded tape head is the single most frequent cause of an unplanned stop on tape-sealed erectors, and it happens most often on the shift with the least experienced operator.

Speed and the Price of Micro-Stoppages

At 35 to 40 cases per minute, a 30-second stop wipes out about 18 cases of output. Individually that is noise. Twenty such events across a shift is 360 cases, or roughly 2 per cent of a three-shift week — which is usually more than the entire difference between a good and an average machine.

The stoppage causes differ by sealing method. On hot-melt machines the dominant fault is nozzle clogging, usually from carbonised glue held above 180 °C for long idle periods; an automatic standby that drops the tank to 120 °C and a weekly nozzle purge remove most of it. On tape machines the dominant fault is tape snap, caused by a dull knife or drifting roll tension; blade inspection every 50,000 cases and a head with automatic tape change cover the same ground. Two failure modes are shared: flaps not held long enough for the seal to set, particularly on double-wall board, and misalignment from a drop in suction cup vacuum, which should be verified at 0.4 to 0.6 bar and cleaned weekly.

PLC Alarms and the Maintenance Rhythm

Modern erectors will tell you what is going wrong, provided the operator reads the alarm rather than clearing it. The four messages below account for the large majority of unplanned glue-sealing stops in the field.

Annoyingly routine maintenance is what keeps those alarms rare. Every shift, check that glue in the tank is clean, that the nozzle is not weeping, and that suction cups are dust-free. Every week, purge the nozzle, verify vacuum at the pump, and inspect tape knife or glue pattern registration. Every month, torque the main drive belts, check the compression section for board dust build-up, and verify squareness against a reference case. Every quarter, replace suction cups on the picking head and clean or replace filters. Annually, re-tension the drive chain, replace the nozzle assembly, and calibrate temperature controllers against a reference probe. Machines that follow that rhythm typically run above 98 per cent uptime across three shifts.

Choosing for Your Line

The decision narrows quickly once volume and board composition are known. Above one million cases a year on one or two box sizes, hot-melt with automatic nozzle standby is almost always correct, and pairing it with the two-lane output of our Double-Head High-Speed Case Former is the standard configuration for that duty. Where a chilled or frozen chain is involved, cold-grade pressure-sensitive adhesive or tape should be specified from the outset rather than retrofitted later.

For operations running more than 200 stock keeping units in small batches, tape or a seal-only head usually produces better uptime than a glue system that spends its day heating and cooling. Where the case carries a high compression specification — display-ready or export pallets stacked eight high — a rigid glued seam protects stack strength better than a hinged taped seam, and the loss of 5 to 15 per cent of compression strength from taping becomes commercially significant. Pre-glued trays and three-piece designs follow a different logic again and are covered by our Three-Piece Carton Erector; specification questions on sizing and board grade are answered in our FAQ.

FAQ

Is hot-melt always cheaper?
On consumables, yes — typically three to four times cheaper per case. On total cost of ownership, only above roughly 400,000 cases per year once capital cost, maintenance hours and energy are included.

Can a glue module be retrofitted to a tape-sealed erector?
Usually yes, if the PLC has a spare output and the machine has 0.6 to 1.0 kW of electrical headroom. Budget three to five days of downtime, including re-validating squareness and compression dwell.

Does tape ever beat glue on strength?
On coated, heavily printed or dusty board, yes — where glue cannot wet the surface. Test both methods on your own board before committing to either.

What destroys hot-melt seals in summer?
Creep at sustained temperatures above 45 °C inside a closed container. Specify a higher softening point grade (above 100 °C ring-and-ball) for tropical routes.

Ready to cut your cost per case? Send your carton specifications — length, width, height, flute grade, target cases per minute and annual volume — to [email protected] or WhatsApp +86 13681839278. Our engineers will return a sealing-cost model and a machine sizing sheet within 24 hours, with no obligation.