Knitted Bag Edge Finishing: Five Methods, What Each Costs Per Piece, and Where Each One Fails

What an Edge Has to Survive on a Knitted Bag

An edge on a knitted bag has to survive three things — unroping, abrasion and handling — and the reason knit edges are a specification problem rather than a detail is that a loop structure has no cut edge to hold: release the last loop and the courses above it come apart.

That is the whole difference from a woven bag. A woven cloth cut on the bias frays a little and then stops, because the yarns are held by the weave at every crossing. A knitted structure is a series of interlocked loops, and a loop released at the edge releases the one above it, and the one above that, which is why the failure is called unroping and why it runs rather than stops. Every finishing method on this page is, in one way or another, a way of preventing that first loop from releasing. Some close it mechanically with a stitch, some wrap it in tape, and some fuse the polymer so that there is no free loop left at all.

Abrasion is the second requirement and it is not evenly distributed. The mouth edge of a tote is handled every time the bag is used, the base edge sits on the ground, and the handle aperture is rubbed by the hand. A pocket edge is barely touched. Specifying one edge treatment for the whole bag therefore over-engineers the pocket and under-engineers the mouth, which is both a cost and a durability error in the same decision.

The Cost of Each Edge, Per Piece and Per Programme

Edge finishing runs from no added cost for a knitted closed selvedge to USD 0.62 per piece for a laser-cut and sealed edge, and on a 1,000-piece programme that spread is the difference between nothing and USD 620 — which is why the edge belongs in the tech pack as a priced line, not as a finish note.

The per-piece figures below cover the operation and the material where there is one. They exclude the knitting cycle, which is unchanged by the choice, with one exception worth naming: a closed selvedge has to be programmed, and doing so adds a small amount of machine time, typically under 30 seconds per piece, because the bed performs the closing action rather than a separate station. Everything else is a downstream operation. Against the reference unit cost of USD 4.80 for a standard E12 tote at specification, a bound edge represents about 8% of the unit cost, which is a substantial fraction for something most briefs do not mention.

Method One: The Knitted Closed Selvedge, at Zero Added Cost

A closed selvedge is produced on the bed by transferring loops at the edge so the structure closes on itself, and it costs nothing in material or downstream labour — the only expense is under 30 seconds of additional machine time per piece.

It is the correct answer more often than buyers expect. Because the edge is closed during knitting rather than afterwards, there is no operation, no component and no thread to match, and the edge stretches with the body instead of restricting it. The limitations are real though: a closed selvedge is 1.5 to 2 mm wide, which is narrow, and it resists unroping but offers little against abrasion. On a bag whose mouth edge will be grabbed several times a day, a closed selvedge alone will eventually show wear. On a bag whose mouth is rarely touched, or where a lining covers the edge, it is the obvious choice and the saving is real.

• Costs no material and no downstream operation, only bed time.
• Stretches with the body, so it does not restrict the mouth or the aperture.
• Edge width is 1.5 to 2 mm, which is narrow against abrasion.
• Cannot be applied to a cut edge: it has to be programmed before knitting starts.
• Best paired with a lining, where the lining protects the edge from handling.

Method Two: Overlocked Edge at 8 to 10 Stitches Per Inch

An overlocked edge wraps the raw edge with a 504 overlock or a 516 safety stitch at 8 to 10 stitches per inch, producing a 3 to 4 mm finished edge for about USD 0.22 per piece, and it is the default for pocket and lining edges on a knitted bag.

The method is economical, fast and familiar to any sewing room, and it works on any fibre, which is its main advantage over thermal methods. The specification detail that matters is the thread: it should match the base polymer, so polyester thread on a polyester body, because otherwise the seam and the fabric age at different rates and the thread fails first. Stitch density matters for the same reason it does on a join. Below 8 stitches per inch the wrap leaves gaps where a loop can release; above 10 it perforates the structure and the edge becomes a line of weakness rather than a line of protection.

Method Three: Bound Edge With a 12 to 20 mm Tape

A bound edge wraps the raw edge in a tape 12 to 20 mm wide, folded and stitched in two passes, and at about USD 0.38 per piece it is the most expensive of the common methods and the most durable on a high-abrasion edge.

Binding is the answer for the mouth of a daily-use tote, and for any edge that will be grabbed, slung over a shoulder or dragged across a surface. It also gives the cleanest appearance and the widest scope for a contrast colour, which is why it survives on retail bags despite the cost. The failure mode is differential shrinkage: a tape whose shrinkage differs from the body will pucker after washing, and the pucker is permanent. Specify the tape in the same polymer family as the body, require a shrinkage figure, and wash-test the combination before bulk rather than afterwards. Two passes, not one, because a single-pass binding leaves the inner fold loose and it rolls after a few cycles.

Method Four: Thermal Cut and Seal at 200 °C

A heated blade cuts and fuses the edge in one pass at about 200 °C, leaving a welded bead 2 to 3 mm wide for about USD 0.18 per piece — the least expensive cut-edge method, and the one that requires a thermoplastic body to work at all.

The economics are attractive and the constraint is absolute. Polyester and nylon fuse; cotton, viscose and most blends do not, and a body with a natural-fibre content above roughly a third will char rather than weld. Where it does work, the result is an edge with no free loops at all, which is the most complete answer to unroping available. The bead is hard, and that is both the advantage and the drawback: it will not unrope, but it also will not stretch, so a sealed edge on a stretch aperture will crack. Thermal cutting also produces a bead of slightly different colour on a piece-dyed body, which is worth approving on a sample rather than discovering on a shipment.

Method Five: Ultrasonic Cut and Seal at 20 kHz

An ultrasonic head vibrates at about 20 kHz and melts the polymer locally as it cuts, sealing the edge without a hot blade, and at roughly USD 0.31 per piece it is the right choice for curved edges and apertures where a straight blade cannot follow the shape.

The advantage over thermal cutting is control. A heated blade follows a straight path well and a curved path badly, because contact time varies as the direction changes, giving an inconsistent bead. An ultrasonic head holds a consistent energy input along a curve, which is why it is the standard method for sealing a handle aperture or a shaped flap. The constraint is the same as thermal cutting: the body needs a thermoplastic content of roughly 65% or more to weld reliably. Energy input has to be set per GSM, and it is worth recording the setting on the run sheet, because a setting carried from a 240 g/m² body to a 520 g/m² one produces an incomplete weld that looks correct until it is pulled.

Closed selvedge

no added material, under 30 seconds of bed time, edge 1.5 to 2 mm

Overlock

504 or 516 stitch, 8 to 10 per inch, edge 3 to 4 mm, about USD 0.22

Binding

tape 12 to 20 mm, two passes, about USD 0.38

Thermal cut

blade near 200 °C, weld 2 to 3 mm, about USD 0.18

Ultrasonic seal

about 20 kHz, weld 2 to 3 mm, about USD 0.31

Laser cut

about 3 mm sealed edge, about USD 0.62, suited to low volume

Choosing by Application: Mouth, Aperture, Base and Pocket

Each edge on a bag has a different job: the mouth is handled, the aperture is rubbed and stretched, the base is abraded, and a pocket is barely touched — so a well-specified bag uses two or three methods rather than one.

The point of splitting the specification is that it costs less and lasts longer. A bag with a bound mouth, an ultrasonically sealed aperture, an overlocked base and overlocked pocket edges spends its money where the wear is, and the total is well below a bag bound on every edge. That is the same argument as reinforcement zoning applied to edges: put the expense where the force and the friction are.

Testing an Edge: Unroping, Abrasion and Peel

Three tests settle whether an edge specification works: an unroping test that pulls the edge and counts courses released, abrasion on the edge under ISO 12947-2, and a peel or grab test on a bound or sealed edge reported in newtons.

The unroping test is the one that catches a bad specification before bulk, and it needs no apparatus: hold the fabric either side of the finished edge, pull apart with steady force, and see whether any courses run. A closed selvedge and a sealed edge should release nothing. An overlocked edge may release one or two before the wrap takes hold, which is acceptable. Anything more means the stitch density is too low or the thread is wrong. Abrasion on the edge is worth specifying separately from abrasion on the panel, because the two are not correlated: a body that returns 30,000 Martindale rubs on the panel can still have an edge that wears through in a month.

Where Edge Finishing Sits in the Programme Windows

Edge finishing sits inside the joining and trim station, which is the largest post-knitting block at about 4.0 minutes of touch time per piece, and it is decided during sampling before the structure is approved rather than after.

That timing matters because three of the five methods cannot be retrofitted. A closed selvedge has to be programmed; a thermal or ultrasonic seal requires a thermoplastic body of sufficient content; and a bound edge needs a tape that has been shrinkage-tested against that body. All three are decisions made at specification, and all three are expensive to change after yarn has been booked. Programme windows sit behind them: MOQ 100 pcs per style and per colourway, 7 working days of sampling, 30 days of bulk, 7 production lines at about 200,000 pcs per month, AQL 2.5 release under ISO 2859-1 and FOB Xiamen as the standard term. For the load-path decisions that sit next to the aperture, see reinforcement zones in a seamless knit body.

Frequently Asked Questions

People Also Ask

The Edge Is a Specification, Not a Finish Note

Most briefs describe the body in detail and the edge in a word. That is backwards in commercial terms: the edge is where knit bags fail, it is priced per piece, and three of the five methods cannot be changed after the programme starts. A tech pack that names the method per edge, with the stitch density, tape width or seal width, costs nothing extra to write and removes the most common source of a disputed shipment.

For the structural decisions around the aperture, see reinforcement zones in a seamless knit body. For the inspection standards used to grade edge faults, see knit bag QC and testing.