Knitted Bag Sample Approval Criteria: What to Check, in What Order, Before Signing Off a Knit Programme

What Is a Sample Approval Actually For?

A sample approval converts a prototype into a reference standard: the signed sample is what the AQL 2.5 inspection under ISO 2859-1 will be measured against, so every criterion that is not recorded at approval cannot be enforced in bulk.

That framing changes what the meeting is about. A team that treats approval as an aesthetic decision approves a bag it likes and then discovers, at inspection, that there is no agreed figure for the things it did not think to measure — pocket symmetry, handle drop, lining twist, shade under a stated light source. A team that treats approval as a specification event records twelve figures and leaves with a document that inspection can use. The second meeting takes an hour longer and removes most of the disputes that otherwise arrive with the shipment.

There is a second function that is easy to overlook. The approval sample is also the reference for the repeat run. A programme that expects to reorder benefits from a signed sample with recorded measurements, because the repeat is then produced against a known standard rather than against a memory of what the first bag looked like. Holding the sample physically, alongside its recorded figures, is the least expensive quality control available on a second programme.

Criterion One to Three: Dimensions, Weight and Gauge

The first three criteria are dimensional: body dimensions within about plus or minus 10 mm, finished GSM within about 5% of the specified figure, and the gauge verified as the one the programme specified rather than the one that knitted fastest.

Measurement discipline is what makes these three meaningful, and it is the part most often skipped. A dimension measured wherever the tape happens to sit is not a measurement; it is an opinion. The pack should name the measurement points, the state of the bag when measured — flat, empty, relaxed — and the relaxation period before measurement, because a knitted body continues to move after it leaves the bed. GSM should be measured on a cut sample of a stated area rather than inferred from the yarn count, and gauge should be read on the fabric rather than assumed from the machine setting.

Capacity is the fourth measurement and the one that connects the dimension sheet back to the specification that produced it. Measuring the volume the sample actually holds, by a stated method such as filling with a calibrated medium or by calculation from the three internal dimensions, confirms that the litres the programme specified are the litres the bag delivers. A discrepancy here is usually the first sign that the dimension sheet and the capacity figure were produced independently, which is a common outcome when a brief gives both and nobody checks that they agree.

The handle or strap drop belongs with these three rather than with the subjective criteria, because it is a measurement that determines whether the bag works. A drop measured at about 220 mm for hand carry or 250 to 300 mm for the shoulder is a figure inspection can check; a strap length on its own is not, since it means nothing without the body height it attaches to. Recording the drop at approval is what allows an inspector to reject a bag that hangs wrongly rather than one that is merely different.

Body dimensions

Within about plus or minus 10 mm at stated measurement points

Finished GSM

Within about 5% of the specified figure

Gauge

Verified on the fabric, not assumed from the machine setting

Measurement state

Flat, empty and after a stated relaxation period

Handle or strap drop

Measured, not the strap length alone

Capacity check

Volume measured against the specified litre figure

Criterion Four to Six: Colour, Surface and Hand

Colour, surface and hand are the three subjective criteria, and all three can be made objective: Delta E 1.0 or better under D65, a stated surface defect list, and a hand comparison against a retained reference swatch.

Colour is the easiest of the three to formalise and the one most often left to judgement. The figure to record is the Delta E against the brand reference, measured on the finished knit under a stated illuminant, with the formula named — CIEDE2000 rather than CIELAB wherever the palette carries saturated colours. Surface is harder and needs a defect list: barre, snagging, dropped stitches, oil marks and shade banding, each with an acceptance threshold. Hand is the hardest and is best handled by retaining a reference swatch from an approved style rather than by describing it in words, because no adjective survives a year and a change of programme team.

• Colour: Delta E 1.0 or better, CIEDE2000, under D65, measured on the finished knit.
• Metamerism: index below 0.5 across D65, TL84 and illuminant A.
• Surface: a named defect list covering barre, snagging, dropped stitches and banding.
• Hand: retain a physical reference swatch rather than describing it in writing.
• Trim match: body and trim approved together, since they are dyed in different processes.

Criterion Seven to Nine: Strength, Abrasion and Fastness

Three performance criteria should be evidenced before approval: about 20,000 Martindale rubs to ISO 12947-2, roughly 600 N of strip tensile strength to ISO 13934-1, and colour fastness of grade 4 or better to AATCC 8 for rubbing.

The timing of these three is the point. Fastness and abrasion results take days, and a programme that waits for the sample to arrive before commissioning them loses those days twice — once waiting and once again if a result comes back marginal. The efficient sequence is to commission the tests against the material specification while the sample is being knitted, so that the results arrive with the sample rather than after it. Where a result is marginal, the decision to re-specify rather than to accept is far easier to make before bulk begins than after 30 days of production.

Criterion Ten: Hardware and Trim Behaviour

Hardware is criterion ten and it is tested on behaviour rather than on appearance: a zip at about 5,000 cycles, plating at 24 hours of neutral salt spray, and nickel release below 0.5 micrograms per square centimetre per week to EN 1811.

Hardware deserves its own criterion because it is the only part of a knitted bag sourced from an entirely separate supply chain, and because its failure modes are different in kind. The body wears; hardware breaks. A zip that has been cycled 5,000 times on the approval sample will behave the same in bulk, and one that has not been cycled at all is an assumption rather than a specification. Salt spray and nickel release are the two compliance-linked figures and should be recorded with the sample, since a hardware declaration requested at customs is far easier to produce from an existing record than from a new test.

Criterion Eleven: Load and Recovery

Load and recovery is criterion eleven and the one most approval meetings omit: load the sample to its stated working load, hold it, unload it, allow it to relax, and confirm the principal dimension has changed by about 2% or less.

This criterion exists because a knitted body behaves differently from a woven one over time, and because the difference only appears after use. A bag that passes every dimensional check flat can still sit lower and hold its opening wider after a month of carrying, and by then there is no remedy available. Running the check on the approval sample takes a few hours and predicts the field behaviour of the whole run. The figure to state alongside it is the working load, which should be derived from the capacity rather than assumed: about 5 kg for an 18-litre tote and about 8 kg for a 30-litre weekender.

Criterion Twelve: Documentation and Compliance

The twelfth criterion is documentation: the fibre declaration, the restriction screen for each material, the hardware declaration and the label content, all confirmed for the destination market before the sample is signed rather than after the bulk ships.

Documentation is the criterion that costs nothing to close at approval and a great deal to close later, because a screen commissioned after production has already begun applies to goods that exist. The list is short and should be requested as a package: a restriction declaration against REACH Annex XVII for the body and each trim item, a Proposition 65 assessment where the destination is California, a nickel release result for metal components, and a confirmation of the label content in the language of the destination market. Each is a fixed fee at any volume, which is why closing them at approval costs the same as closing them later and is worth considerably more.

What Happens After the Sign-Off

After sign-off the sample becomes the inspection reference and bulk begins: 30 days of production, then release against AQL 2.5 under an ISO 2859-1 sampling plan, then freight from FOB Xiamen at about 21 days by sea.

Two actions at sign-off make the rest of the programme run more smoothly, and neither takes long. The first is to write the defect list for inspection, using the twelve criteria as its structure, because an unnamed defect cannot be inspected and the twelve criteria are exactly the list. The second is to retain the sample physically, with its recorded figures attached, and to name where it is held — because the repeat run, the inspection dispute and the next season's revision will all need it, and a sample that cannot be found is a standard that cannot be enforced.

A third action is worth taking even though it costs a little more: approving a second sample and holding it separately. Bulk is inspected against one reference and the repeat run against another, and a single sample that travels between the two will eventually be lost or damaged. A duplicate kept at the destination, alongside the recorded figures, means a dispute can be settled on both sides without shipping anything, and it is the least expensive insurance available on a programme that expects to reorder.

Frequently Asked Questions

People Also Ask

Approve a Standard, Not a Bag

The difference between a programme that ships cleanly and one that argues for a month is decided in a single meeting. Approving a bag means forming an impression and hoping bulk matches it. Approving a standard means recording twelve figures, writing them into the defect list and retaining the sample — after which bulk can be inspected, a dispute can be settled and a repeat can be produced against something that exists rather than something remembered.

For the sequence that follows approval, see custom knitted bag order timeline. For the second run that the approval sample serves, see knitted bag reorder and repeat runs, and for the volume structure behind the programme, see custom knit bags.