How to Check Fabric Weight (GSM) During a Bag Factory Inspection

GSM — grams per square metre — is one of the few material facts that is simultaneously a commercial figure and an engineering one. Change it and you change the price of the material, the durability of the article, the appearance of the finished bag, and sometimes whether the customer's own specification is met. It is also one of the easiest things to check accurately on a table, and one of the things most often not checked at all.

The reason it gets skipped is timing. By the time finished units exist, cutting and coating are long finished, and nobody wants to destroy a unit to cut a coupon. So the check has to be scheduled — at goods-in, before cutting — rather than bolted onto the end of a pre-shipment visit. Saying this early saves the argument later.

How Do Buyers Catch Weight Substitution?

By identifying the roll and keeping a retained sample from it. The economic incentive to ship lighter material than specified is straightforward, and the commercial exposure from it often looks very small when set against a whole programme — which is exactly why systems rather than intentions are what stop it.

Require the roll identity to be recorded at goods-in, require the reading to be taken against that identity, and retain a sealed swatch from the same roll with the same label. Where a programme runs for years, those swatches become the evidence that settles a question nobody could otherwise answer from paperwork alone.

A counter sample works alongside this. Keep one unit from each shipment, identified by lot, in a place where somebody can actually find it in three years' time. Every weight argument that has ever dragged on for a fortnight could have been settled in an afternoon with a bag off a shelf and a pair of scissors.

What Happens When GSM and Hand Disagree?

Trust your hands, then find out why. When a material meets the stated weight but feels noticeably different from the approved reference, something has changed: the construction, the yarn, the finish, the calendering, or the amount of dressing on the surface. Weight alone will not identify which.

The useful next step is to compare what is straightforward: count the yarns in a stated length under magnification, check thickness with a calliper, and look at the edge of the weave under light. Where those all agree with the reference and the hand still differs, a finishing change is the usual answer, and it belongs in the specification rather than being accepted silently.

What Does GSM Actually Measure?

Mass per unit area: the weight of a known area of material, expressed in grams per square metre. Nothing about strength, nothing about weave, nothing about composition. A heavy fabric is not necessarily a strong one, and two fabrics of identical weight can behave completely differently.

That limitation matters because GSM is often used as a proxy for quality. It is a fine specification control and a poor quality guarantee. Where strength is what you actually care about, ask for a tensile test using an accepted method rather than assuming weight tells you what you need.

The measurement itself is straightforward. A round cutter of 100 square centimetres produces a coupon whose weight in grams can be read directly as a fraction of a square metre — so the arithmetic is simply the reading multiplied by one hundred. Put simply, each 0.01 gram on the scale represents one gram per square metre, which is why the second decimal place is not fastidiousness.

Why Does GSM Matter on a Bag?

Three consequences follow directly. Cost — material is quoted and consumed by weight or area, and a lighter fabric than specified is a silent price change. Durability — abrasion resistance, tear behaviour and how long a bag keeps its shape correlate strongly with how much material is in it. And fit — the same bag made in a lighter material handles differently, drapes differently and often reads as a different product altogether.

There is a fourth, less obvious one. Where the destination market or the customer's own filing requires a stated construction, GSM is often the number filed. A departure from it becomes a documentation problem in addition to whatever else it is.

And a fifth worth stating plainly, because it explains why the check earns its place: nobody argues about a number that exists. Where weight is measured and recorded, discussions stay short and factual. Where it is assumed, every disagreement about performance eventually becomes a disagreement about what was actually supplied, and neither side has anything to point at.

Where Should the Sample Be Taken From?

From the roll, at goods-in, before cutting. Take coupons across the width rather than from the edge only, because many constructions — coated fabrics especially — are not uniform across their width. Take them from more than one roll where the lot uses more than one, and record which reading came from which roll.

Reading a finished unit is possible, technically, but the result carries every subsequent process with it: coating, lamination, print, adhesive, soil. A finished unit reading above specification may indicate any of those, and a reading below it is nearly impossible to interpret. Use the roll, and reserve unit readings for investigating a specific question.

What Tools Are Used, and How Accurate Must They Be?

A round cutter of known area with sharp blades, a balance reading to at least 0.01 gram, and a flat surface. That is the whole kit. Accuracy comes less from the equipment than from the technique: blades must be sharp so the coupon is not compressed at its edge, the balance must be levelled and zeroed each session, and the coupon must be free of loose threads and debris.

Check the cutter itself. A blunt blade compresses the material at the edge rather than cutting through it, which adds a little material and quietly raises every reading, and a cutter worn from use may no longer be quite the area it claims. Weighing a reference material of known value at the start of each session takes under a minute and tells you whether the session is worth anything.

Conditioning matters as much with weight as with dimensions. Textile materials take up and release moisture, and a fabric read straight out of a damp store reads heavier than the same fabric after an hour in the room. Standard practice conditions the sample for a stated period before weighing — say so in the clause, or compare readings taken under different conditions and get an argument instead of an answer.

How Many Samples Should Be Weighed?

Enough to know whether the variation is within the roll or between rolls. Three coupons across the width of one roll establishes the profile; doing that on each of several rolls establishes whether the lot is consistent. For a lot drawn from many rolls, treating three coupons as representative of everything is a common and expensive shortcut.

Run the count explicitly. Where the number matters to price or to a filed specification, record every reading rather than the average alone. A mean inside tolerance with readings scattered either side is a different product from one clustering tightly at nominal, and only the full set tells you which you have.

What Tolerance Belongs on GSM?

Around ±5% is common for woven constructions, looser for nonwovens, and best stated rather than assumed. Whatever number you use, state whether it applies to the mean, to each individual reading, or to both — because those are three different requirements and an inspector will otherwise pick one.

Directional requirements deserve particular care. Most buyers care about the lower limit, because a light fabric means less material for the price agreed. Where that is true, write "not less than" rather than a band, so there is no question about the intent.

How Does Coating or Lamination Change the Reading?

It adds whatever has been applied. A fabric specified at one weight can end up substantially heavier once coated, and a single total hides which of the two moved. Take a reading before finishing and another after, and specify both figures — the substrate and the total — rather than only the result.

There is also a timing element. Some coatings and adhesives cure for a while after application, during which weight changes measurably. Read after full cure, or compare like with like and note which point in the process each reading came from.

Is GSM the Right Measure for Every Material?

No, and insisting on it is a frequent error. Foam is governed by thickness and density; a weight figure tells you little about whether padding will protect anything. Film is governed by thickness in microns. Webbing is governed by width, thickness and strength. Insulation is governed by loft as much as weight. Use the measure that governs behaviour, and state it clearly.

Where several measures apply together, write them together. A padding specified only by thickness can be met with a very low density material that collapses in use; specifying density alongside thickness removes the loophole without adding appreciable expense.

How Do Yarn Count and Weave Interact With Weight?

Two fabrics can weigh the same and behave differently because one uses fewer, thicker yarns and the other more, finer ones — which changes abrasion behaviour, tear behaviour, opacity and hand. Where those properties matter, specify yarn count or construction alongside weight so the two together pin down what you actually want.

This is the reason a weight check alone is weak evidence in a specification argument. A fabric meeting the stated GSM but made to a different construction meets the letter of the clause and fails its intent. Naming both closes the gap.

How Should a GSM Finding Be Recorded and Classed?

Record each reading, the roll it came from, its position across the width, the stated specification, the tolerance, the tool, and whether the material was conditioned. That is a short list and every item on it has produced an argument somewhere.

Classification should follow consequence, as always. A reading outside tolerance where function or appearance is genuinely affected is a major. A reading just outside that nobody would notice and nothing depends on is commonly treated as minor — but only if the class was agreed before the visit. Where a filed specification or a customer requirement is breached, the class rises regardless of how the unit looks.

Where weight findings and strength findings disagree, believe the strength test. Weight tells you what was bought; a tensile result tells you what will hold. Both belong in the file, and neither substitutes for the other — which is why the two checks should be written into the same schedule and read together.

What Should Happen When a Weight Reading Fails?

Stop the cutting first, then argue. Material that has been cut into panels cannot be put back in the roll, and whatever the outcome of the discussion, the exposure grows with every metre converted. A short quarantine until readings are confirmed is far less disruptive than discovering the problem three hundred units later.

Ask for a roll map: which rolls were delivered, what reading came from each, and where they were used. Where the failure is isolated to one roll, the exposure may be identifiable and containable; where every roll is light, the lot is the lot and there is no partial answer. Insist on that distinction before anybody starts proposing one.

Then decide using the clause, not the relationship. If the tolerance bound the mean and the mean fails, the material fails; if it leaves room for readings to scatter and only some did, the decision is about risk rather than compliance, and whoever owns the risk should make it in writing with the date consequence stated beside it.

Frequently Asked Questions

People Also Ask