A padded envelope lands on your desk five working days after you sent the enquiry. Inside is a 300 mm by 200 mm piece of black rubber belt, three plies, a thick top cover, a thin bottom cover, and a handwritten label with the construction on it. The question every buyer asks at that moment is the right one: what does a conveyor belt sample actually prove before placing an order, and what is it quietly failing to tell you?

The honest answer is that a swatch is a strong test of the recipe and a weak test of the kitchen. It shows you what the supplier can build when they are building one small piece with full attention. It says almost nothing about what comes off the press at the far end of a 300-metre roll on a Thursday afternoon. Treating the two as the same thing is how buyers end up with a production belt that measures a millimetre thinner on the top cover than the swatch they approved.

This article splits the swatch into what you can legitimately conclude from it with an hour and a few hand tools, what you cannot conclude from it at all, and what evidence fills the gap.

What the sample lead time tells you before you open the envelope

Start with the number on the envelope. Jiutong Conveyor Belting, a Boye County works listed on ChinaMakersHub, states standard samples in 5 working days and custom samples in 10 — company-stated figures, like everything on its profile. That split is worth reading carefully, because it hints at where the sample came from.

A five-day standard sample is almost certainly cut from something already made: a stock construction, an offcut from a recent production roll, a retained piece. That is not a bad thing. A cut from a real production roll is the most honest sample there is, because it came off the press under production conditions, not sample conditions. A ten-day custom sample, by contrast, has had time to be built: plies laid up to your spec, covers applied, cured, trimmed. It is closer to your construction but further from production reality, because a small piece built by hand on a sample press or in a corner of the main press does not experience the same heat history or pressure as a full-width belt.

So the first question to ask on the day the sample arrives is not "is it good" but "was this cut from a production roll or built as a sample?" — and if built, on which press. The fact basis for Jiutong does not answer that, and you should not assume it. Ask, and write the answer on the label.

What a swatch legitimately proves

With a durometer, a vernier caliper or thickness gauge, a sharp knife, a steel rule, a kitchen scale and daylight, a 300 mm swatch will tell you the following with reasonable confidence.

Ply count and carcass type. Cut a clean cross-section with a fresh blade and count the plies. Look at the fabric weave under a loupe: a plain woven fabric with visible warp and weft is what you expect in an EP (polyester warp, nylon weft) or NN belt; you cannot reliably distinguish polyester from nylon by eye, but you can confirm the count and the fabric-to-rubber ratio. If your spec says four plies and you count three, the conversation is over before it starts.

Cover gauge at the cut edge. Measure top and bottom cover thickness at several points along the cut edge, not just one. A cover that measures 4.0 mm at one end and 3.4 mm at the other on a piece 300 mm long tells you something about calender control. Measure through the full thickness too, and check that ply count times a plausible ply thickness plus the two covers adds up; if it does not, either the skim coats are unusually thick or thin, and either is worth asking about.

Cover hardness. A Shore A durometer pressed on the cover at three or four points gives a fast read on the compound. You are not verifying a specification with a hand durometer — bench testing on a proper specimen does that — but you are checking that a cover claimed as a general-purpose abrasion grade reads in the range you are used to on comparable belts, and that it is consistent across the swatch. A cover that reads noticeably different at two corners has not cured uniformly.

Surface finish and mould condition. Look at the top cover in raking light. Press-cured belt carries the texture of the platen or the release cloth; pitting, pinholes, blisters or a visible ridge running across the cover are signs of platen wear, trapped air or a bad cure. Small, but they are the kind of defect that becomes a hot spot under an impact idler.

Edge construction. Is the edge moulded (rubber wraps the carcass, plies not exposed) or cut (plies visible)? Both are legitimate; they cost differently and behave differently in wet or abrasive duty. Check that the sample edge matches what you asked for and what you are being quoted.

Inter-ply and cover-to-ply adhesion — crudely. Slit into the swatch between a cover and the first ply, grip the tab with pliers and pull. You are not measuring newtons per millimetre; a hand peel is not a test. But a cover that comes away cleanly with the fabric weave printed on its underside and little rubber left on the ply has poor adhesion, and you will see it. A peel that tears rubber rather than separating at the interface is what you want to see.

Weight per square metre. Weigh the swatch, measure it, and divide. Compare to what the supplier's construction ought to weigh. A belt that is unexpectedly light for its stated gauge usually has thinner skims or a lighter fabric than claimed.

That the supplier can read a specification. This is underrated. If you sent width, ply count, top and bottom cover gauge, compound grade and edge type — the fields Jiutong's profile says are specified per order — and the swatch matches all of them, the supplier's order intake works. If two of six are wrong on a sample, expect the same on a roll.

What a swatch cannot prove

Property Swatch verdict Why not What proves it
Full-thickness tensile strength (N/mm) No Needs a standard specimen and a tensile machine, and a piece cut from the roll, not a sample build Third-party test report on a production sample
Belt strength across the full width No Edge-to-centre variation only exists on a full-width belt Width-position specimens from a production roll
Longitudinal consistency No A 300 mm piece is one moment of one press cycle Pre-shipment inspection at several positions along the roll
Camber and straightness No Only measurable on a length of belt laid out flat Roll inspection; a tracked trial length
Troughability No Requires a full-width transverse specimen Test to the troughability method on your specification
Splice quality No A swatch has no splice Trial roll with a factory splice, or a splice sample built to your joint spec
Cover abrasion loss No Bench test on a prepared specimen Laboratory report, with the method and the grade named
Compound batch consistency No One compound batch, one piece Batch records; durometer and gauge readings logged on every roll
Cure state of the full belt Weakly A small build cures more evenly than a full-width belt Hardness and adhesion readings across the width of a production belt
Length and width tolerance No Not represented at all Roll measurement before shipment

Read that column of "no" the way you would read any specification: the swatch is answering a different question than the one you are about to spend money on. It confirms the recipe. It does not confirm the roll.

There is one more thing a swatch cannot prove, and it is the one most buyers forget: the swatch cannot prove it came from this plant. In a belting cluster with dozens of works in the same few towns, a sample can be a cut from a neighbour's roll. That is why the sample and the video walkthrough of the press and vulcanising line belong together. One shows you the recipe; the other shows you the kitchen.

Turning the swatch into a golden sample

If the swatch passes your bench check, do not throw it in a drawer. It is the most useful contractual object you will hold until the first roll arrives.

  1. Photograph it — top, bottom, and the cut cross-section with a rule in frame.
  2. Record your readings on a single sheet: ply count, top and bottom cover gauge at five points each, total thickness, Shore A at four points, weight per square metre, edge type, peel observation.
  3. Cut it in half. Keep one half. Send the other half back to the supplier, or ask them to retain a matching cut from the same piece, with both halves initialled and dated.
  4. Reference it in the purchase order as the approved sample, by date and the construction on the label, with a line stating that production belt is to match the approved sample on cover gauge, hardness and construction within the tolerances of the standard you name.
  5. Ask for the same readings on the production roll, taken at both ends and the middle, before shipment — and compare them to your sheet.

This is the same discipline as any sample-to-production handover in China, applied to a product that cannot be re-cut once it is in a shipping container. The general sample-ordering process covers payment and courier mechanics; belting adds the cross-section, the peel and the weight.

Where a trial roll fits

For belting, the gap between swatch and container is unusually easy to bridge, because a stated MOQ of one piece means you can order one roll. Jiutong's profile states MOQ 1 piece and production in 15–20 working days, with below-MOQ quantities at around 10 working days — company-stated terms, to be confirmed at quote. A single roll, run on your own line for a month, answers every "no" in the table above except the laboratory ones: camber, tracking, splice, edge wear, cover wear under your material. The economics and mechanics of that step are covered in ordering a single trial roll before a container.

The sequence that works is: swatch → bench check → golden sample → one trial roll → third-party test on a cut from that roll → container. Each step is cheap relative to the one after it, and each step is allowed to end the process.

Common questions

The sample was free. Does that mean the trial roll will be at sample price?

Not necessarily, and the fact basis for Jiutong does not say either way. Ask two separate questions: what a custom sample costs, and whether the sample or trial-roll cost is credited against the first production order. Get the answer in the quote, not in a chat message.

The swatch reads 3.5 mm top cover but I ordered 4 mm. Is that a rejection?

It depends on the tolerance your specification allows. Belt standards carry a thickness tolerance on covers and total thickness, and a supplier working to a named standard should be able to tell you which tolerance they hold. If neither your spec nor the quote names a tolerance, that is the problem to fix before the roll, not the 0.5 mm.

Can I send the swatch to a lab?

You can, and for hardness, cover thickness and a basic tensile on a small specimen it is worth doing. Be aware that a sample-built piece can pass tests a production roll would fail, because it was built with more care. The report you want on file before the container is one on a specimen cut from a production roll, ideally the trial roll.

The supplier is only two years old. Does the sample carry more weight or less?

Less, on its own. A company registered in 2024 has a short shipping record to lean on, so the profile's own advice is to ask for shipped-order references and third-party test reports rather than relying on system certificates or a good-looking sample. The swatch is the start of that evidence file, not the whole of it.

What to ask the supplier next

Send these with your bench-check readings attached:

  • Was this swatch cut from a production roll or built as a sample, and on which press?
  • Which fabric grade and which cover compound went into it, by name, and will the same go into the trial roll?
  • What is the sampling cost for a custom piece, and is it credited against a first order?
  • What thickness and hardness tolerance do you hold, and to which standard?
  • Will you retain a matching half of this swatch, initialled and dated, as the reference for production?
  • Can you take gauge and durometer readings at both ends and the middle of the trial roll and send them with photos before shipment?

Jiutong Conveyor Belting is a workable example of a supplier this sequence suits: company-stated standard samples in 5 working days and custom in 10, MOQ of one piece, production 15–20 working days, and a ChinaMakersHub profile that plainly lists what has and has not been verified about the company. Treat the swatch as the first document in that file, and the pre-shipment inspection on the roll as the one that closes it.