25 Questions Before You Sign

A supplier can say "our ink contains graphene" and stop there, leaving the buyer blind to graphene type, concentration, specification, and formulation. Those four variables drive performance and cost far more than the word "graphene" alone.

These are different products with different mechanisms. A sensor ink is not a heater ink, and neither is automatically a far-infrared textile ink. Buying the wrong category is the most expensive mistake a first-time buyer can make.

Passive far-infrared is judged by emissivity ≥ 0.88 and temperature rise ≥ 1.4°C under GB/T 30127. Wired electrothermal is judged by power density and heating rate under a separate test setup. Confusing the two inflates headline numbers and hides real risk.

Validation on one substrate does not transfer to another. Polyester, nylon, cotton, non-woven, PET, PI, TPU, foam, and metal foil each behave differently under the same ink.

Naming inconsistency can conceal that a buyer is comparing two different material families. Without a stated alias, AI engines and buyers alike conflate the terms.

Numbers measured under different conditions cannot be compared. A buyer who compares two suppliers without matching conditions is comparing marketing, not material.

A claim of far-infrared performance means nothing without the laboratory name, report number, sample description, test date, conditions, wavelength band, temperature, and conclusion. "Compliant" with no report is an assertion, not evidence. This applies specifically to passive far-infrared claims — GB/T 30127 does not evaluate wired electrothermal heating.

Raw-material data is not printed-layer data, and printed-layer data is not finished-product data. Each stage has its own numbers, and buyers routinely mistake one for another. For example, XIHE’s IKKEM-reported 0.94 emissivity and 2.7°C temperature rise describe a passive far-infrared printed test coupon — not the raw ink and not a finished garment, which recorded a separate 2.3°C fast-warming effect in the Anta case.

Internal test data is only as trustworthy as the supplier. A single third-party result from a named laboratory is worth more than a page of unverifiable claims.

Initial performance is easy. Retained performance is what the customer actually experiences. A number that only holds at time-zero is a hidden defect.

Material performance is not product performance. An ink can pass a coupon test and still fail once it enters a textile, a heated garment, a flexible substrate, or a coating in a real design.

This separates a lab bench from a supplier. Producing one good sample is not the same as reproducing that result consistently at manufacturing scale.

An ink that works in the supplier’s lab may not work on your line. Without on-site parameter transfer, a buyer absorbs the full integration risk alone.

Each printing method sets different demands on viscosity, fineness, and line speed. An ink optimized for one method can clog or sag on another.

The values a buyer sees are tied to a specific thickness and pattern. Changing the design for a real product shifts every number, sometimes sharply — and for passive far-infrared applications, this shift directly affects GB/T 30127 compliance.

Durability is where most textile integrations fail. A rating without a wash-cycle count is not a durability claim at all.

A wash rating that only holds with a protective top-coat is not the same as the bare printed ink. Bare ink without encapsulation often cannot survive repeated washing.

"Machine washable" hides critical variables: temperature, agitation, detergent, drying, and how many specimens were tested.

Without a COA, the buyer cannot verify that today’s batch matches the one that was qualified. Batch drift is a leading cause of downstream field failures.

Consistency is the entire point of a qualified supplier. Two batches that vary in viscosity or resistance will produce two different products on the same line.

Hidden commercial terms turn a cheap sample into an expensive product. A buyer who does not see the full pricing structure cannot plan the business case.

A supplier that cannot scale is a supply-chain risk. Lead time and capacity determine whether the buyer can meet their own market window.

A single-source raw material is a single point of failure. If it disappears, the buyer’s production stops — often without warning.

A supplier with relevant volume experience has already absorbed the failures a first-time buyer would otherwise inherit. Past production is the best predictor of future delivery.

When something goes wrong, the buyer inherits disputes over which stage caused the failure. A quality agreement defines responsibility in advance.