A fitness set assembled from several materials looks like one product to the buyer and behaves like four or five different substrates to the factory. EVA foam, thermoplastic elastomer, polyurethane, natural rubber and silicone each hold ink differently, accept a debossed depth differently, and fail a wear test for different reasons. The most common decoration problem in a mixed range is not a bad printer — it is a single marking specification copied from the item that was developed first and applied to everything else. This note sets out what changes from surface to surface, and how to specify a mark that survives handling rather than only surviving the sample round.
Why one mark specification does not transfer across a set
Ink adhesion is a surface-energy problem before it is a chemistry problem. A substrate that wets out easily lets the ink film form a continuous bond; a substrate with a low surface energy or a moulding aid on its skin repels the same ink, and the print looks acceptable on the day and lifts at the edges after a few weeks in a warehouse. The parts of a fitness set sit at both ends of that range, so the same pad, the same plate and the same ink batch can produce a mark that is durable on one item and marginal on the next.
The second variable is geometry. Flat moulded blocks and rolled mats present very different surfaces to a printing head: a mat has a compressible, slightly textured face and a curvature that changes contact pressure along the print, while an EVA block has sharp edges where ink film thins. A specification written as "logo, one colour, 40 mm wide" says nothing about either, so each supplier fills the gap with its own assumption.
Printing on EVA, TPE, polyurethane and silicone
EVA foam is forgiving to print and unforgiving to clean. Its closed-cell surface takes a solvent-based or UV ink well after a proper pretreatment, but the same foam is easily attacked by aggressive solvents, which can dull the surface or raise a texture around the print. The practical rule is to keep the pretreatment as mild as the ink will tolerate, and to test on the actual density: a 45 kg/m³ block and a 90 kg/m³ block do not respond identically to the same wipe.
Thermoplastic elastomer mats and bands are harder. TPE frequently carries a mould release or a bloom on the surface, and both sit between the ink and the polymer. Wiping with the wrong solvent smears the release agent instead of removing it, which produces a print that passes a fingernail test and fails after a month of cleaning. Where a mark has to be permanent on TPE, the more reliable route is often a moulded-in or debossed feature rather than an applied ink.
Polyurethane has the widest process window of the group but is sensitive to moisture during cure, so a print applied in a humid month can behave differently from the same print in a dry one. Silicone is the least cooperative substrate for conventional inks: its low surface energy means standard pad printing sits on top rather than bonding, and a durable mark usually requires a silicone-compatible ink with a proper post-cure, or a moulded-in mark. On a mixed set this is the item that most often needs a different decoration method rather than a different ink colour.
Debossing: depth, wall thickness and legibility
Debossing looks like the simpler option because it removes the ink question, but it moves the constraint into tooling and wall thickness. A debossed logo needs enough material behind the recess that the part does not thin to the point of tearing or shining on the reverse face, and the achievable depth scales with the thickness of the section rather than being a fixed number. On a 4 mm mat skin the depth that reads crisply is a fraction of what a 25 mm EVA block will accept.
Legibility has its own limit. Fine counters and thin strokes fill in or round over when the recess is shallow, and they tear when it is deep, so the artwork has to be redrawn for the process rather than scaled down from a print-ready file. A mark with a 1 mm stroke that prints well may need 1.5 mm or more as a deboss, and that change is a design decision, not a factory adjustment.
Debossing also interacts with colour. A recess in a dark part reads through shadow, not contrast, and it disappears under the low, diffuse light of a studio floor. Where the mark has to be read at a glance — size, weight, resistance level — a deboss alone is usually the wrong choice, and a filled or printed element is added on top.
What an adhesion test can and cannot tell you
Adhesion testing is worth specifying, and it is worth understanding before it is quoted. ASTM D3359, the standard test methods for rating adhesion by tape test, is the reference most suppliers mean when they say the print was tested. In its current form, D3359-23, it rates adhesion on a 0 to 5 scale after a tape is applied over cuts in the film and removed, with Method A intended for films above 125 μm and Method B below that; the standard notes that no single tape is likely to be suitable for all coatings, and that the result indicates a minimum bond strength was met rather than an absolute force.
Two limitations matter for a mixed set. First, the method was developed with metal as the substrate, and the standard itself records that precision and bias data for non-metallic substrates such as plastics is lacking; a similar method, ISO 2409, permits testing on non-metallic substrates. A pass reported on an EVA block and a pass reported on a coated metal insert therefore do not carry the same weight, even when both are written as the same grade. Second, the scale is coarse by design — it is described as insensitive to all but large differences, and operators are told not to use intermediate values. It detects that a mark is failing badly; it does not rank two adequate marks.
The test is also operator-dependent: the standard points to the angle and rate of tape removal as key steps, and states that results between operators can differ. When a supplier reports a grade, ask which method, which substrate and who performed it. The useful evidence is not the number but whether the item tested was the item you are buying.
Approving a sample across a mixed set
Sample rounds fail in a predictable way: the item that was decorated first is approved, and the rest of the set is assumed to follow. Approve the whole set in one round instead, with one mark from each surface, and compare them side by side under the same light. Colour that matches on a white block and on a black mat is not the same colour measurement, and a single approved swatch cannot stand in for both.
Condition the samples before judging them. A print judged straight from the sample bag has not seen warehouse heat, a cleaning wipe or a month of handling, and those are the conditions that separate a durable mark from a marginal one. Where the intended use includes repeated cleaning, the sample should be wiped on the same schedule the product will meet in service.
Finally, record the approved sample as a physical reference with a date and a revision, and keep it with the artwork file. A photograph in an email is not a standard; a retained part with a signed label is.
Writing a marking specification a factory can follow
A workable specification is a table, not a paragraph: one row per surface, with the decoration method, the artwork file, the colour reference, the pretreatment, the cure condition and the acceptance criterion for that row. Where the same logo is applied five ways across one set, say so explicitly and give the reason, so that a new supplier does not "harmonise" the methods back into one and quietly change what the buyer approved.
Add the handling rules that decide whether the mark survives the first week: nesting and stacking, the use of a liner between printed faces, and how long the parts cure before packing. Most early print failures we see are not adhesion failures at all; they are parts packed face-to-face while the ink was still soft.
Our customisation programme treats decoration as a per-surface specification for exactly this reason, and the production background explains which processes we run in-house and which we qualify with partners. The development notes record sample rounds where a mark had to change method rather than change colour. If you are preparing a mixed set, send the artwork and the material list and we will return a per-surface marking table before tooling is cut.
Frequently asked questions
Can the same pad print be used on every item in a set?
Rarely. EVA, TPE, polyurethane and silicone differ in surface energy and in what sits on the surface after moulding, so the ink, the pretreatment and sometimes the method itself have to change per substrate.
Is a debossed logo always more durable than a printed one?
It removes the ink adhesion question, but it introduces depth limits set by wall thickness and it can lose legibility at small sizes. On dark parts a deboss alone is often hard to read and needs a filled element.
What does an ASTM D3359 grade actually prove?
That a tape applied over cuts in the film removed less than a defined amount of coating, under one method and on one substrate. It indicates a minimum bond strength was met and does not rank two adequate marks or transfer between different substrates.