Can I compare tensile strength between two epoxy data sheets?

Only if both sheets name the same test method and conditions. Most epoxy sheets we read use ASTM D638, the method for plastics, but Key Resin and Leggari test their polyaspartics by ASTM D412, a rubber and elastomer method, and Simiron tests its polyaspartic by ASTM D2370, which pulls the coating as a free film. One Armorpoxy sheet uses D412 for tensile strength and D638 for elongation. Figures from different methods are not on the same scale.

What does concrete failure mean in a pull-off adhesion result?

It means the test pulled out a piece of the slab before the coating let go. Sika, Sherwin-Williams, APF and Elitecrete all report their adhesion that way. The psi beside it is then the strength of that piece of concrete, not the limit of the coating, which is why sheets write it as a minimum such as more than 400 psi. A lower figure on another sheet with concrete failure does not mean a weaker coating.

Is a lower Taber abrasion number always better?

Lower weight loss is better only between results run the same way: the same wheel, the same load per wheel and the same number of cycles. Sika and Simiron state all three. RINOL uses the softer CS-10 wheel, Elitecrete gives a wheel and load but no cycle count, and Ghostshield gives a weight loss with no wheel, load or cycle count. LATICRETE uses a different test entirely, falling sand. Those results cannot be ranked against each other.

Which viscosity figure on the sheet matters?

Usually the mixed figure, at the temperature the sheet states. Elitecrete prints part A, part B and mixed separately, and they differ several times over. Temperature moves it too: EcoPoxy measures at 28°C, Elitecrete at 23°C, and SureCrete warns that its part B thickens below 60°F until it needs a heated space to pour. Westcoat prints viscosity in two units, cPs and KU, which are different instruments and do not convert by eye.

Epoxy Data Sheet Performance Properties: What the Strength Numbers Mean

The performance block is the part of a data sheet that looks most like hard evidence and compares worst. Every figure in it belongs to a test method and a set of conditions, and two makers rarely use the same ones.

Quick answer

The performance block on a coating data sheet lists lab results: tensile strength and elongation, compressive strength, abrasion, adhesion and viscosity. Each one is only meaningful beside the test method and conditions printed with it, and makers use different methods for the same property. Before comparing two sheets, check that the method, the wheel or specimen, the temperature and the cure age match. For adhesion on concrete, read the failure mode: concrete failure means the slab broke first.

Where this block sits, and what it is

Most floor coating sheets carry two kinds of number. The application fields, meaning solids, coverage, pot life, recoat and cure, tell you how to put the product down, and how to read a technical data sheet works through those. The second kind sits in a table headed physical properties, performance properties or technical data, and it describes what the cured material did in a laboratory test.

That table usually tells you how it was made. Sika prints every property on its Sikafloor 264 sheet at 73°F (23°C) and 50% relative humidity. Ghostshield says its results are averages obtained in a controlled environment and that reasonable variations should be expected. RINOL calls its figures approximate values that do not constitute a guarantee. Read the block as a description of a sample in a lab, not a promise about a floor.

Hardness sits in the same block and has its own page: Shore hardness covers what the Shore D figure measures and why harder is not automatically better. The rest of this page is the other six properties.

Read the test method before the number

Every row in the block names, or should name, a test method, usually an ASTM number. That code is not decoration. Two sheets can print the same property under the same heading from tests that are not built to agree, and the only place that shows is the method column.

Tensile strength is the clearest case. Most epoxy sheets we read test it by ASTM D638, the tensile method for plastics. Key Resin and Leggari test their polyaspartics by ASTM D412, a method written for rubber and elastomers. Simiron tests its Polyaspartic HS by ASTM D2370, which pulls the coating as a free film. One Armorpoxy sheet prints tensile strength by D412 and elongation by D638, two methods for two halves of the same pull. A figure from one of those is not on the same scale as a figure from another, however alike the units look.

Tensile strength and elongation

Both come from one test. A shaped specimen of cured material is pulled until it breaks. Tensile strength is the stress it carried at that point, in psi or MPa. Elongation, or elongation at break, is how far it stretched before it went, as a percentage of its starting length. A stiff material stretches little before it breaks and a flexible one stretches more, and the polyaspartic sheets we read print higher elongation than the epoxy ones.

Two things limit what either figure tells you about a floor. The specimen is not bonded to anything: ASTM D2370 tests coatings as free films, and its own scope notes that tensile results vary with specimen thickness, preparation and how fast the load is applied. A coating on a slab is held by the concrete under it, so it is not free to stretch the way the specimen did. And some figures are not for the coating alone: DCP footnotes the tensile figure on its Strongcoat PA500 sheet as measured on a full system with its recommended primer.

No sheet we read says how much elongation a floor needs, or claims that a given figure lets a coating ride over a moving crack. Where a slab moves, the joint and crack treatment decides what happens, which concrete floor joints explained covers.

Compressive strength: a block of resin or a mortar

Compressive strength is the load a specimen takes in crushing. On a coating sheet it is often a large figure, and the method explains why. Simiron, Elitecrete, Westcoat and APF test by ASTM D695, a method for rigid plastics, on a cast piece of the resin. Sika tests Sikafloor 264 by ASTM C579, which is written for chemical-resistant mortars, grouts, monolithic surfacings and polymer concretes, so the specimen can be a filled mix rather than neat resin. RINOL quotes compressive strength for its whole system against DIN EN 196 and ASTM C109, standards from the cement world.

A thin film on a slab passes the load straight into the concrete, so the figure says most about thick, filled builds where the coating is the wearing body of the floor, such as epoxy mortar and slurry. RINOL makes the point from the other side: its system sheet asks for a substrate with a minimum compressive strength of its own, because the coating cannot supply what the slab lacks.

Abrasion: a Taber number needs its conditions

The usual abrasion test for floor coatings is ASTM D4060, the Taber test. A coated panel turns under two abrasive wheels for a set number of cycles, and the result is the weight it lost, in milligrams. Lower is better, but only against a result run the same way, and three settings change it:

The wheel. The CS-17 wheel abrades harder than the CS-10.
The load. D4060 applies 1,000 g to each wheel, and a sheet should still say so.
The cycles. More turns, more loss.

Some sheets give all three. Sika prints CS-17, 1,000 cycles and 1,000 g. Simiron prints CS17 wheel, 1000 g load, 1000 cycles. Others do not. Sherwin-Williams names the wheel and the cycles but not the load. Elitecrete names a CS-17 wheel and a 1 kg load and no cycle count. RINOL uses the softer CS-10 wheel against DIN 53754 and ASTM D1044. Ghostshield prints a milligram loss with no wheel, load or cycles at all. LATICRETE uses a different test entirely, ASTM D968, which measures how much falling sand wears through a dry mil of coating. None of those ranks against another.

Even a clean comparison is a comparison of coatings under a wheel. How a floor wears in service also depends on what is carried across it and what the top layer is, which is why a broadcast floor’s wear is discussed in flake vs quartz rather than in a Taber figure.

Adhesion: read the failure mode, not just the psi

Adhesion on concrete is usually a pull-off test. A metal dolly is glued to the cured coating and pulled straight up until something gives. ASTM D7234 is the method for coatings on concrete, and ASTM D4541 covers other rigid substrates; Sherwin-Williams cites ACI 503R. The result is two things, a force and where the break happened, and the second matters more.

On a sound coating the weak point is usually the slab. Sika reports more than 400 psi with 100% concrete failure. Sherwin-Williams reports 300 psi, concrete failure. APF reports more than 400 psi with the concrete failing, and Elitecrete reports concrete failure too. In each case the test pulled out concrete before the coating let go, so the psi measured that piece of slab, not the limit of the bond. That is why the figures carry a more-than sign, and why a lower figure with concrete failure on one sheet does not mean a weaker coating than a higher one on another.

It also explains why the slab gets its own requirement. LATICRETE asks for a prepared surface with a tensile pull-off strength of 200 psi or greater, and RINOL for a substrate tensile strength of at least 1.5 N/mm². A bond can only be as strong as the concrete it holds, which is why surface prep decides more than the adhesion row does, and why peeling is almost always a prep story.

Not every adhesion row is a pull-off. Westcoat rates EC-36 by ASTM D3359, a tape test scored on a scale rather than in psi, and gives a result on steel as well as on concrete. A tape rating and a pull-off figure are different tests, and a result on steel tells you nothing about a slab.

Viscosity: which part, at what temperature

Viscosity is how thick the liquid is, in centipoise (cP or cps). It belongs to application more than performance, but it is printed in the same block and it is the figure most often compared loosely. Three things change it:

Which liquid. Elitecrete prints part A, part B and the mixed product separately for E100, and part A is several times thicker than part B. Simiron prints only the mixed figure for 1100SL. Compare mixed against mixed.
The temperature. Elitecrete measures at 73°F (23°C) and EcoPoxy at 28°C. Cold thickens resin: SureCrete warns that below 60°F its part B will thicken until it needs moving into a heated space to pour.
The unit. Westcoat prints viscosity in both cPs and KU. Those come from different instruments and do not convert by eye.

A single figure also cannot show a material whose thickness changes as it is worked, which is what thixotropic means. Where a low figure genuinely earns its place is a primer, which has to get into the surface of the slab, as epoxy primers for concrete floors explains.

Comparing two sheets

For any property you want to compare, check five things before the number:

The test method is the same, not just the property name.
The specimen conditions match: wheel, load and cycles for abrasion, liquid and temperature for viscosity.
The temperature, humidity and cure age the sample was tested at.
Whether the figure is for the product alone or for a full system with its primer.
For adhesion, where the break happened.

If a sheet leaves one of those out, the figure cannot be compared, and the maker’s technical service line can usually supply the missing condition. And remember what no sheet we read offers: a figure that a given floor needs. The block describes a product. Which system a floor needs is a different question, and the epoxy system builder starts from the floor rather than the sheet.