Moisture, Hot Tires and Cold Cure: Why Floor Coatings Fail on Long Island Slabs

A failed garage floor coating almost never fails because the resin was bad. Modern epoxies, polyureas and polyaspartics are mature chemistry sold by manufacturers who publish real data sheets. Coatings fail at the interface - either the bond between resin and concrete, or the bond between one coat and the next - and on Long Island there are three things that reliably break those interfaces. Water vapor moving up through the slab. Hot tires pulling down on a film that was never properly anchored. And a cure that ran too cold or too slow to finish cross-linking before the garage went back into service. Every one of those is testable and every one of them is cheap to check compared with paying twice.

This page is the technical version of a conversation we have on most estimate visits in Smithtown. It covers what actually holds a coating on concrete, why Suffolk County ground makes moisture the dominant failure mode here, how epoxy and polyaspartic behave differently in a climate that records 101 nights a year at or below freezing, what hot tire pickup really is, and how to read the surface preparation you are being sold. There are no product recommendations in it. The point is to give you enough to evaluate any quote you receive, including ours.

What actually holds a coating onto concrete?

Mechanical interlock, almost entirely. Resin does not form a meaningful chemical bond with cured portland cement. What it does is flow into the pores and the roughness of the surface, cure hard, and lock itself into that geometry. Everything about adhesion follows from that. A smooth, hard-troweled garage floor has almost no geometry to lock into. A floor with a curing compound or an old sealer on it has none at all, because the resin is bonding to that layer rather than to concrete. And a floor that was ground open and then primed over its own dust has a bond line made of powder.

That is why adhesion is measurable rather than a matter of opinion. ASTM D7234 covers pull-off adhesion strength of coatings on concrete using portable testers: a dolly is glued to the finished floor, pulled perpendicular until something lets go, and the failure plane is recorded. The plane matters more than the number. If the dolly comes away clean with bare concrete under it, the coating never bonded. If it comes away with a layer of concrete stuck to it, the coating was stronger than the substrate and the slab surface was the weak link. Those two results point at completely different fixes, and neither one is buying a more expensive bucket of resin.

• Adhesion on concrete is mechanical interlock, not chemistry - profile is the product.
• Curing compounds, sealers, old paint and grinding dust all become the bond line if left in place.
• ASTM D7234 pull-off testing reports which plane failed, which tells you what to fix.
• A coating is only ever as strong as the top quarter inch of concrete beneath it.

Why does the slab under a Smithtown garage hold so much water?

Because of what is under it and what was not put between. USGS mapping of the Huntington-Smithtown area describes glacial deposits that are largely crudely stratified sand and gravel with kames and interspersed kettles, running more than 300 feet thick beneath the terminal moraines and buried valleys, plus a south-sloping outwash plain of stratified sand and gravel. The dominant soils mapped across this part of Suffolk County, the Haven and Riverhead series, are both classified well drained, but USDA describes their saturated hydraulic conductivity as very high in the substratum. Well drained does not mean dry. It means water arrives fast and keeps moving, and a slab poured on that material has a continuous supply available at its underside.

The thing that is supposed to interrupt that is a vapor retarder - a polyethylene sheet between the slab and the base course. That is a modern requirement. The 2021 International Residential Code sets the minimum at 10 mil material conforming to ASTM E1745 Class A with joints lapped not less than six inches, and earlier editions specified 6 mil, with exceptions in both. A garage slab poured in Smithtown or St. James decades before any of that has nothing beneath it but graded fill and is in direct capillary contact with the ground. Even where a sheet exists, one torn during concrete placement performs about as well as none at all. So the build year is a hint, not an answer, and the only thing that settles it is a test.

• ASTM D4263, the plastic sheet method, is the cheap screen: an 18 inch square of poly taped down and read for condensation.
• ASTM F1869 is the calcium chloride test, reporting pounds of moisture per 1,000 square feet over 24 hours.
• ASTM F2170 places probes inside the slab and reports relative humidity within the concrete itself.
• None of those documents set a pass limit - they define methods. The limit comes from the coating manufacturer.
• ASTM F3010 covers the class of two-component resin membrane-forming moisture mitigation systems used when readings run high.

Epoxy or polyaspartic in a climate that freezes 101 nights a year?

The Islip station - Long Island MacArthur Airport, the nearest long-record weather station to Smithtown - shows 101 days a year with a minimum temperature at or below 32 degrees over a record running from October 1963 through December 2023. The 1991 to 2020 normals put the daily maximum above freezing and the daily minimum below it in December, January and February, with March sitting right on the line at a 32.0 degree normal minimum against a 47.7 degree normal maximum. An unheated garage tracks that pattern with a lag. The slab is a large thermal mass, so it runs colder than the air on a mild winter afternoon and takes days to come up.

That matters because epoxy cure is strongly temperature dependent. Standard epoxy formulations want the substrate at roughly 55 degrees and rising, and cure slows dramatically below it - a film that would be hard in twelve hours at 75 degrees can still be soft days later at 45. A soft film is not just slow, it is vulnerable, because it goes into service before it has finished cross-linking and it takes damage it will never recover from. Polyaspartic and polyurea chemistry was developed in part to solve exactly this, curing well into freezing conditions and reaching service hardness in hours rather than days. The tradeoff is pot life measured in tens of minutes and a much less forgiving application.

• Epoxy: cheapest per mil, best film build, ambers under UV, needs the slab near 55 degrees and rising.
• Polyaspartic: cures cold and fast, aliphatic so it holds color under sunlight, builds thin per coat, short working window.
• The common answer on Long Island is both - epoxy or a moisture-tolerant primer for build, polyaspartic clear for the wear surface.
• Dew point is the other cold-weather limit: a slab within about 5 degrees of dew point is collecting condensation, and Islip averages 76 percent morning relative humidity.

Hot tire pickup: what it is and what actually stops it

A tire that has been running at highway speed generates real heat, and it arrives home hot, soft and slightly tacky. It parks on the coating and cools while pressed into it under the weight of the car. If the film underneath is thin, undercured, or not properly anchored to the concrete, the rubber wins the contest when the car pulls out and takes a patch of coating with it, in the shape of the contact patch. The tell is unmistakable: a rounded rectangle of bare or primer-colored concrete exactly where a tire sat, often with the film peeled back at one edge like a lifted sticker.

The fixes are not exotic. First, real mechanical profile, so the coating is locked into the concrete instead of sitting on it - this is where acid-etched and scuffed floors lose. Second, film build. A 3 to 5 mil roll-on kit has no reserve; a 20 to 30 mil broadcast system does. Third, full cure before vehicles return. A polyaspartic topcoat is generally good for vehicles at about 24 hours, but a cold epoxy build can want five to seven days, and driving in early is a decision you cannot undo. Fourth, the topcoat chemistry itself - an aliphatic polyaspartic clear does not soften under warm rubber the way a thin, undercured epoxy will.

How to read the surface prep you are being sold

Preparation is invisible once the floor is finished, and it is the largest labor line in the job, which makes it the first thing a cheap quote removes. It is also the one part of the process with a published, physical standard attached. ICRI Technical Guideline 310.2R-2013 defines ten Concrete Surface Profile levels, CSP 1 through CSP 10, and sells molded replica chips a contractor can lay against the floor to verify what was actually achieved. Profile is matched to film thickness: thin sealers key into a shallow profile, general-purpose coatings sit in the middle of the scale, and heavy overlays need a deep one. For a typical residential garage system at 15 to 25 mils, a diamond-ground floor around CSP 2 to CSP 3 is the usual target and the primer data sheet is the final authority.

Acid etching cannot be placed on that scale, which is the whole argument against it. It relies on a reaction with cement paste that is uneven or absent on a hard-troweled or previously sealed floor, it cannot touch an existing coating, it leaves salt residue that must be neutralised and rinsed, and it produces no profile anyone can hold a chip against and verify. Diamond grinding and shot blasting both produce verifiable, repeatable results - grinding at the shallower end while also flattening the surface, blasting deeper and faster on large open floors but unable to reach edges. On a Suffolk County garage the honest answer is nearly always grinding for the field with an edger around the perimeter.

• Ask what CSP number the contractor is targeting and which primer data sheet requires it.
• Ask for the dry film build in mils across the whole system, not the number of coats.
• Ask what the moisture test was, what it read, and what limit that reading was compared against.
• Look for a shrouded grinder on a HEPA extractor and a vacuum pass before primer - not a broom.