Concrete Garage Floor Finishes

Start here: your garage slab probably has no vapor barrier

The IRC requires a vapor retarder under a concrete slab on ground, and then it exempts garages by name. Section R506.2.3 requires the sheeting "between the concrete floor slab and the base course or the prepared subgrade", with the first exception covering "garages, utility buildings and other unheated accessory structures". Newer code editions raised the requirement to a 10-mil ASTM E1745 Class A retarder and kept the exemption for unheated accessory structures.

Now put that next to what coating manufacturers require. One resinous flooring specification states that "due to hydrostatic, capillary and moisture vapor pressure, substrates in contact with ground must have a properly installed, effective vapor barrier." The code does not require the thing the coating requires. That single mismatch explains most garage floor coating failures, and almost nobody selling a garage floor kit mentions it.

What goes wrong is not mysterious. Water vapor moves up through the slab from the ground, reaches the underside of a film that will not let it through, and pushes. Sherwin-Williams describes the result directly: moisture "can react with the coating's adhesive layer, forming bubbles or causing the finish to lose adhesion", and when a coating goes down before moisture stabilizes "the vapor pressure beneath the surface can cause blistering or delamination."

Test before you coat, and know that the kit will not tell you to

Professional specifications set numeric limits. The common ones:

• ASTM F1869, the calcium chloride test, measured in pounds of moisture per 1,000 sq ft per 24 hours. Sherwin-Williams adopts "no more than 3 pounds"; other specifications allow 5.
• ASTM F2170, in-situ relative humidity probes. The usual limit is 75 percent, though at least one polyaspartic allows 80.

Consumer garage kits require neither. They ask for a taped-down plastic sheet left overnight, which is the ASTM D4263 check. That is worth doing, but understand its limit, and Sherwin-Williams states it plainly: the plastic sheet test indicates moisture is present but "will not quantify the amount of moisture movement." The same coating chemistry is sold to professionals with a numeric acceptance criterion and to homeowners with a plastic bag. If you are about to spend a weekend and a few hundred dollars, a calcium chloride kit is cheap insurance.

Hot tire pickup, and why it is not a heat problem

This is the defining garage floor failure: the coating lifts in two strips exactly where the tires park. The usual explanation is that tires get extremely hot and cook the coating. That is not what happens.

Euclid Chemical's technical bulletin names the mechanism: "tire marking or hot tire pickup on concrete sealers and coatings is caused by a phenomenon called plasticizer migration." Plasticizers are the compounds that keep tire rubber flexible. Driving warms the tire, the plasticizers soften and leach out, and they attack the coating film from above while a hot tire presses on it.

The temperatures involved are lower than the folklore. The only primary measurement study we could find, a US Army tire temperature program, put passenger tire tread at roughly 100 to 120 degrees F after highway running at normal inflation, reaching about 130 only at a badly deflated 14 psi. The widely repeated "tires reach 200 degrees" has no measurement behind it.

That reframing matters, because it changes the fix. A coating that lifts under a tire is usually failing at the bond, not from heat. Poor surface preparation and moisture are the causes; the tire is the load that finds them.

What resists it, per the same bulletin, is crosslink density. Acrylic and styrene-acrylic cure-and-seals have minimal crosslinking and mark most. "Epoxy and urethane sealers and coatings are much more crosslinked and are therefore less likely to tire mark, but it is still possible." Note "less likely", not immune. Euclid's own recommendation where tire marking is a concern is a penetrating silane or siloxane sealer, which has no film to lift at all.

The manufacturers are not unanimous here, which is worth knowing: the ASCC sealer guide lists hot tire resistance as a property of solvent-based acrylics, while Euclid names acrylics as the most susceptible chemistry. Both are industry sources and they disagree.

Surface preparation, and whether acid etching really works

Coating adhesion needs a profile to key into, and the industry measures it on the Concrete Surface Profile scale, published by the International Concrete Repair Institute as guideline 310.2R. It runs CSP 1, nearly smooth, to CSP 10, very rough. What each method achieves:

And what coatings ask for, by film thickness: thin film up to 20 mils wants CSP 2 to 3, medium film up to 40 mils wants CSP 3 to 5, and self-leveling mortars want CSP 4 to 6.

So the honest answer on acid etching is neither "it never works" nor "it is fine". Etching reaches CSP 1 to 3, which is adequate for the thin-film products that specify CSP 1 to 3, and inadequate for anything specifying CSP 3 to 5. It is a standardized practice with its own ASTM method, and one major manufacturer's two-part industrial epoxy names CSP 1 to 3 and references that method directly. Every consumer garage kit we checked still ships etch rather than telling you to rent a grinder, and for those products that is defensible. It stops being defensible the moment the product is a 100 percent solids or high-build system.

One detail from a kit data sheet worth repeating: it tells you not to use the etch on a basement floor, and to use a TSP solution instead. Same product, different slab, different prep.

The options, honestly

Before the table, one finding that shapes how to read every comparison you will see: not one manufacturer data sheet we read states an expected service life in years. Not one. Every "lasts 15 years" figure in circulation traces to marketing or editorial. Treat the life column below as the weakest column.

Bare concrete

Free, and its failure mode is dusting, which NRMCA defines as "formation of loose powder resulting from disintegration of surface of hardened concrete." It is a finishing defect, not wear: bleed water worked back into the surface during troweling produces a very high water-cement ratio in the top quarter inch. One cause is specific to garages, and it is worth knowing if you are building: running an engine or an unvented salamander heater in a closed garage over green concrete carbonates the surface and "greatly reduces the strength and hardness."

Penetrating sealer

Silane or siloxane soaks in and leaves no film. Nothing to peel, nothing for a tire to lift, and it is Euclid's own recommendation where tire marking matters. It also has published evidence behind it for the other garage problem: an Iowa State literature review found silane and siloxane sealers improved resistance to deicer scaling and reduced moisture uptake, while noting efficiency "degrades with time, and re-applications are required." It will not change how the floor looks, which is either the point or the disappointment.

Acrylic film sealer

Cheap, about $0.10 to $0.20 per square foot in material, and short lived: the ASCC sealer guide gives water-based acrylics 1 to 2 years and solvent-based 1 to 3. This is the chemistry Euclid names as most prone to tire marking. Reasonable on a floor you will re-coat; poor value under a parked car.

One-part epoxy paint

The cheapest thing in the flooring aisle, and the one to be most careful with. Data sheets typically want 7 days before tires touch it and exclude hydrostatic pressure and industrial use. The clearest evidence against it in a garage is a major manufacturer's own sheet for a one-part acrylic floor coating: "not recommended for areas subject to hot tire pickup." Its listed uses are laboratories, offices and hospitals. Garages are not on the list.

Two-part epoxy

The default answer, and a reasonable one when the prep is right. Consumer kits cover roughly 200 to 250 sq ft per pouch and want 3 days before vehicle traffic. A professional two-part product runs 3 to 5 mils dry, 206 to 350 sq ft per gallon, and 48 to 72 hours before vehicles at 77 degrees, stretching to 4 or 5 days at 50 degrees, which matters if you are doing this in an unheated garage in spring. Note what that professional sheet does not say: it makes no hot tire claim at all.

Polyaspartic and polyurea

The premium option, usually professionally installed, and genuinely better in two ways: it is more UV stable than epoxy, and it goes down in fewer visits. A manufacturer's QUV comparison to ASTM D4587 measured color change after 500 hours at 0.89 for polyaspartic against 8.83 for epoxy, where the acceptance threshold is 2.0. That is an interested test, but it names a method, a criterion and numbers. Relevant if your garage door stands open in the sun.

The "back in service in a day" claim is product specific rather than chemistry wide. A standard-set polyaspartic data sheet we read lists vehicle traffic at 7 days. Fast-set versions exist; ask which one you are buying.

Polished concrete

Covered fully on polished concrete floors, so just the garage-specific part here. It is the only option with nothing on top to fail, which neatly sidesteps both the vapor barrier problem and hot tire pickup. Against that: it is the most expensive, it needs a slab flat enough to take the finish you want, and it is not stain-proof. A densifier manufacturer says its product "will improve the resistance of concrete surfaces to staining" and then points users to a separate stain-protection product "for improved resistance to water or oily stains." In a garage, that second product is not optional.

And read that stain guard's own limitations before you commit. One states it is "not for use on surfaces exposed to standing or pooling water." A garage bay where snow melts off a car and pools is arguably outside what the product is warranted for.

Road salt, if you live somewhere it snows

A car parked in a garage all winter drips brine onto the slab every day. That is a chemical attack, not just a mess, and it does not need freezing weather to happen.

FHWA's report on deicers and concrete describes calcium hydroxide in the cement paste reacting with calcium chloride to form calcium oxychloride, a reaction that is "highly expansive, with the resulting damage likely due to crystallization pressures." The detail that matters for a garage: the phase change occurs "at temperatures between 32 degrees F and 122 degrees F", so a heated or semi-heated garage does not escape it. Magnesium chloride contributes indirectly by producing calcium chloride in the first place.

Here is the gap: no garage floor coating data sheet we read makes any quantified salt or deicer resistance claim. Not one gives a test method or a figure. Products say "excellent chemical resistance" and leave it there. For the single most important durability question a buyer in a snow market has, the industry publishes nothing.

What does have evidence behind it is the penetrating sealer route, from the Iowa State review above. It is also Euclid's recommendation against tire marking. A penetrating silane or siloxane is the one option where both of the garage's main problems have published support behind the answer, which is a strange thing to be able to say about the cheapest and least glamorous choice on the list.

A new garage slab: how long to wait

Every specification we read says 28 days, and most qualify it with a temperature: "cured at least 28 days at 75 degrees F". Consumer kits say 28 or 30 days.

But the number is a convention rather than a physical threshold, and a coatings manufacturer says so on the record: "some coatings are designed to tolerate higher moisture content, so as long as the moisture content falls within the coating manufacturer's published information, then we can coat before 28 days." The 28 days comes from compressive strength testing convention, and in a data sheet it functions as a proxy for dryness.

Which means the real criterion is the moisture number, not the calendar. On an unheated garage slab with no vapor barrier under it, that test can fail at 28 days, at 90 days, or permanently, depending on your water table and your soil. Waiting longer helps. It does not guarantee anything, and only the test tells you.

One other risk of coating early: concrete shrinks as it cures, and shrinkage cracks that form afterwards telegraph through the coating. If you are pouring the slab now, how to get a garage slab right covers thickness, air entrainment, the slope requirement and the joint layout, all of which decide how good a candidate the floor will be later.

Slope and drainage

The IRC does require slope, and notably does not say how much. Section R309.1: garage floor surfaces "shall be of approved noncombustible material" and the parking area "shall be sloped to facilitate the movement of liquids to a drain or toward the main vehicle entry doorway." The figures the trade uses instead, and where they come from, are on the concrete slope calculator: ACI 302.1R asks for a quarter inch per foot and the Portland Cement Association asks for an eighth, which is a factor of two with no national figure between them.

We could not verify any numeric slope requirement from a primary source. The 1/8 inch per foot figure that circulates is trade practice, not code, and we would rather say so than repeat it as a rule.

Slope matters to the finish choice through pooling rather than through chemistry. Standing water is where the stain guard limitation bites, where a moisture-sensitive coating gets the longest exposure, and where brine sits longest in winter. A flat spot that holds a puddle will be the first place any finish fails.

Claims the data sheets do not support

Worth knowing before you stand in the aisle reading boxes.

• "No hot tire pickup" on a product page whose own technical data sheet makes no hot tire statement at all. We found exactly this on a major consumer epoxy kit: the marketing page claims it, the TDS is silent.
• "20X stronger than epoxy" on a retail listing, naming no property and no test method, on a product whose manufacturer documentation says nothing about hot tire pickup either.
• A one-part product claiming hot tire resistance while a major manufacturer's one-part acrylic sheet says the opposite and Euclid explains why minimally crosslinked acrylics are the most susceptible chemistry.
• Any service life in years. No data sheet gives one.
• "Epoxy does not yellow." The one published QUV comparison puts gray epoxy at more than four times the accepted color change threshold after 500 hours.

So what should you actually do

In order of how we would think about it:

1. Test the slab for moisture before choosing anything. It is the only measurement that can rule options out, and it costs less than redoing the floor.
2. If the moisture number is bad, a penetrating sealer or bare concrete is the honest answer. No film will survive, and no amount of prep changes that.
3. If you want cheap and durable and do not care how it looks, penetrating silane or siloxane. Published support against both tire marking and deicer scaling, nothing to peel, reapply periodically.
4. If you want it to look finished and you are doing it yourself, two-part epoxy with proper prep, matching the CSP the product asks for rather than the one the kit makes convenient. Skip one-part paint.
5. If you are hiring it out and the budget allows, polyaspartic for UV stability and a faster return to service, or polished concrete if the slab is flat enough and you accept the stain guard as part of the job.

Related reading on this site: epoxy and concrete floor coatings for the coating types in depth, sealing a concrete floor for sealer chemistry, painting a concrete floor if paint is what you want, floor repair for cracks and pitting before you finish anything, and the concrete floors hub for the whole picture. To work out the concrete for a new slab, use the concrete floor calculator.

Sources

• Euclid Chemical CP-22, Tire Marking and Pickup of Concrete Sealers, for the plasticizer migration mechanism.
• IRC R506.2.3 as adopted, for the under-slab vapor retarder requirement and the garage exemption.
• Sherwin-Williams concrete surface preparation guide, for the moisture limits and CSP by film thickness.
• ICRI 310.2R concrete surface profile scale, for what each preparation method achieves.
• FHWA-HIF-17-008, Chemical Deicers and Concrete Pavement, for the calcium oxychloride mechanism.
• Iowa State CP Tech Center, evaluation of penetrating sealers.
• NRMCA CIP 1, Dusting Concrete Surfaces.
• ASCC Decorative Concrete Council sealer guide, for acrylic service life and cost.
• Tnemec on the 28 day cure, for why the number is a convention.