Will the Hollow Sound Come Back After Injection Repair?

Injection repair works by drilling a small hole into a grout joint or the tile face, pumping epoxy or polyurethane resin into the void beneath the tile, and letting it cure into a rigid bond between tile and substrate. It's standing in for the mortar contact that should have been there from the start — ANSI A108.02 requires 80% mortar coverage in dry interior installations and 95% in wet or exterior ones. When the original setter simply missed that coverage and nothing else about the floor is wrong, the resin fills the exact gap the mortar left, and the repair holds indefinitely because the installation was the problem, not the substrate underneath it.

That's also why the same repair can look identical on two floors and behave completely differently a year later. The resin doesn't know or care why the void was there — it just fills it. Whether the sound stays gone depends entirely on what was moving, or not moving, before the hole was drilled.

Injection repair doesn't correct structural deflection. The TCNA Handbook calls for tile substrates to hold to a deflection limit of L/360 — a joist span shouldn't flex more than 1/360th of its length under load. If the floor still flexes past that limit after the resin cures, the same flexing that broke the original bond keeps working on the new one, and the hollow sound tends to return in the same spot, sometimes within a single heating and cooling season.

Missing movement joints cause the same pattern for a different reason. TCNA guide EJ171 calls for a soft, flexible joint roughly every 25 ft in an interior tile field and every 12 ft where the floor is exposed to direct sun, exterior weather, or radiant heat. Skip those joints and the field has nowhere to absorb seasonal expansion and contraction — the stress concentrates at the nearest grout line, transmits through the new resin bond, and reopens the void right where the repair was made.

Moisture is the third variable, and the hardest to see. A slab that's still curing, wicking ground moisture, or expanding under a leak keeps the substrate dimensionally unstable no matter how well the resin cured. Injection repair on an actively wet substrate tends to fail faster than a repair on a floor with a coverage gap alone, because the resin is bonding to a surface that's still changing shape underneath it.

Ignoring a hollow sound that keeps returning doesn't freeze the problem in place. Each cycle of movement that reopens an injection repair also works on the tiles around it: grout joints crack first, tile edges start to lip, and the hollow area creeps past the roughly 5% of a tile's face that NTCA guidance treats as normal variation into territory that reads as an actual bonding failure. A full diagnosis of that progression, including how lippage and cracked grout show up together, is covered in the hollow-sounding floor tile guide.

What started as one tile treatable with a resin shot turns into a section that needs the tile pulled, the substrate corrected — deflection braced, a movement joint cut in, moisture resolved — and new tile set to match. At that point the job shifts from a repair budget to a reinstall budget, and it's worth running the affected square footage through the floor tile calculator or the tile estimator before committing to a second round of injections, since replacement material, matching lot numbers, and mortar quantity all change the cost picture more than another few holes of resin would.

Map the hollow area as a share of the total floor before deciding anything. If it's stayed under the 5% NTCA threshold and hasn't grown since the first repair, the original fix is probably permanent and there's nothing further to do. If the affected zone has expanded, a moisture meter reads high, or the run between the last two movement joints is longer than the 25 ft interior / 12 ft exterior spacing TCNA recommends, treat the repeat sound as a diagnosis problem rather than a repair problem — get the cause identified before paying for a second injection that's likely to fail the same way the first one did.