Sika AcouBond: Cutting Impact Noise by Up to 20 dB Under Engineered Oak Boards

Impact noise ratings sit or fall on the substrate long before the SikaBond cartridge opens. The Sika AcouBond system combines a rolled acoustic mat, typically SikaLayer, with SikaBond T52 or a comparable elastic adhesive applied in triangular beads through a notched application shoe. Laboratory data cited by Sika puts the impact sound reduction at up to 20 dB against a bare concrete reference, measured to ISO 10140 procedures. The gap between that number and what a building acoustician measures in situ under EN ISO 16283-2 is where most of the arguments happen.

The headline figure assumes an engineered board of at least 14 mm, a flat and sound substrate, and a fully floating perimeter. Strip any one of those conditions and the delta collapses. A board glued rigidly to the mat but pinched at a skirting or a door threshold creates a flanking bridge that shorts the whole assembly. The membrane does its work by decoupling the board mass from the structural slab, and a single hard contact point defeats decoupling across a surprisingly large radius.

Why the substrate flatness tolerance decides the outcome

SikaLayer is thin, usually around 2 mm to 3 mm depending on the variant, and it does not level anything. It follows the substrate profile. That matters because the AcouBond adhesive beads need consistent bearing to transfer load without point-loading the mat. Sika specifies a substrate flatness broadly in line with the SR1 category under BS 8204, meaning no more than 3 mm deviation under a 2 m straightedge. Anhydrite screeds poured to a good finish often meet this. Sand cement screeds hand floated by a rushed gang rarely do without correction.

Where the screed is out of tolerance, the fix is a smoothing compound, and this is where trades collide. On calcium sulphate screeds the laitance must come off first, usually by mechanical sanding, before any Mapei anhydrite screed primer or Sika equivalent goes down. Skip the sanding and the primer sits on a weak surface film that later delaminates, taking the acoustic mat and the oak with it. The primer is not optional on anhydrite because the material is moisture sensitive and the adhesive chemistry reacts badly to residual surface dust.

Moisture is the second gate. Calcium sulphate screeds must reach around 0.5 percent CM moisture before bonding, tested with a carbide hygrometer, not a surface meter. Cementitious screeds sit near 2 percent CM. Bonding oak over a wet anhydrite base traps moisture against a material that swells, and the acoustic layer becomes irrelevant when the boards cup and lift.

The perimeter detail that eats the decibels

Every expansion gap around the room, and every penetration, has to stay free. The AcouBond method sheet calls for a compressible edge strip at all vertical abutments, kept proud of the finished floor and trimmed only after the skirting is fixed off the wall rather than onto the board. A skirting screwed through the oak into the wall reconnects the floating field to the structure and reintroduces the flanking path the membrane was installed to break.

Door thresholds are the worst offenders. A metal trim bar bedded in rigid adhesive across a doorway couples two rooms and, more damagingly, couples the board to the slab under the trim. The clean detail uses a floating threshold sitting on the same membrane, with the elastic adhesive carried through continuously and no mechanical fixing driven into the slab.

A worked comparison on a 30 square metre apartment floor

Take a first floor flat with a 65 mm calcium sulphate screed over 25 mm of mineral wool on a precast concrete plank. The baseline weighted impact sound level, L’nT,w, measured before the finish, comes in around 58 dB. Building regulations in many jurisdictions, including Approved Document E in England and Wales, require 62 dB or better for the separating floor, so the slab already passes, but only just, and the client wants engineered oak.

Glue the 15 mm oak directly to the screed with a standard hard adhesive and the impact level worsens by several dB because you have added a hard radiating surface. Install the same oak through AcouBond with SikaLayer and the on site improvement typically lands in the region of 8 dB to 15 dB, not the laboratory 20 dB, because the field measurement includes flanking through the walls that the lab rig excludes. That still moves L’nT,w comfortably clear of the limit. The lesson buried in the arithmetic is that the 20 dB claim is a component figure, and the number that gets signed off is always the room to room field figure.

Where AcouBond meets the tiling scope

On a mixed refurbishment the oak lounge often shares a slab with a tiled kitchen or bathroom, and the two systems interact at the doorway. A large format porcelain installation next door is heavy and rigid, and its own acoustic strategy usually runs through a separate decoupling layer. An uncoupling membrane for cracked tiles, such as a Schluter Ditra or Mapei Mapetex type mat, handles crack isolation but does very little for impact noise, so combining the two goals in one wet zone means stacking systems, not substituting one for the other.

Wet room tanking is the flashpoint. A wet room tanking kit terminates at the door zone precisely where the AcouBond floating field wants a clean decoupled edge. The tanking must be dressed up the wall and lapped correctly while the oak side stays free floating, which means the threshold detail carries both a waterproof upstand and an acoustic break within a few centimetres of each other. Diagonal tile setting out in the wet zone changes nothing acoustically but does shift where the perimeter cuts fall, and awkward perimeter slivers are exactly where installers are tempted to bed a tile hard against the threshold and bridge the gap.

Slip performance belongs to the tiled side, not the oak, and it is measured independently. A pendulum slip resistance test to BS 7976 gives a PTV that the specifier reads against the wet zone requirement, commonly a PTV of 36 or above for a barefoot wet area. That test result has no bearing on the AcouBond claim, but the two data sets end up in the same handover file, and confusing a slip PTV with an acoustic dB rating in the O and M documentation is a recurring source of client queries.

The adhesive open time nobody plans for

SikaBond elastic adhesives skin over. On a warm site the working window before the bead skins can drop under 20 minutes, and once skinned the board no longer keys into the adhesive, so acoustic contact is lost across those beads.

What the datasheet does not settle

The unresolved question on most jobs is whose measurement governs at completion. Sika publishes a component reduction, the building control body wants a field L’nT,w, and the acoustician tests a finished room that includes flanking paths outside the flooring contractor’s scope. When the field result lands between the two, and it usually does, the argument over whether the floor finish underperformed or the separating wall leaked sound is rarely something the flooring package alone can answer.