PUR adhesive for glulam and engineered timber

PUR wood adhesive is chosen for two reasons: it is solvent-free, and it bonds wood to a wider set of materials than a water-based glue. Both advantages come from the same chemistry, and so does its main weakness.

The reaction is with water

One-component PUR cures by reacting with moisture — from the wood, from the air, or from a damp sponge on the joint. That is why the adhesive needs no mixing, and why the process variable that matters most is not time but moisture content.

KA505 is 100% solids with a viscosity of 5,000–10,000 mPa·s at 25 °C. No solvent means no shrinkage gap in the glue line, no VOC in the press hall, and a spread rate that is what you applied rather than what survived evaporation.

Open time, clamp time and cure time are three different numbers

Parameter	What it controls	Consequence if ignored
Open time (30–70 min, seasonal)	How long after spreading the joint can still be closed	Dry glue line, no bond
Clamp time (1–4 h)	How long pressure must be held	Spring-back and a foamed, weak line
Full cure	When the joint can be machined or loaded	Premature handling damage

The catalogue gives 1–4 hours cure. That range is not sloppiness: cold or damp timber pushes to the top of it, dry warm timber to the bottom.

Moisture: the window is narrower than most workshops assume

Structural glulam is normally bonded at 8–12% moisture content. Outside that:

• Wet timber — the reaction produces CO₂ faster than the glue line can release it, so the bond foams and weakens. It looks cured and fails under load.
• Frozen timber — the water in the cells is unavailable and the cure stalls.
• Very dry timber — the reaction has too little water to proceed at rate, and clamp times stretch.

This is why “the glue did not cure” is usually a timber-condition report, not a product complaint.

Where PUR beats the alternatives, and where it does not

Adhesive	Bonds	Water resistance	Gap filling	Typical issue
PUR (one-component)	Wood, metal, plastic, insulation	Good	Poor	Moisture sensitivity
PVAc (white glue)	Wood only	Poor	Poor	Creep under sustained load
MUF	Wood	Good	Moderate	Formaldehyde classification
EPI	Wood	Moderate	Moderate	Two-component handling

PUR’s advantage is the mixed-material bond: a glulam beam with steel connectors, a sandwich panel with insulation, a truck body with aluminium skin. Its disadvantage is that it is not a gap filler — the joint has to close and the pressure has to be even.

Failure modes worth knowing

1. Starved joint — too little spread, or open time exceeded; the line looks thin and fails in shear.
2. Foamed glue line — wet or frozen timber; visible as a thick, crumbly line.
3. Spring-back — clamp released before cure; the joint opens a fraction of a millimetre and takes water.
4. Contaminated surface — planer glaze, dust or extractives on dense species; abrade before bonding.

When this is not the right choice

• The joint is a gap. PUR will not fill it; use a gap-filling adhesive or re-machine the joint.
• The assembly is structural steel. A timber adhesive is not a steel adhesive; use a structural product designed for metal.
• The environment is continuously wet. Water-resistant is not waterproof; exposed timber needs its own protection.

A press schedule worked through

Numbers make the scheduling problem concrete. Take a glulam beam bonded at 18 °C with timber at 10% moisture content, 45 mm lamellae.

open time at 18 °C      ~45 min (within the 30–70 min seasonal range)
clamp time to plan      2–3 h  (catalogue range 1–4 h, mid-temperature)
release check           break a sample joint: cohesive wood failure

Shift the same job to 10 °C and every number moves: open time lengthens toward the top of the range, cure slows, and clamp time goes to 3–4 hours. The adhesive has not changed; the reaction rate has.

Where scheduling fails

• Planning to the bottom of the range in winter. The press is released while the glue line is still green.
• Using air temperature instead of timber temperature. A cold board pulled from an unheated store behaves like a cold day.
• Skipping the release test. The clock is a proxy; a broken joint is evidence.

A press schedule that includes temperature and a release test costs a few minutes per batch and prevents the failure mode that is expensive to detect later: a joint that opens in service.

100% solids changes the arithmetic of a glue line

A water-based adhesive is bought as a wet weight and paid for as a dry polymer. At 50% solids, 400 g/m² of wet glue leaves 200 g/m² in the joint. A 100% solids PUR leaves exactly what you spread.

PVAc, 50% solids        400 g/m² wet  → 200 g/m² dry
PUR, 100% solids        200 g/m² wet  → 200 g/m² dry
water to be removed     ~200 g/m²     → 0

The difference is not just accounting. Every gram of water that leaves a PVAc glue line has to go somewhere, and it goes into the wood, swelling the cells and then shrinking them as it dries. That shrinkage is one reason a water-based line can show a fine crack along the joint in a dry room, and why thick assemblies bonded with PVAc are prone to creep under sustained load.

With PUR there is nothing to evaporate. The glue line stays where it was spread, the adjacent wood does not swell, and the bond is not competing with the substrate for moisture. That matters most in thick sections and in species that move a lot.

Spread rate, glue line and pressure

Three numbers decide whether a PUR joint holds, and they are set at the press, not by the adhesive.

Parameter	Working range	What happens outside it
Spread rate	150–250 g/m² per glue line	Starved joint below 120; squeeze-out and waste above 300
Cured glue line	0.1–0.3 mm	Thicker lines foam and lose strength
Press pressure	0.6–1.0 N/mm² for softwood glulam	Too low: gaps and spring-back; too high: crush and starve
Open time	30–70 min, temperature dependent	Exceeded: the film skins and never wets the second face
Clamp time	1–4 h	Released early: the joint creeps open

Two practical notes:

• Pressure is a function of species density. A dense hardwood needs more pressure to bring the faces into contact, but too much crushes the cells and starves the line. The press gauge is a starting point; the check is a broken joint showing wood failure rather than a clean glue line.
• Glue line thickness is set by machining tolerance, not by spread rate alone. A joint with a 0.5 mm gap will not close at any sensible pressure; adding adhesive only makes a thicker, weaker line.

Species and extractives

PUR bonds most species, but some woods are actively hostile to adhesion:

Species group	Issue	Response
Oak, chestnut	High tannin content	Abrade freshly; test a coupon
Teak, iroko, rosewood	Oily extractives	Wipe with solvent, abrade, bond soon after
Dense tropical hardwoods	Poor penetration	Roughen the surface; increase spread slightly
Douglas fir, larch	Resin pockets	Remove resin; avoid bonding across a pocket
Spruce, pine (glulam)	None significant	Standard preparation is enough

The extractive problem is time-dependent: a freshly planed surface bonds; the same surface left overnight may not, because extractives migrate to the surface. “Bond within a few hours of machining” is a process rule that costs nothing and removes a whole failure class.

Measuring moisture honestly

The moisture content that matters is the one at the glue line, 10–20 mm below the surface, not the one at the face.

• Pin meters read the surface layer; in a thick lamella that can be 2–3% away from the core.
• Capacitance meters read to a depth that depends on the electrode spacing; check the manual rather than assuming.
• Oven-dry testing is the reference method, and a small sample cut from a representative board is worth more than ten surface readings.

Equilibrium moisture content depends on the service environment: a heated building in winter can sit at 6–8%, an unheated workshop at 12–15%. Bonding at the wrong end of that range means the joint spends its life either shrinking away from the glue line or swelling against it.

Standards your customer may ask about

Two standards sit behind most load-bearing timber specifications:

• EN 301 classifies adhesives for load-bearing timber structures into types by their durability under exposure. If a project specifies a type, the adhesive has to be classified to it — a brand name is not a classification.
• EN 14080 governs glued laminated timber production, including bonding, and refers back to the adhesive classification.

If your market requires either, ask the supplier for the classification document and the test report behind it, not a certificate of company registration. A factory can be well run and its adhesive still not classified for the exposure class your project requires.

Failure diagnosis, by what the broken joint looks like

Appearance after breaking	Most likely cause	What to change
Clean wood failure across most of the area	Good bond — the wood gave way first	Nothing
Smooth, glossy glue line	Starved joint or contaminated surface	Spread rate, machining, wiping
Thick, crumbly, foamed line	Wet or frozen timber	Dry to 8–12%, condition above 15 °C
Wood failure at the edges, glue failure in the middle	Uneven pressure or a crowned face	Press platens, machining tolerance
Joint open at one end after clamping	Spring-back, released too early	Clamp time, pressure, check with a sample
Dark stain around the line	Water in the line or in the wood	Moisture control, adhesive storage

The pattern is the diagnosis. A supplier who asks for a photograph of the broken joint before answering is a supplier who has seen the failure before.

Running the press in a cold workshop

Below about 15 °C the reaction slows enough that a schedule written for summer produces a joint that is still green when the clamps come off. The fixes are mechanical, not chemical:

• Condition the timber and the adhesive in the same space for at least 24 hours.
• Heat the press area, not just the office.
• Extend clamp time toward the top of the range — 3–4 hours rather than 1–2.
• Break a sample joint at release; if it shows glue-line failure, the batch waits.

A press schedule is a table of temperatures and times, and it should be written down and posted at the press. Schedules that live in an operator’s head are the ones that fail when that operator is on holiday.

The practical position

Buyers should treat moisture content as a process control point, not a supplier specification. Ask for the timber’s measured MC at bonding, hold clamp time at the top of the range in cold weather, and test a glue line by breaking a sample joint rather than by trusting the press schedule. The adhesive is the cheap part of a failed glulam beam.

Standards referenced

• EN 301:2023 — classification of phenolic and aminoplastic adhesives for load-bearing timber structures. If a project specifies an adhesive type, this is the classification it refers to: CEN listing.
• EN 14080:2013 — glued laminated timber requirements, including bonding: CEN listing.
• EN 717-1 — formaldehyde release from wood-based panels, the test behind E1/E0 claims: CEN listing.

PUR is not classified under EN 301 — that standard covers phenolic and aminoplastic adhesives. A glulam specification that requires EN 301 is asking for MUF or PRF, not PUR, and that distinction belongs in the enquiry before the quotation.