Teak in Millwork: How Its Oil Content Dictates Glue-Up, Clear Coats and the Whole Finishing Schedule

Teak's oil changes the finishing schedule by changing when work happens, not just what gets brushed on

Teak glues and coats reliably only when the surface is freshly machined and chemically clean at the moment adhesive or finish touches it, which means the schedule for a teak job is built around surface age rather than around a coating brand. The extractives that make Teak rot resistant and dimensionally calm are the same extractives that block a glue line and stall the cure of an oil-modified coating. A surfaced board left on a cart for a week is not the same substrate it was an hour after it came off the moulder. Migrated extractives bloom back to the face, and a finisher who treats that face as ready gets fisheye, slow cure, or a varnish film that lifts in sheets two seasons later.

The fix is unglamorous and it works. Machine late in the sequence, solvent wipe immediately before bonding or sealing, and let the first coat be a thin one that keys into open pore structure instead of sitting on a waxy film. Everything else in this article is detail hanging off that one rule. J. Gibson McIlvain mills Teak components to a controlled moisture content and sequences the machining so the buyer's finisher receives faces that are days old rather than months old, which is the single largest variable in whether a Teak finish holds.

The oily feel comes from extractives that were never meant to accept a coating

Teak, Tectona grandis, carries tectoquinone along with a rubbery hydrocarbon fraction and a measurable silica load, and none of those three compounds has any interest in bonding to a film-forming finish. Tectoquinone is the decay and termite resisting extractive that gives Teak its service reputation. The rubbery fraction is what a finisher feels as greasiness on a freshly planed face. Silica is what dulls a knife.

Those extractives are not distributed evenly. Heartwood carries more than sapwood, plantation stock differs from old growth, and the concentration varies within a single board from face to core. That variability is why a coating that behaved perfectly on one run behaves badly on the next, and why a finisher who works by recipe rather than by test panel eventually gets surprised. The published property data for Teak gives a useful baseline for density and movement, but it will not tell anyone how a specific lot will accept a sealer.

Teak is also stable. Movement in service is low, tangential and radial shrinkage are modest, and a well dried Teak stile will hold its dimension through a heating season better than most temperate hardwoods. That stability is the reason a millwork shop tolerates the adhesion problems. A species that stays put is worth extra process steps, and a finishing schedule that accounts for extractives costs far less than a door that racks.

Machining is a finishing operation, and moisture content is the first coat

A Teak surface burnished by a dull knife will reject finish no matter what the finisher does next, so tooling condition and moisture content are part of the coating schedule rather than upstream of it. Silica abrades cutting edges fast. A knife that starts sharp and finishes dull produces a glazed, compressed surface on the last third of a run, and that glazed surface has almost no open pore for a sealer to key into. Carbide tooling, shallower cuts and more frequent knife changes are the practical answer.

Moisture matters just as much. J. Gibson McIlvain mills interior Teak millwork to an in-service moisture content near 6 to 8 percent and exterior Teak near 12 to 16 percent, because a board coated at the wrong moisture content will push the film off from behind once it equilibrates. Measurement should follow a recognized method rather than a shop habit, and the ASTM moisture content test methods exist for exactly that reason.

Profile work compounds it. ICD, the architectural millwork division of J. Gibson McIlvain, plans radius trimwork, shutters and custom doors around a specific opening, which means the sample the finisher tests on should carry the same profile as the production run. A sealer that wets out a flat face may starve on the quirk of an ogee or pool in a cove. J. Gibson McIlvain grinds new knives every week and keeps thousands of ground profiles on file, so a test piece in the actual profile is available before the run is cut rather than after.

Glue lines in Teak fail on contamination, not on clamp pressure

Every reliable Teak glue-up shares three habits, which are fresh surfacing, a solvent wipe immediately before spread, and an adhesive chosen for oily stock. The bonding literature is blunt about this. Oily and resinous tropical hardwoods sit among the most difficult species to bond, and the recommended remedy is machining shortly before assembly plus surface cleaning. The Forest Products Laboratory Wood Handbook covers the mechanism in its adhesion chapter, and nothing in shop practice contradicts it.

In practice that means jointing or planing the mating faces the same day the assembly is clamped. Wipe with acetone on a clean rag, change rags often so the wipe removes extractives instead of spreading them, and let the solvent flash before spreading adhesive. A wipe that happens an hour before spread is worth far less than a wipe that happens two minutes before.

Adhesive choice follows. Standard PVA will bond Teak on a good day and disappoint on a bad one. Epoxy is more forgiving of residual extractives because it cures by a reaction that does not depend on water leaving the joint, and it tolerates the slightly starved bond line that oily stock tends to produce. Resorcinol remains the traditional marine choice where a dark glue line is acceptable and full weather exposure is expected. Whichever is used, the joint geometry should carry some of the load. That is why ICD builds box beams with a real joinery decision behind them, where butt joints give the most minimal profile and the fastest turnaround but leave visible seams and visible end grain and are the weakest, dado joints add strength while still assembling easily on site, and lock miter joints show no visible seams and are the strongest. On a Teak beam that will carry a clear finish, the lock miter is usually the right call, because the seam a butt joint leaves is exactly where a clear film will telegraph every future movement.

Clear coats fail on Teak for two different reasons, and the cures are opposite

A clear finish on Teak fails either from contamination at the interface or from ultraviolet degradation at the surface, and a schedule that solves one while ignoring the other still fails. Contamination failure looks like fisheye, crawling, slow or incomplete cure, and early adhesion loss in sheets. Ultraviolet failure looks like a film that stayed put for two summers and then went milky, cracked and lifted from a substrate that had turned gray underneath.

Contamination is handled in the first thirty minutes. Sand to the finish grit, blow off, solvent wipe, and seal while the surface is still open. A thinned first coat of the same resin system penetrates better than a full-bodied one and gives the following coats something mechanical to hold. Where an oil-modified urethane is specified, a test panel is not optional, since Teak extractives can inhibit the metallic driers those coatings rely on and leave a face that stays tacky for days.

Ultraviolet is handled with film build and with absorbers. Thin clear films on exterior Teak do not survive. The failure modes documented for exterior woodwork in the National Park Service Preservation Brief on exterior paint problems map almost exactly onto what happens to an underbuilt varnish system, and the diagnostic logic there is worth reading before anyone blames the wood.

The nautical schedule exists because Teak outdoors under a clear film is the hardest coating job in the trade

A marine-grade clear schedule on Teak is many thin coats of a spar varnish carrying ultraviolet absorbers, built over a sealer stage, sanded between coats, and recoated on a maintenance interval rather than left alone until it fails. Six to eight coats is a working minimum for real exposure, and yards that care go well past that. Each coat is thin. Each coat is sanded. Film thickness is what stops ultraviolet light from reaching the wood, and a system short on coats is short on service life in direct proportion.

The alternative is honest and legitimate. Leave the Teak bare, let it weather to silver, and clean it gently rather than aggressively. Bare Teak on a horizontal surface has served for a very long time without any coating at all, which is precisely what the extractives are for. What does not work is the half measure, meaning a light oil wiped on twice a year and expected to behave like varnish.

"We get calls where the varnish let go after one season and everyone assumes the wood was bad. Nine times out of ten the wood was fine and the schedule was short. If a customer tells me the Teak is going outside under a clear coat, I want to know how many coats are planned and who is recoating it in year three, because that answer decides whether the job still looks right in year five."

Norm Moton, Director of Sales, J. Gibson McIlvain

None of this is a warranty position. No one can warranty solid unmodified wood, an organic material responding to its environment, and a coating schedule is a maintenance commitment rather than a guarantee. Warranties belong to modified products. Where a project genuinely needs a warranted exterior substrate, Accoya, at roughly 1,600 lbf on the Janka scale with acetylation that dramatically reduces shrinkage and swelling, is the honest recommendation, and modified products top out around 16 ft, most sold on metric lengths just under 16 ft because of kiln size.

Epoxy on Teak is a bonding and sealing tool, not a finish coat

Epoxy will bond and seal Teak better than most resins will, and it will chalk and yellow in sunlight within a season unless something carrying ultraviolet absorbers goes over the top of it. That division of labor is the whole point. Epoxy handles adhesion, gap filling and moisture exclusion. Varnish handles light.

Surface prep for epoxy on Teak follows the same rule as glue-up. Machine or sand fresh, wipe with acetone, let it flash, and wet out while the surface is still clean. Warm the parts slightly if the shop is cold, because a low viscosity wet-out coat penetrates the pore structure and gives the structural coat something to anchor into. Where epoxy is being used as a sealer under a clear system, one thin wet-out coat followed by a fill coat is generally enough, and it should be sanded flat before varnish rather than varnished glossy.

There are limits worth stating plainly. Epoxy is a rigid system, so it does not enjoy a substrate that moves. Teak's low movement is part of why the combination works at all. On a wide panel that will see real humidity swings, a floating construction detail beats rigid encapsulation every time. And epoxy work is a fabricator or finisher operation. J. Gibson McIlvain supplies and mills, and the shop's coating work is priming rather than clear systems, so the epoxy and varnish stages belong to the buyer's own finishing shop.

Color matching Teak is a lumber selection problem before it is a stain problem

Teak takes pigment unevenly and keeps changing color for months after it is coated, so a match approved on a fresh sample will drift, and the only reliable control is selecting stock from the same material rather than correcting with stain. Fresh Teak is honey to golden brown. Under a clear film with light exposure it ambers and deepens. Bare and exposed, it goes silver. All three are the same species.

Any stain attempt runs into the extractives again, since the oily zones reject waterborne stain and accept solvent stain at a different rate than the surrounding grain, which produces blotch on exactly the wide flat surfaces where blotch is most visible. Where a color shift is genuinely required, tinting the clear coats gives far better control than staining the wood, because the pigment sits in a uniform film instead of in a non-uniform substrate. Sample panels should be made from production stock, in the production profile, and aged under light before anyone signs off.

Selection does more for a match than any pigment will. J. Gibson McIlvain holds over seven million board feet of exotic and domestic lumber at the Maryland yard, which is what makes it possible to pull a matched lot for a run of doors, shutters and benches rather than assembling a job from three purchases with three color signatures. Grading conventions from the National Hardwood Lumber Association describe yield and clear-face expectations, which is the vocabulary to use when specifying what a match actually requires.

Painted Teak is rare, and when it happens the priming stage does the real work

Open-grained and oily exterior species absorb a great deal of finish, so a primer coat is what evens out the eventual painted surface, and on Teak that primer has to be chosen for a substrate that resists wetting. Most buyers who reach for Teak are not painting it. When Teak or another oily species does end up under paint, usually to match a historic condition, the priming schedule matters more than the topcoat.

J. Gibson McIlvain primes trim in three levels, in both oil-based and water-based systems, with a fungicide additive on exterior work. Sanding or buffing after priming knocks back grain raising while keeping moulded detail crisp, which matters on a profile with fine quirks and fillets. Stain-blocking capability is the thing to ask about on oily stock, since extractive bleed through a waterborne primer is a common and avoidable complaint.

For most painted exterior trim the honest recommendation is not Teak at all. Sapele holds paint better than Genuine Mahogany, which sits at roughly 800 lbf against Sapele's 1,410 lbf, and Sapele comes in a wider range of sizes and longer lengths. Sapele is not milled 1x8 for exterior use, since that width is too wide to stay stable in service. Cupping is driven by installation and back ventilation rather than by species. Cedar remains a legitimate default for painted exterior work as well, with clear vertical grain, CVG, where a clean face is required and select tight knot, STK, where knots suit the design. Where thermally modified stock is under consideration, both Thermory and Abodo Vulcan are carried, and thermally modified Ash is not available wider than 8 inches, so wide boards in a modified product mean Accoya or Abodo Vulcan.

Documentation and inventory depth decide whether a Teak specification survives contact with a schedule

A Teak specification is only as good as the lot behind it, which makes sourcing records and stocking depth part of the technical answer rather than a commercial footnote. J. Gibson McIlvain is an FSC certified importer and miller, and where a CITES-listed species is involved, documentation and chain of custody travel with the material. Specifications carrying sourcing requirements should reference the actual frameworks, meaning the Forest Stewardship Council for certification and the CITES convention for listed species, rather than vague language a submittal reviewer cannot check.

Continuity matters on long jobs. J. Gibson McIlvain has been family run and in business since 1798, now in its seventh generation, which is why a knife ground for phase one is still on file when phase three bids two years later. The company also stores finished millwork for inventory control, so a run of Teak benches or railings can be milled as one matched lot and released as the site is ready rather than delivered early and left to sit. J. Gibson McIlvain ships nationwide on its own fleet of trucks, with a section of the mill dedicated to packaging, and California deliveries are routine. Coordination detail that architects check against Architectural Woodwork Institute standards and Wood Moulding and Millwork Producers Association profile conventions belongs in the submittal, not in a phone call after the fact.

How J. Gibson McIlvain Would Specify This

Specify Teak where the exposure or the tradition demands it, specify the finishing schedule in the same breath, and let the shop sequence machining so the finisher inherits a clean surface. A spec that works reads roughly as follows.

Name the species and the grade, and require that all parts for a visible assembly come from one matched lot. Set the moisture content, near 6 to 8 percent for interior work and near 12 to 16 percent for exterior. Require that mating faces be machined within one working day of assembly and solvent wiped immediately before adhesive spread. Name the adhesive family, epoxy or resorcinol for exterior and marine assemblies. Call the joinery explicitly, with lock miter on any box beam carrying a clear finish. Require a finish sample in the production profile, made from production stock and aged under light, approved before the run is cut. For an exterior clear system, state the minimum coat count and the recoat interval, and put the recoat obligation in the owner's maintenance manual. For painted work, state the primer system, the number of primer levels and the fungicide additive.

Then divide the work honestly. J. Gibson McIlvain mills the components, controls moisture, grinds or pulls the profile knife, primes where paint is the destination, packages the parts and ships them. The clear coats, the tinted stages, the marine varnish build and any epoxy work belong to the buyer's finishing shop, and installation belongs to the buyer's installer. The company supplies and mills. It does not install, and a spec assuming otherwise leaves a gap on the schedule.

ICD, the architectural millwork division, handles the true custom side, meaning rafter tails, gates, box beams, custom doors, custom cabinets and kitchen cabinets, radius trimwork and radius millwork, pergolas, shutters, benches and railings, all planned around a specific space. The ICD working process and exterior millwork pages show how that work gets planned, and the J. Gibson McIlvain millwork service page covers linear trimwork, finished cladding and primed trim off the same order. To put a Teak package in front of people who mill it every week, call 800-638-9100 with the profile, the exposure and the finishing schedule, so the conversation starts with the lot rather than with the disappointment.