Matching New Lumber to Wood Already in the Building

Start With a Control Sample, Not a Species Name

The match is decided by a physical piece of the wood already in the building, not by the species written on a drawing set from twenty years ago. If you are the construction manager on an occupied renovation where one wing of trim stays and the other gets replaced, the first thing that should leave the site is a cut-off of what is installed. Pull it from a removal, not from the back of a closet that never saw daylight. Keep it at least 12 inches long with the full profile intact.

Cut one end square and clean so the end grain is readable. Do not sand the face. Do not strip it, and do not wipe it with solvent to see what the color used to be. The aged, finished, light-exposed face is what the new material has to live beside, and a sanded sample throws away the one thing you could not get any other way. If the run was painted, take a second piece and strip a short section at the end so the profile can be measured without decades of coating buildup sitting on it.

That control sample is the document the order runs on. It goes out with the request for quote, it comes back attached to the submittal, and a duplicate stays in your trailer so the finish carpenter and the supplier are arguing about the same object. Photographs are useless for this. A phone camera adjusts white balance on its own, and two shots of the same board taken an hour apart will not agree.

Four things can be wrong at once between the old wood and the new. The species identification, the sawing pattern, the color, and the moisture content. Two of those close on their own and two never do. Sorting out which pair you are looking at is what decides whether you order now and wait, order and pre-weather, or stop and change the approach entirely.

The Species on the Drawings Is Usually a Guess

Species identification on existing work is wrong often enough that it should stay an open question until someone reads end grain under magnification. Finish, age and dirt flatten the differences that make identification easy on raw lumber. The reason this matters more than it sounds is that a wrong species name sends the whole order down a path no amount of waiting corrects.

Red Oak and White Oak are the pair that gets missed most. Hardness will not separate them, since Red Oak runs roughly 1,290 lbf on the Janka scale and White Oak roughly 1,360 lbf, close enough that a thumbnail proves nothing. The pores do. White Oak carries tyloses that plug the earlywood vessels, so a clean cross section cut on the control sample settles it in about ten seconds with a hand lens. Published anatomical descriptions are worth having open while you look.

The second common miss is Sapele read as Genuine Mahogany. Under a dark finish the ribbon figure on quartered Sapele looks like what most people have been told mahogany is. The two are not interchangeable in hand, with Sapele near 1,410 lbf against roughly 800 lbf for Genuine Mahogany, and they do not take stain the same way. Order one believing it is the other and the color work at the end turns into a chase.

Painted interiors hide a third case. A good deal of older painted trim that everyone calls pine is Yellow Poplar, which runs near 540 lbf and machines to a crisper arris than pine does. On stained work, Black Walnut at roughly 1,010 lbf has been finished to pass for Cherry for a century. The Forest Products Laboratory publishes the anatomical keys professionals use for this.

Here is the boundary no supplier can move. A finished board sitting in a wall cannot be identified with certainty from a photograph or a description, and short of a clean cross section read by someone trained in wood anatomy, every identification is an inference carrying a real chance of being wrong. J. Gibson McIlvain pulls stock against the sample in hand rather than against the species name typed on a transmittal, which is the only defense against an identification that was wrong before anyone was asked.

Sawing Pattern Is the Mismatch That Never Closes

Cut orientation is the one source of mismatch that no amount of waiting repairs, and it is the one most often discovered after delivery. Get the species right and the sawing wrong and the new work reads as a different room from across the hall, in exactly the right color.

Read it off the end grain of the control sample. Growth rings meeting the face at 60 to 90 degrees is quartersawn, 30 to 60 degrees is rift, and anything under 30 degrees is plainsawn. The visual difference is loud. Plainsawn White Oak throws cathedral figure. Quartersawn White Oak throws ray fleck, those broad shimmering flecks that no stain will imitate on a plainsawn board. Rift gives a straight, quiet grain with the fleck suppressed on purpose.

Older interiors, particularly anything built before the Second World War, are quartersawn far more often than people expect, and that creates a supply problem rather than a technical one. Quartersawing yields fewer usable boards per log and gives up width to do it. A 7 inch quartersawn board with consistent fleck is a different request from a 7 inch plainsawn board, and the two will not be filled out of the same pile on the same afternoon.

There is a side benefit worth knowing while you are writing the order. Quartersawn and rift stock moves less across its width than plainsawn, which is why it keeps turning up on painted panel work and on anything that has to stay flat through a heating season. If the existing material is quartersawn, matching it is not only a visual decision.

Color Is on a Clock That Varies by Species

The color gap you see on delivery day is the mismatch that moves on its own, and how fast it moves depends entirely on which species you are holding. Light does most of the work, so the clock runs on exposure rather than on calendar days.

Cherry is the sprinter. It darkens visibly in weeks of indirect daylight and lands close to installed material inside a few months. Sapele and Genuine Mahogany deepen toward red brown over a similar stretch. Black Walnut goes the other direction, lightening and mellowing rather than darkening, which catches people who assume every species gets darker with age. White Oak and Hard Maple move slowly and amber over years, not months.

Exposure is what you are actually scheduling against. A south-facing corridor closes the gap in a fraction of the time a north-facing stair hall takes, and a run of base behind a casework bank may never catch up to the run across the room. Pre-weathering is the practical answer where the species moves fast and the schedule allows it. Lay the new stock face up in daylight on site for a few weeks before the carpenters touch it, flipping nothing, and let the delivery do part of its aging before it goes on the wall.

One honest limit belongs here. No supplier can promise that new material will look like the installed material on the day it arrives, and no one can warranty a color match, because solid wood is an organic material that keeps responding to its environment long after it leaves the mill. Anyone who tells a construction manager otherwise is describing something other than solid wood.

The finish layer is its own variable. You are rarely matching raw wood to raw wood. You are matching new wood plus new coating against old wood plus a coating that has yellowed and ambered for decades, so the sample that settles the question is finished with the system the job will actually use and viewed in the room under the room's light. Where the work is painted, the color question drops out and the surface takes its place. Open-grained and oily species absorb a great deal of finish, so a primed and lightly buffed face is what keeps a new run from reading as coarser than the old one beside it. Preservation guidance on repairing historic woodwork is blunt about this, and it is worth reading even on a building with no historic designation.

Moisture Closes Fastest and Gets Blamed the Most

A moisture difference between new stock and installed wood equalizes in a matter of weeks, and almost every joint that opens after install gets blamed on the supplier when the cause was the building. This is the easiest of the four to manage and the one most often skipped.

Interior millwork is milled to an in-service moisture content near 6 to 8 percent, and J. Gibson McIlvain mills and ships interior stock to that range. What the number does not account for is the building you are putting it in. A structure closed up and conditioned for two years sits where it sits. A renovation running on temporary heat with a wet slab somewhere below does not.

Keep a moisture record, dated, in the same binder as the control sample. Take pin meter readings on the existing installed trim in three or four locations, take readings on the new material when it arrives, and take them again before anyone cuts. If the two columns sit more than a couple of points apart, the material is not ready and the calendar is the fix. If permanent HVAC is not running, every reading you take describes a building that stops existing the day it starts.

Acclimation has to happen in the conditioned space, stickered so air moves on all faces, not stacked solid in a garage or a corridor still open to the weather. Moisture behavior in wood assemblies is well documented, and standard test methods exist for verifying meter readings when a dispute is live.

Reading the Four Mismatches Against the Calendar

Two of the four close on their own, and which group a problem falls into decides what you do at order time rather than what you do later. The table below is the version worth taping inside the binder cover.

Run the four in that order and the decision usually makes itself. If both of the unresolvable items check out, order the full quantity now and let color and moisture close while the rest of the job catches up. If the sawing pattern or the width is not available, the honest move is to change the approach while it is still cheap, by refinishing an entire run instead of patching it, or by relocating the existing material into the room that shows and putting new stock where it does not.

Matching a Profile Is a Separate Problem From Matching the Wood

The wood match and the profile match are two different orders, and the sample that answers one rarely answers the other. A perfect species and orientation match in the wrong section still reads as a repair from across the room.

Paint buildup is what defeats most profile measurements. Six coats fill a cove and round a bead until the contour gauge reports a profile that never existed. Strip a short section at the end of the sample, or take the section from a cut end where the layers show in cross section. Send the physical piece rather than a tracing when you can, since a tracing loses depth on undercuts every time.

Some profiles in a building are standard patterns that have been in production for a century, and the published moulding pattern numbers will name them. Most older work in a building of any quality is not standard, and the only path is a knife ground to the traced section. J. Gibson McIlvain grinds knives to a submitted section and keeps thousands of ground profiles on file, which is what lets the second phase of a job match the first without anyone re-measuring a wall.

Where the match involves more than linear moulding, that is, a stair part, a cased opening or fitted casework built to the room rather than assembled from modules, the work moves to ICD, the architectural millwork division of J. Gibson McIlvain. Profile work handled there starts from the same submitted section.

Widths and Lengths the Original Job Had

Old work was frequently cut from boards wider, longer and tighter ringed than anything a routine order fills today, and that is a sourcing fact rather than a technique. It is also the quiet reason a match that looked easy on paper falls apart during submittal review.

A 1920s baseboard running 9 1/4 inches in a single piece is ordinary in that building and a real request now. So is a 16 ft run of casing with no joint in it. Neither is impossible. Both are questions to ask before the shop drawings commit to them, because the answer decides whether the detail survives.

Grade enters here in exactly one way. Hardwood grades under the National Hardwood Lumber Association rules describe the yield of clear cuttings a board gives, not how good the wood is, so a grade that yields plenty of short clear pieces is the wrong call for long uninterrupted runs even though the material itself is perfectly sound. Specify against the run length you need. J. Gibson McIlvain holds over seven million board feet of domestic and imported lumber at its Maryland yard, which is what makes it realistic to pull eleven boards that look like each other instead of eleven boards that merely meet a grade.

On exterior patching, the grade call outranks the species call. Cedar is a legitimate default on trim and siding and always has been, but a patch into clear vertical grain material, CVG, will not disappear if it is filled with select tight knot, STK, and the reverse is just as visible. Confirm which one is on the building before ordering, and check the published grade descriptions if the original submittal is long gone.

Staging One Lot Against a Phased Schedule

On a phased renovation the match has to survive the gap between the first floor turned over and the last, which means the whole quantity comes from one purchase even though it does not ship at once. Lot to lot variation inside a single species is real, and it is the failure that shows up at month nine.

Buy the full take-off plus attic stock in one order against one set of samples, then release it on a schedule. J. Gibson McIlvain stores finished material for customers, so a quantity bought for four floors can be drawn down floor by floor instead of sitting in a job trailer collecting damage and taking on whatever the building is doing that month.

"I ask for the sample before I ask for the quantity. A board that matches a name on a purchase order is easy to find. A board that matches the piece in your hand is the one we have to go look for, and on a phased job we have to find all of it at the same time or not at all."

Norm Moton, Director of Sales, J. Gibson McIlvain

Attic stock has a wrinkle nobody plans for. Material kept in a dark storage room does not age with the installed run, so the repair made in year three arrives lighter than everything around it. Store the attic stock in the building where it gets the same light the installation gets, or write into the closeout documents that it will need to sit out before it is used.

Field verification of what is in the wall, the fitting and the scribing stay with your carpenters, and the delivery ticket is where the two halves meet. Check it against the submitted sample on arrival, not after the crew has opened six bundles, since a mismatch caught on the truck is a conversation and a mismatch caught after cutting is a claim.

How J. Gibson McIlvain Would Specify This

A matching order that actually works names six things, and the control sample is the first of them. Everything else on the list exists to keep that sample from being the only thing the shop has to go on.

Send a physical control sample, 12 inches or longer, with one clean crosscut end and the aged face untouched. State the sawing pattern required rather than assuming it, using the growth ring angle read off that end. State the in-service moisture content the material is being milled to and the building condition it will be delivered into. State whether the match will be judged raw or finished, and if finished, send the finish schedule with the order.

State the full quantity for every phase, attic stock included, in one order with a release schedule attached. And state plainly what is not being asked for, so nobody quotes a profile you are not buying. Milling services at J. Gibson McIlvain cover the lumber side of this work, with ICD handling the custom architectural pieces, and both start from the same submitted sample.

J. Gibson McIlvain ships nationwide, including regular runs to California, and it mills to the condition stated on the order rather than to a house default. Call 800-638-9100 with the sample in your hand and the ring angle already read, and the first conversation gets to the real question instead of circling a species name on an old drawing. Specification language for architectural woodwork is covered in depth by the Architectural Woodwork Institute standards if you want the long form.