How Much a 12-Inch Wood Panel Moves: Garage vs. House, With Real Numbers
How Much a 12-Inch Wood Panel Moves: Garage vs. House, With Real Numbers - A step into the workshop.
Stop Wasting Wood
Generate clear cut diagrams instantly with our new Cutlist Optimizer. Plan your cuts to make your next build easier.
A friend of mine glued up a beautiful 12-inch-wide red oak panel in his garage shop in March. Perfect fit, flush seams, straight edges. By the time football season started, he'd brought it into the house as a console top and one corner had cracked a breadboard end clean off. He blamed the glue. It wasn't the glue. It was moisture, and it was predictable the whole time.
Here's the thing that gets people in trouble: wood in an unconditioned garage and wood in a climate-controlled house are living at completely different moisture contents, and if you machine and glue up in one and deliver to the other, the panel will move. We covered the moisture side of this in our earlier article on equilibrium moisture content, where our numbers put garage-stored wood around 11–12% MC versus 6–8% indoors in typical conditions. Those figures came from our own EMC writeup and I haven't found an independent garage-specific dataset to confirm them, so treat them as our shop numbers, not gospel. But the physics behind them is well documented, and the USDA Forest Products Laboratory's Wood Handbook backs the direction: at 70°F and 65% RH, EMC runs about 12%; at 70°F and 35% RH, it's about 7% (SensoraHome, citing the Wood Handbook). A garage tracks the outside humidity. A house does not.
The Grit: How Much Does a 12-Inch Panel Actually Move?
The standard shop formula, per Popular Woodworking, is:
Movement = board width × species coefficient × annual MC change
It's only valid for wood between roughly 6% and 14% MC, which covers furniture service. The species coefficient is the key. Per the USDA Wood Handbook (GTR-282, Table 4-3, as tabulated by SensoraHome), the flat-sawn tangential coefficients per 1% MC change are:
| Species | Tangential coefficient |
|---|---|
| Red oak | 0.00369 |
| Hard maple | 0.00353 |
| Black walnut | 0.00274 |
| Black cherry | 0.00190 |
(Note: Flowyline lists walnut at 0.00279, a small discrepancy between sources. When values conflict, I go with the USDA figure.)
Run the numbers for a 12-inch flat-sawn board using SensoraHome's movement-per-percent figures (~0.044" per foot for red oak, 0.042" hard maple, 0.033" walnut, 0.023" cherry):
- Red oak, 12″ wide, 5-point MC swing (garage wood at 12% moving to a 7% indoor EMC): about 0.22″, a hair over 7/32″.
- Red oak, same board, 2-point swing in a stable house: about 0.09″, a smidge over 3/32″.
- Hard maple: ~0.21″ / 0.084″. Walnut: ~0.165″ / 0.066″. Cherry: ~0.115″ / 0.046″.
That 5-point swing scenario is exactly what happens when a panel is glued up at garage EMC and brought into a dry heated house. It doesn't have a choice. It shrinks. If you built it tight, something breaks.
Quarter-sawn stock helps: flat-sawn wood moves roughly 40–50% more across the grain than quarter-sawn (Flowyline; SensoraHome puts red oak's ratio near 1.6, so quartersawn moves about 38% less), which puts quartersawn red oak in the same scenario closer to 0.14″. Still enough to matter. And before you say "I'll just leave it long and trim it," length is a non-issue: longitudinal movement is about 0.1%, so an 8-foot board moves maybe 3/32″ end to end (Flowyline). Across the grain is the whole fight.
Popular Woodworking's own worked example tells the same story at bench scale: a 24″-wide quartersawn yellow-pine top with a 3-point annual swing moves about 1/8″; flatsawn, about 3/16″. Small numbers on paper. Enough to split a glued-up top in real life.
Why the Garage Blows Up Your Glue-Ups
WOOD Magazine lays out three rules that ought to be carved into every bench top: wood moves, wood never stops moving, and if you constrain the movement, either the joint or the wood fails. A finish slows movement. Attaching the panel to a rigid substrate doesn't stop it either. Nothing stops it.
Now apply that to the garage-to-house handoff. A 12-inch-wide panel glued in cross-grain fashion far exceeds WOOD Magazine's rule of thumb that you can ignore cross-grain movement up to about 4 inches of width. Beyond that, you must build in room to move. And the failure isn't always immediate. Woodwork Science describes delayed cohesive failure: a joint assembled at high MC in a drier end-use environment carries residual cross-grain stress, and years later, when seasonal cycling exceeds the hardened glue's plasticity, the seam lets go. If you pull a failed joint apart and see torn fibers on both faces, that's cohesive failure from overstress, not a bad glue batch.
SensoraHome puts hard numbers on the stress: a 24″ flat-sawn red oak tabletop assembled at 11% MC facing a 4-point drop to 7% shrinks about 0.355″, more than 3/8″, which puts something like 18–22 lbs of lateral force across a glued seam. (That pound figure is one source's estimate with no published test method, so file it under "directionally right.") Either way, 3/8″ of shrinkage against a fixed breadboard end is not a negotiation. The panel wins or the joint does.
WOOD Magazine's chest example makes the case for dividing wide glue-ups: a single 29″-wide quartersawn white oak panel can change ~3/8″ in width, while three narrow panels side by side each move under 1/10″. Fewer inches per panel means less total fight.
Fix Your Shop Before You Fix the Joint
The whole problem starts before glue touches wood. Dr. Eugene Wengert's rule, quoted by Wagner Meters, is that wood should be dried to within 2 percentage points of the EMC of its end-use location. If your end use is a house at 6–8% MC (Wagner's acceptable range for interior furniture-grade work; AWI specifies 6–9% for premium casework) and your garage lumber has equilibrated at 11–12%, you're out of spec before you make the first cut.
Find your shop's actual number: log temperature and RH at bench height for three consecutive days with a calibrated hygrometer, then cross-reference the USDA Wood Handbook EMC tables (GTR-282, Table 4-2). That EMC is your assembly target (SensoraHome). And when checking a board itself, Wagner Meters recommends taking meter readings at both ends and the center, because end grain dries faster; if any reading is out of range, the whole board waits.
Build It So It's Allowed to Move
If the panel is wide, don't constrain it. WOOD Magazine's proven strategies:
- Floating frame-and-panel construction — no glue in the groove, panel loose side to side.
- Breadboard ends attached only at the center, with slotted peg holes outboard.
- Battens screwed, not glued, with slot-cut or counterbored screw holes.
- Figure-8 fasteners and Z-clips for solid tops over cases.
- Drawer bottoms glued only at the front; 1/16″ perimeter gaps on inset fronts.
- Finish panel edges before glue-up so shrinkage never exposes raw wood.
And when solid wood meets plywood, remember sheet goods don't move appreciably. WOOD Magazine's fix: round pilot holes up front, slots at the rear, so all the movement heads toward the back and your reveals stay put.
The Verdict
I won't pretend I ran a calendar-controlled test with a caliper on a shelf; the honest math above is built from USDA coefficients and published EMC tables. A 12-inch panel is going to move between 1/8″ and 1/4″ over a year depending on species and where it lives, and the garage-to-house move is the worst of it. Glue up at the EMC of the room the piece will live in, not the room you work in. Widen your tolerance, divide your panels, and give the wood somewhere to go. If you build a 12-inch panel tight and try to pin it in place, you've already lost.