Technical note 1

TN-1: The Panel Problem

Flat panels for a thousand year chair. First draft 17 Aug 2026, from the CLT discussion of that date. Status: DRAFT, two sources in the queue.

The claim

No wooden panel is a thousand year part; the thousand year move is to make the panel a replaceable module built to a published layup from stock that will still be sourceable in five hundred years, held by interfaces that never change, in a construction any competent joiner can remake and any owner can repair.

The problem, in numbers

Wood moves across the grain with moisture and barely at all along it. White oak shrinks about 10.5 percent tangentially and 5.6 percent radially from green to oven dry [STANDARD: FPL Wood Handbook values; pin the table number before print], which works out to roughly 0.37 percent of width per point of moisture content for a flat sawn board and about half that for quarter sawn [DERIVED]. A realistic indoor seasonal swing of 3 points of moisture content therefore moves a flat sawn 400 mm board about 4 mm in width, and a quarter sawn one about 2 mm [DERIVED, assumptions stated]. The standard's argument runs through hole stations held at true distance. A solid 400 board cannot hold a cross grain hole pattern: fix it and it splits, slot it and the datum is surrendered. This is the panel problem.

The candidate constructions, ranked for the millennium

  1. GAPPED BOARDS (the crate solution). Dissolves the problem rather than solving it: no glue, every board moves freely, every board is individually replaceable. Proven by centuries of ledged doors and by the 1934 sheet this project began from. The most millennium credible answer because there is no panel, only boards.

  2. SHOP CLT (the current proposal). Three face pieces of 6 x 140 white oak per side, grain vertical, cross layers between, 18 mm total, hide glue. Three 140s make 420, the 400 module plus trim, so the face layup restates the module. Cross lamination cancels seasonal movement almost entirely, so the panel carries plain round holes at published stations. Narrow stock stays sourceable when 400 wide old growth will not: that is the reproducibility argument. The honest cost: the movement is not deleted but internalized, loading the glue lines in shear and the cross plies in rolling shear every humidity cycle, and a thousand years is on the order of a thousand cycles. The glue line is the panel's clock. Rolling shear is the recognized weak mode of cross laminated timber [STANDARD: CLT engineering literature; pin a handbook citation before print]; thin cross plies reduce the demand, and 6 mm is already thin.

  3. TONGUE AND GROOVE. Honorable, glue free, movement absorbed in the joints, proven in floors and wainscot for centuries, but it is not a panel until a frame or battens make it one, and then the frame owns the problem.

  4. THE SINGLE 400 BOARD. Beautiful, doomed, and unreproducible. Survives only quarter sawn with slotted fixings, which forfeits the datum, and requires a log that gets rarer every decade.

Why hide glue, of all glues

Hide glue is the weakest structural adhesive available and the only one with a demonstrated multi century service record plus the one property no modern adhesive offers: reversibility. A hide glue delamination is a repair, executed with the same glue, by anyone, forever. A failed polyurethane or epoxy line is a dead panel. Two cautions attach: hide glue creeps under sustained stress and dislikes liquid water, so the panel lives indoors and wears a breathable renewable finish, oil and wax renewed each generation, never a film that cracks and traps moisture. TN-2 owns the full adhesive argument.

Best shot practices, in order of leverage

Symmetric balanced layup, same species and thickness both faces, odd ply count, or the panel cups by construction. Quarter or rift face stock, halving the stress each cycle fights. Thin cross plies. Straight knot free heartwood only: white oak's decay resistance lives in the heartwood [STANDARD]. Seal end grain. Show the layup at the edges rather than lipping it: the panel declares its construction to its future repairer. Ventilate both faces in service, no trapped water, no ground contact. No iron against oak: tannins corrode the fastener and stain the wood black [STANDARD]; the bronze and brass doctrine is compatibility here, not just aesthetics. And publish the panel drawing, because the deepest longevity is the Ise move: the Ise shrine buildings have been ritually rebuilt on a roughly twenty year cycle for over a millennium, and what survives is the information plus the interfaces [VERIFIED in general outline, common scholarly knowledge; pin one citation before print].

Test protocol (proposed, one afternoon)

Build one 18 mm panel to the layup with hot hide glue. Run accelerated humidity cycling alongside the printed frame test: the chair votes on the bracket and the panel in the same season. Log the result here.

Sources and verification queue

UNVERIFIED, do not repeat in print yet: the claim that Egyptian coffins used cross banded lamination millennia ago, often cited as the oldest plywood and directly on point as CLT's deep precedent. Needs a primary source, likely Lucas and Harris, Ancient Egyptian Materials and Industries, or a museum catalog record.

TO LOCATE: the conservation literature on cradled panel paintings, where rigid cross battens caused the very splits they were meant to prevent. The cautionary mirror image of cross lamination; worth one careful citation.

TO PIN: FPL Wood Handbook shrinkage table for the white oak figures; a CLT handbook citation for rolling shear as the governing mode.

Revisions

17 Aug 2026: first draft, from the panel discussion in the build sessions.

Technical Notes of the Hundred-Year Chair: public from first draft, refined iteratively, forever. A claim is VERIFIED, STANDARD, DERIVED, or UNVERIFIED, and says which. The register · the verification ledger · hundredyearchair.com