Technical note 2

TN-2: Adhesives and Reversibility

What glue does, whether the chair needs it, and if so which. First draft 19 August 2026, from KU's questions and the fastener thread. Companion to TN-4 (Fastening Systems Survey) and TN-7 (Why Joints Go Slack). Status: DRAFT; the adhesion mechanics are STANDARD, the hide-glue record is VERIFIED against the sources cited, and the recommendation is DERIVED from the project's repairability doctrine.

The claim

Glue transfers load across a joint. In a chair designed the Keel way, faces abutting in compression with a fastener that only positions, glue is at most a secondary, renewable aid against racking and working loose, never the primary load path, and a well-made compression joint may need none at all, which is itself a longevity virtue: nothing to creep, nothing to reverse. If glue is used on a chair meant to be maintained for a century, use a REVERSIBLE glue, hot hide glue, for the same reason the finish note reaches for a renewable finish and the fastener note for a repairable fastener: reversibility is repairability is longevity.

What glue actually does

Glue holds by adhesion to the wood and cohesion within itself, and its strength depends almost entirely on grain orientation [STANDARD]. On LONG-GRAIN to long-grain, a good bond is stronger than the wood around it; the joint breaks in the wood, not the glue. On END-GRAIN, the open tubes drink the glue and starve the bond, so it is weak and unreliable. Most chair joints present the weak case: the shoulder of a mortise and tenon, or a panel meeting a keel, is largely end- or cross-grain. So glue helps exactly where the chair needs it least and is least reliable where the loads concentrate.

Does glue help longevity? The evidence cuts both ways.

FOR: virtually every surviving antique wooden chair is hide-glued, and the glue does resist the joint working loose and racking [VERIFIED against the accounts of hot hide glue in antique furniture, 19 Aug 2026]. Glue demonstrably contributes to medium-term survival. AGAINST: a rigid, irreversible glue line fights cross-grain wood movement, and over centuries it creeps, fails, or cracks the wood. The survival of antiques is often DESPITE glue failure: the mechanical mortise and tenon holds while the glue has long since let go and been renewed. A glue that cannot be reversed also makes every repair destructive, because you must break the joint or the wood to open it.

The resolution is the one the whole project keeps reaching. If you glue, use a glue you can undo. Hot hide glue is the only adhesive with a multi-century furniture record; it is what holds the antiques; and conservators re-glue with it precisely because it sticks to itself and releases with heat and moisture, so a joint can be opened, cleaned, and re-glued without destroying the wood [VERIFIED against the wood-conservation adhesives literature, Williams, and the hide-glue accounts, 19 Aug 2026]. Its weaknesses, creep under sustained load and failure when damp, are tolerable precisely because it is renewable. The modern glues invert this: PVA is convenient and strong but creeps and is hard to reverse cleanly; epoxy and polyurethane are strong and gap-filling but effectively permanent, so a joint made with them can only be repaired by destroying it. Christopher Schwarz, in the Anarchist's Tool Chest and his hide-glue writing, argues the same, and is worth reading through the dogma: hide glue for reversibility, and skepticism of leaning on permanent adhesives in furniture meant to be repaired.

For the Keel Chair specifically

Because the Keel Chair is a compression truss with mechanical fasteners, glue is not the load path and should not be treated as one. Three positions, in order of preference:

  1. NO structural glue. If the compression joints are well-fit and the fasteners keep them closed, the chair stands without glue, and "no glue" means nothing to creep, nothing to yellow, nothing to reverse. This is the thousand-year chapter's own rule: no glue structural, because no glue but hide has a long record and every rigid glue line is an unrepairable joint.
  2. Hide glue as a renewable AID where a joint tends to rack or rattle, chosen because it can be renewed on the maintenance covenant like the finish.
  3. Never a permanent adhesive (PVA, epoxy, polyurethane) on any joint meant to be opened for repair across generations.

For the Standard

A glue clause: structural glue is not required and not recommended; the chair's strength comes from compression joints and fasteners, not adhesive. Where glue is used as a secondary aid, it must be reversible (hot hide glue). Permanent adhesives are excluded from any joint intended to be serviceable.

Open questions

  1. Does the Keel Chair rack or rattle without any glue once the fasteners are set? A build-and-shake test answers it; if it is quiet, glue is simply omitted.
  2. Hide glue on plywood: the panels are ply, and hide glue's behavior on the thin face veneer and the exposed edge grain of ply is not the same as on solid wood. Worth a small test.
  3. The damp-storage case again: hide glue fails with moisture, so a chair that might get wet argues either for no glue or for a different, and less reversible, adhesive, at the cost of repairability.

Revisions

19 Aug 2026: first draft, from KU's glue question. The central finding, that in a compression-fastened chair glue is a secondary renewable aid at most and often unnecessary, and that if used it should be reversible hide glue, dates from this pass.

Renumbered 19 Aug 2026: first written the same day as TN-14 (Glue), then folded into its reserved slot, TN-2 (Adhesives and Reversibility), to retire the duplicate. The content is unchanged.

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. Sources: the hot-hide-glue accounts, Donald Williams's "Adhesives for Wood Restoration," and Christopher Schwarz's Anarchist's Tool Chest. The register · the verification ledger · hundredyearchair.com