Technical note 16

TN-16: Comfort: Pressure, Contour, and Cushion

Why the seat and the back are different comfort problems, how flat panels with cushions compare to contoured seats and tensioned mesh, and the layered cushion stack. First draft 24 August 2026. Status: DRAFT. The seated-posture and pressure literature is VERIFIED per the project's founding ergonomics brief (notes/chair-project-ergonomics-brief.md, 24 Jul 2026, references checked against publishers at research time); the foam mechanics are STANDARD polymer behavior; the load-share figures are STANDARD but should be pinned to a specific source before print; KU's cushion findings are EMPIRICAL, from the bench, 24 Aug; the layered-stack analysis is DERIVED and its test is QUEUED.

The claim

The seat and the back of a chair are not the same problem wearing two orientations. The seat's problem is pressure and is solved by redistribution: contour, cushion, or membrane. The back's problem is geometry and is solved by angle, contact height, and shape, with almost no cushion at all. The project's bench practice, 50 mm of medium foam on the seat and 12 mm of closed-cell on the back, is this claim discovered empirically, and it matches what the literature took pressure-mapping rigs to establish.

What the body presents

Seated upright at dining rake, the majority of body weight, on the order of two-thirds, passes through the seat, with the remainder split between the feet and the backrest; the backrest share grows with recline [pin the split against Chaffin/Andersson or Humanscale before print]. The seat load concentrates under the ischial tuberosities, and anatomy is unhelpful here: flexing the hips to sit rotates the gluteal muscle away from exactly those points, so the seated body has less natural padding than the standing one. On a hard flat panel the pressure peaks under the two sit bones dominate the whole comfort budget, and discomfort over time tracks those peaks; sustained pressure above roughly capillary level, about 32 mmHg, is the standard threshold at which tissue begins to object over long durations. The back presents the opposite case: lower load, delivered over broad muscle, with only the spinous processes as sharp points. The founding brief's line stands: on an unpadded chair the backrest does more comfort work than the seat, and it does that work with geometry. KU's open-lumbar observation from the first night is on the tape: "Few commercial dining chairs have what I would consider lumbar support... The lumbar area is open."

Three routes to the same area

Contour, cushion, and mesh are one strategy in three materials: grow the contact area so the peaks fall. A carved contour does it geometrically, meeting the body's envelope before loading it, which is why a Windsor saddle is tolerable for hours with zero compliance; Li et al. (2020) measured contours cutting peak pressures roughly 18 to 31 percent against flat surfaces, with the caveat that a wrong contour is worse than none. A cushion does it by deformation: the panel stays flat and the foam generates the contour under load. A tensioned membrane, the Aeron's pellicle, is a contour that re-derives itself continuously for every sitter and every shift, and it also wins the second comfort axis, microclimate, since foam traps heat and moisture while mesh ventilates and bare wood sits between them. Duration is the moderator: a dining chair is a 45-to-90-minute machine, and at that duration flat panel plus modest cushion, honest contour, and mesh converge to nearly the same experienced comfort; the eight-hour claims that justify the Aeron's machinery are answering a question the dining table does not ask. The brief's other finding belongs here too: over-soft is a real failure mode, loading gluteal muscle and destabilizing the pelvis. Firm is not the enemy; the peak is.

The back: geometry, then 12 millimeters

The Eek lesson, as KU recalls it, is that his scrap chair became comfortable when the backrest was curved, and the mechanism is almost certainly plan-view: a flat back panel meets the thoracic spine's convexity as a vertical line directly on the spinous processes, and curving the panel in plan moves the contact outboard onto the paraspinal muscles. A large gain, at zero cushion [Eek's account: KU's recollection; locate a citable source or mark as told]. For the Keel Chair, whose rake and seat-back angle were frozen in the first week, the back's remaining variables are contact height and a thin dry layer: the bench verdict is that 12 mm of closed-cell foam is sufficient, and the theory concurs, since the back needs contact-softening and a little thermal warmth, not immersion. Two precedents inside the project record: the founding brief's faceted-back proposal, two flat slats at slightly different angles approximating Keegan's curve from straight stock; and Rietveld's own 1934 build rules, which specify a roll cushion at the back top, kapok or cork, with the seat carried plain. Rietveld made the same discovery.

The seat: bottoming, and the layered stack

The bench finding: even 50 mm of medium upholstery foam can bottom out under the sit bones. The mechanism is the shape of polyurethane foam's stress-strain curve: a plateau through the middle strains, then steep densification at roughly 60 to 70 percent compression, at which point the foam transmits load almost like a solid and the sit bone effectively meets the panel. Local strain under the ischial tuberosities far exceeds the cushion average, so the peaks densify first, exactly where redistribution is needed most.

The proposed stack, 12 mm of closed-cell foam against the panel with 35 mm of medium open-cell above it, helps, for four reasons, and the layer order is correct. First, the soft upper layer does the envelopment: it conforms around the sit bones and grows the contact area, which only a compliant layer touching the body can do. Second, the firm lower layer converts the bottoming event from a step to a ramp: when the topper densifies locally, the load lands on a layer that still has nearly all its stroke, so the sitter meets firm rubber rather than plywood, and the pressure spike is capped. Third, 12 mm of closed-cell is stiff enough to act slightly plate-like, bridging the concentrated load over more panel area. Fourth, the practicalities: closed-cell resists compression set, serves as a spill barrier protecting the open-cell from below, and takes the abrasion against the ply. This two-layer architecture, a soft immersion layer over a firm support base, is standard practice in upholstery and in wheelchair seating, which is the literature that has studied it hardest. Soft always goes toward the body; a closed-cell top layer would surrender both envelopment and breathability.

Two refinements worth testing alongside the stack. If bottoming persists, the medium topper's firmness is the suspect, not the total thickness: a high-resilience (HR) grade at the same 35 mm carries a more supportive curve and outlasts conventional foam, so 12 closed-cell plus 35 HR is the predicted sweet spot. And there is a flat-stock-native way to buy stroke exactly where it is needed: a relief well under the sit bones, made not by carving but by stacking, a 12 mm rim piece under the cushion with a cutout in the ischial zone, which gives the cushion somewhere to immerse. It is the founding brief's faceted-saddle idea turned upside down: the contour built from rectangles, and hidden under the upholstery.

The experiment

The founding brief already prescribed the method: Corlett and Bishop body-map discomfort ratings at timed intervals, across variants, with a sample spanning small to tall sitters. The comfort tournament, on one Keel Chair frame: seat conditions bare, 50 medium, 12 closed-cell + 35 medium, 12 closed-cell + 35 HR, and the relief-well variant; back conditions bare, 12 mm closed-cell, and a kapok-style roll at the top rail. Ratings at 20, 60, and 90 minutes. The brief noted that no peer-reviewed study directly compares contoured wood, flat wood, and cushioned seats; the tournament is product development and publishable territory at once.

Sources

The project's founding ergonomics brief (notes/chair-project-ergonomics-brief.md, 24 Jul 2026), whose references were verified at research time: Keegan 1953; Andersson et al. 1974; Mandal 1981; Corlett and Bishop 1976; Helander and Zhang 1997; Li et al. 2020; Sydor and Hitka 2023; the BIFMA and EN 1729 envelopes. The tape, 24 to 25 Jul: the seating-science run, the thinner-stiffer-foams question, and the open-lumbar observation. Rietveld's build rules: the 1934 parts sheet, per TN and the Pattern appendix. Foam mechanics: standard polyurethane behavior; pin the densification figures to a foam handbook or ASTM D3574 context at print. Load shares and the capillary threshold: standard ergonomics values, to be pinned. KU bench findings: 24 Aug 2026.

The Hundred-Year Chair · hundredyearchair.com · technical notes, public from first draft. home · the technical notes · the book