Cabinet Box Construction

The Forest Products Laboratory publishes a predictive equation for the ultimate withdrawal load of a screw from particleboard. The load depends on shank diameter, depth of thread embedment, the specific gravity of the board, and a coefficient that changes with where the screw goes: in inch-pound units, 2,655 for the face and 2,055 for the edge. For fibreboard — the class that includes dry-process MDF — it is 3,700 and 2,860.

Two things fall out of those four numbers. The edge is systematically weaker than the face in both materials, by around 22 percent — the arithmetic reason a dado, which puts load into bearing rather than withdrawal, is a different class of connection rather than a slightly better one. And the fibreboard coefficients sit roughly 39 percent above the particleboard ones, which at equal board specific gravity is a directly proportional advantage. It is not the ranking most people expect MDF to hold.

Plywood is the interesting omission: the Wood Handbook gives no equivalent screw coefficient, so rather than infer one, the honest report is what it does say — nail withdrawal is 15 to 30 percent lower than in solid wood of the same thickness, but plywood resists splitting near an edge considerably better. Withdrawal also scales with the square of specific gravity, so a modestly denser panel is not modestly better at holding a screw.

The Architectural Woodwork Standards define the difference in a line each: frameless construction is where the front edge of the cabinet body components is edgebanded, face frame construction is where that edge is overlaid with a frame. The standards decline to prefer either, and note that flush overlay is the default door interface for both.

In load terms that matters. In a frameless cabinet the front edge of the side panel is the front of the cabinet: it carries the hinge, resists racking, and is the edge that took the lower of the two coefficients above. Its protection is edgebanding, which the standards note varies from 0.017 to 0.118 inch — 0.45 to 3 millimetres — by manufacturer practice. That range spans nearly seven to one: thin banding is decorative, thick banding is a structural edge. In a face frame cabinet an applied hardwood frame takes the hinge instead, so the panel edge behind it is no longer the front line.

The first is creep. The Architectural Woodwork Standards define it as the increase in deflection over time, fluctuating with temperature, humidity, and load stress, and are explicit that creep is not a defect. A shelf that passes a load test on day one can continue to sag for years without anything having gone wrong. The standards list four ways to reduce it, none a repair: load the shelf less, use a stiffer material, change the support arrangement, or specify a tighter deflection factor at the outset.

Two different deflection criteria are in circulation. For shelving specification the standards give L/144 as the industry standard maximum, permitting a quarter inch of deflection in a 36-inch shelf. Their casework integrity test is stricter: load the shelf uniformly toward 200 pounds and allow a maximum deflection of one one-hundred-eightieth of the span, not exceeding a quarter inch. The standards also adopt design capacities of 50 pounds per square foot for school, hospital, and library shelving and 40 for all other shelving — well above the 15 pounds per square foot at which ANSI/KCMA A161.1 loads residential shelves for seven days.

Racking is cumulative in the same way: every uneven load, every shim that settles, every time a run is moved applies a small diagonal force the corner joints and back must absorb. A drifted box shows it first at the doors, as reveals that are no longer parallel. Moisture is the only one of the three usually an event — a slow leak reaches a cut panel edge, absorption runs fastest in exactly that direction, and in a fibre-based panel the swelling is permanent.