Shelf Sag Calculator

Every sagging bookshelf was built by somebody who thought the board looked thick enough. The reason judgement fails here is that deflection does not scale the way intuition expects: span enters the formula to the fourth power and thickness to the third.

The formula, and what each term does

A shelf carrying an evenly distributed load and resting on supports at both ends deflects at mid-span by

δ = 5 w L⁴ ÷ (384 E I)

where w is load per unit length, L is the clear span, E is the modulus of elasticity of the material, and I is the second moment of area of the cross-section. For a rectangular board, I = b d³ ÷ 12, with b the depth front to back and d the thickness.

Two consequences follow immediately, and between them they explain almost every sagging shelf.

Span to the fourth power. Double the span and deflection goes up sixteenfold. A 30 inch shelf that sags an invisible 1/32 of an inch will sag half an inch at 60 inches under the same load. This is why adding one support in the middle of a long run beats every other intervention: halving the span divides the sag by sixteen.

Thickness cubed. Going from ¾ inch to 1 inch is a 33 percent increase in depth but a 137 percent increase in stiffness. Going from ¾ to 1½ is eight times stiffer. Depth front to back, by contrast, enters only linearly: a 16 inch deep shelf is twice as stiff as an 8 inch one, not eight times.

Material matters less than people assume

The modulus of elasticity for common shelving materials, in pounds per square inch:

• Particleboard about 350,000
• Melamine-faced board about 400,000
• MDF about 500,000
• Plywood about 1,300,000
• Pine and SPF about 1,400,000
• Red oak about 1,800,000
• Hard maple about 1,830,000

The interesting comparison is not oak against pine, which differ by less than a third. It is plywood against particleboard, which differ by nearly four times. That is why flat-pack shelving sags and site-built plywood shelving does not, and it has nothing to do with craftsmanship.

Switching from pine to oak buys about 30 percent less sag. Adding a quarter inch of thickness buys 137 percent. Adding a support buys 1,500 percent. They are not comparable interventions.

How much sag is acceptable

Structural practice expresses deflection limits as a fraction of span:

• L/360 is the limit used for floors under plaster ceilings, and is effectively invisible.
• L/240 is a common general limit.
• L/180 is the usual working limit for shelving, where a slight bow is tolerable.

In absolute terms, the eye starts to notice a bow at roughly 1/8 of an inch over a typical 32 inch shelf, and finds it obvious by 1/4. A shelf against a wall shows its sag more than a freestanding one because the straight line of the wall gives the eye a reference.

Ways to stiffen a shelf without making it thicker

Since I depends on depth cubed, adding material at the edges is far more effective than adding it everywhere:

A front edge. Glue a strip of solid wood on edge along the front, say ¾ by 1½ inches. It adds very little weight and can double or triple the effective stiffness, because it puts material where the bending stress is highest. This is the single best trick in shelf building.

A back cleat. A strip screwed to the wall along the back supports the rear edge continuously and, on a shelf fixed at both ends, changes the load path substantially.

Torsion box construction. Two thin skins with a light grid between them, which is how long museum and gallery shelves are built. Extremely stiff for its weight, and considerably more work.

What this calculation assumes

It models a simply supported shelf: resting on supports at both ends, free to rotate there, with an evenly spread load. That is the conservative case and the right one for a shelf sitting on pins or brackets.

A shelf fixed rigidly into cabinet sides, glued and dadoed, behaves as a partially fixed beam and deflects noticeably less. A shelf carrying one heavy object at the middle rather than an even load deflects more for the same total weight. Real shelves fall somewhere between, and the figure above is the safe end of that range.

One thing the formula does not capture at all: creep. Wood and especially particleboard deflect further over months under a constant load than they do the day it is applied. A shelf that looks acceptable when first loaded can be visibly bowed a year later, which is another argument for designing to the stiff end rather than to the limit.