Calculators

Shelf span and sag calculator

Sag is not a defect that appears later. It is present the day the shelf is loaded, at a size you can predict.

On this page (7 sections)
  1. Shelf span and sag calculator
  2. The six inputs, and the one people enter wrong
  3. Why span is raised to the fourth and thickness to the third
  4. 1/180 on day one, 1/120 after creep
  5. A 36 in chipboard shelf, worked by hand
  6. What it tells you to do when the shelf fails
  7. What it will not tell you: strength, fixings, tipping

Short answer

A 36 in (91 cm) span of 3/4 in (18 mm) melamine chipboard, 11 in (28 cm) deep, carrying mixed hardbacks at 22 lb per foot, deflects 0.30 in (7.5 mm) immediately and about 0.59 in (15 mm) after years under load. Both fail: the limits are 1/180 of the span on day one and 1/120 in the long term. The same shelf in birch plywood deflects 0.08 in and passes comfortably.

The numbers, at a glance

Visible sag threshold, day one
1/180 of the span0.20 in (5 mm) over 36 in (91 cm)
Long-term threshold
1/120 of the span0.30 in (7.6 mm) over 36 in (91 cm)
Creep factor, chipboard and MDF
2.0Deflection roughly doubles over years
Creep factor, plywood
1.51.35 softwood, 1.3 hardwood
Mixed hardbacks
22 lb/ft (33 kg/m)The default load band
Vinyl records
50 lb/ft (74 kg/m)Heaviest domestic shelf load
Front stiffener credit
About 4.9 times less sagModelled as 1.7 times the section depth
One support at mid span
About 16 times less sagSag follows the fourth power of span
Thickness options
5/8 to 1 1/2 in (16 to 38 mm)Stiffness follows the cube of thickness

Shelf span and sag calculator#

Enter the clear span, the board and the load. The calculator returns the sag on the day it is loaded and the sag years later once the panel has crept, and checks each against its own limit.

Enter your measurements above.

Deflection is calculated for a simply supported beam under a uniform load, and is checked twice: the day it is loaded against a 1/180 limit, and years later against a 1/120 limit once the panel has crept. Creep factors are 2.0 for chipboard and MDF, 1.5 for plywood and 1.3 for solid timber.

Nothing you type leaves your browser: the maths runs on this page, and no values are stored or sent anywhere.

The six inputs, and the one people enter wrong#

Span is the clear distance between the inner faces of the supports, measured face to face. It is not the length of the board. A 36 in (91 cm) shelf sitting in a case with 3/4 in (18 mm) sides spans 34.5 in (88 cm), and that 1.5 in matters more than it looks, because the answer follows the fourth power of this number.

A support is anything the shelf cannot deflect past: the sides of a carcass, a vertical divider, a bracket. Shelf pins count, even though the shelf only rests on them. A cleat behind the shelf does not, unless the shelf is fixed to it along its whole length, and neither does a back panel.

The depth, material, thickness, load band and front edge complete the picture. All six run through one formula in your browser as you type. Nothing you enter is sent anywhere.

Why span is raised to the fourth and thickness to the third#

A loaded shelf is a simply supported beam under a uniform load, so

sag = 5 x w x L^4 / (384 x E x I)

where w is the load per inch of run, L the clear span, E the modulus of elasticity of the material and I the second moment of area, which for a rectangular board is b x d^3 / 12 with b the depth and d the thickness.

Two exponents carry the whole result. Span is raised to the fourth power and thickness to the third. Load, depth and stiffness are all linear. The consequences of those exponents are worked through at length in shelf span, thickness and sag; what the calculator adds is the second test that comes after them.

1/180 on day one, 1/120 after creep#

The panel checks the same shelf twice.

  1. The day it is loaded, against a limit of 1/180 of the span. That is the point at which the eye stops reading a shelf as level: 0.20 in (5 mm) over 36 in (91 cm).
  2. Years later, against a limit of 1/120 of the span, with a creep factor applied. Wood panels keep deforming under sustained load, and the factors are in the table below.

It reports whichever test is closer to failing. A shelf that is comfortable on day one and bowed in five years is the normal failure mode, and the second test is the one that catches it.

MaterialModulus of elasticityMetricCreep factor used
Melamine-faced chipboard350,000 psi2.4 GPa2.0
MDF400,000 psi2.8 GPa2.0
Birch plywood1,300,000 psi9.0 GPa1.5
Softwood, pine1,300,000 psi9.0 GPa1.35
Hardwood: oak, ash, maple1,650,000 psi11.4 GPa1.3

The plywood value assumes the face grain runs along the span. Turned the other way it loses roughly half its stiffness, and site-cut shelves often are, because it wastes less sheet.

A 36 in chipboard shelf, worked by hand#

A 36 in span of 3/4 in melamine chipboard, 11 in (28 cm) deep, carrying mixed hardbacks.

  1. Second moment of area: 11 x 0.75 x 0.75 x 0.75 divided by 12 = 0.387 in4.
  2. Load: 22 lb per foot divided by 12 = 1.83 lb per inch of run.
  3. Span to the fourth: 36 x 36 x 36 x 36 = 1,679,616.
  4. Top of the fraction: 5 x 1.83 x 1,679,616 = 15,396,000.
  5. Bottom: 384 x 350,000 x 0.387 = 51,975,000.
  6. Sag: 0.30 in (7.5 mm), against a 1/180 limit of 0.20 in (5 mm). It fails on day one.
  7. Long term: 0.30 x 2.0 = 0.59 in (15 mm), against a 1/120 limit of 0.30 in (7.6 mm). It fails again, by a factor of two.

Swap the board for birch plywood and only line five changes: 384 x 1,300,000 x 0.387 = 193,000,000, and the sag falls to 0.08 in (2 mm), or 0.12 in (3 mm) after creep. Same shelf, same load, same span, and the difference between an obvious fault and one nobody notices.

What it tells you to do when the shelf fails#

When the ratio goes over 1, the panel stops describing and starts prescribing. It gives the span the shelf would need to pass, which on the chipboard example is about 28.5 in (72 cm), and notes that one support at mid span cuts the sag to roughly a sixteenth. If the front edge is plain, it shows what a 2 in (51 mm) stiffener would bring the long-term figure down to, which here is 0.12 in (3 mm). And if the load is vinyl it says so directly, because records at 50 lb per foot (74 kg per metre) are the one domestic case where ordinary shelving practice is not conservative enough. The formats behind that figure are in book, vinyl and media storage dimensions.

What goes on itLoad per footMetric
Folded linen10 lb15 kg/m
Paperbacks15 lb22 kg/m
Mixed hardbacks22 lb33 kg/m
Art and reference books32 lb48 kg/m
Crockery, stacked32 lb48 kg/m
Vinyl records50 lb74 kg/m

What it will not tell you: strength, fixings, tipping#

It predicts deflection under a uniform load. It is a planning tool, not a structural check.

  • It does not test strength. A shelf can be inside a deflection limit and still break under a point load, and glass is governed by breaking strength rather than sag, which is why it is not in the list.
  • It assumes an even load. One heavy object at mid span deflects a shelf about 1.6 times as much as the same weight spread along it.
  • It treats every shelf as simply supported. Housing a shelf into a dado gains a little end restraint, and brackets set in from the ends behave differently again.
  • It says nothing about fixings or tipping. Anchoring and carcass proportions are in bookcase dimensions and built-in cabinetry dimensions.
  • It has no view on whether the shelf is at a useful height, set out in reach zones and useful storage heights, or on how a run of shelving is proportioned, which is in bookshelf and shelving dimensions.

For the rest of the set, see the calculators index.

Frequently asked questions#

How far can a shelf span before it sags?

For mixed hardbacks on a 3/4 in (18 mm) board, about 28 in (71 cm) in melamine chipboard, 36 in (91 cm) in birch plywood and 48 in (122 cm) in oak. Those are the spans at which the calculator's long-term check, deflection after creep against a 1/120 limit, is just satisfied. Change the load and every figure moves, because sag is directly proportional to weight.

What sag is acceptable on a shelf?

One 180th of the span on the day it is loaded, and one 120th once the board has crept. Over a 36 in (91 cm) span that is 0.20 in (5 mm) and 0.30 in (7.6 mm). Below 1/180 the eye reads a shelf as straight. The calculator applies both tests and reports whichever is closer to failing, since a board can pass the first and fail the second.

Why does the calculator show two sag figures?

Because wood panels creep. The first figure is the elastic deflection the moment the load goes on. The second multiplies it by a creep factor of 2.0 for chipboard and MDF, 1.5 for plywood, 1.35 for softwood and 1.3 for hardwood, which is the deformation the board takes on permanently over years of sustained load. Chipboard creep is largely irreversible, so unloading a tired shelf does not straighten it.

Does a front edge stiffener really help that much?

Yes. The calculator models a 1.5 to 2 in (38 to 51 mm) strip glued on edge along the front as a 1.7 times increase in the effective depth of the section, which cuts sag by about 4.9 times because depth is cubed. That is worth roughly as much as moving from chipboard to hardwood, for the price of an offcut, and it is why cabinetmakers have detailed shelves this way for centuries.

What is the clear span, exactly?

The distance between the inner faces of the two supports, not the length of the board. A 36 in (91 cm) board in a case with 3/4 in (18 mm) sides has a clear span of 34.5 in (88 cm). Measuring the outside of the case overstates the span and makes a workable shelf look marginal. Shelf pins count as supports; a back cleat does not.

Does making the shelf deeper reduce sag?

In the calculator, yes, because it holds the load per linear foot constant while depth adds stiffness in proportion. In a real room, no, because a deeper shelf gets filled with more. Enter the depth you will actually build and choose the load band for what will actually sit on it. If a 14 in (36 cm) shelf will carry two rows of books, use the next load band up.

Sources and standards referenced#

  1. Wood Handbook: Wood as an Engineering Material USDA Forest Products LaboratoryModulus of elasticity and creep behaviour for timber and wood panels
  2. Panel Design Specification American Wood CouncilBending stiffness values for plywood and composite panels
  3. BS EN 16121: Non-domestic storage furniture British Standards InstitutionDeflection and stability test criteria for shelving
  4. Architects' Data Ernst Neufert, Wiley-BlackwellShelf depths, spacings and domestic storage loads
  5. Time-Saver Standards for Interior Design and Space Planning De Chiara, Panero and Zelnik, McGraw-HillStorage unit dimensions and load allowances

We cite published codes, industry planning guidelines and manufacturer specifications. Where a figure is our own recommendation rather than a published standard, the text says so. See how we measure.