Beams & Bending Members

Imagine a simply supported beam - a plank resting on two supports - with a load pushing down in the middle. The beam sags into a gentle curve. Look closely at what that curve does to the material. The bottom face lengthens: its fibres are being pulled apart, so the bottom is in tension. The top face shortens: its fibres are being pushed together, so the top is in compression. Somewhere between the two, there is a layer that neither lengthens nor shortens - the neutral axis - where the bending stress is zero. As you move away from the neutral axis toward either face, the stress grows; it is greatest at the extreme top and bottom fibres.

This is the whole physics of a beam in one picture, and it has a profound design consequence: the material near the neutral axis is barely working, while the material at the top and bottom edges is working hardest. A beam is therefore an inefficient shape if it is a solid rectangle, because so much of its middle is idling. The efficient move is to concentrate material at the top and bottom - far from the neutral axis, where it resists bending best - and to remove it from the middle, where it does little. That is exactly the logic of the I-section (or the steel joist, or a concrete T-beam): two flanges top and bottom to carry the compression and tension, joined by a thin web that mostly holds them apart and carries shear. Nature and engineers converge on the same answer: get the material away from the neutral axis.

The same logic explains why depth is a beam's superpower. Because bending resistance grows with how far the extreme fibres sit from the neutral axis, a deeper beam is dramatically stronger and stiffer than a shallower one of the same material - depth counts far more than width. Doubling a rectangular beam's width doubles its bending strength, but doubling its depth multiplies its bending strength roughly fourfold and its stiffness roughly eightfold. This is why beams are drawn tall and slim, not short and fat, and why the fight over a building's floor-to-floor height is so often a fight over beam depth.

Once you can see the tension face and the compression face, a great deal of structural detailing becomes obvious. Consider reinforced concrete, which is strong in compression but weak in tension (Module 3). In a simply supported RC beam that sags, the bottom is in tension - so that is exactly where the main steel reinforcement must go, near the bottom face, to supply the tension the concrete cannot. The concrete happily takes the compression at the top. Put the main steel in the top of a simply supported beam and it would sit uselessly in the compression zone while the tension face cracked and failed - a classic and dangerous error.

But where the tension face is depends entirely on how the beam bends, and that is not always the bottom. The rule generalises beautifully: the main reinforcement follows the tension. In a beam that sags (a sagging moment, curving like a smile), the tension is on the bottom, so the steel goes low. In a beam or region that hogs (a hogging moment, curving like a frown - which happens over the supports of a continuous beam, and along the whole length of a cantilever), the tension is on the top, so the steel goes high. This is why the reinforcement in a real continuous beam weaves up and down: low in the middle of each span where it sags, rising to the top over each support where it hogs. Reading a beam's bending tells you, without any calculation, roughly where its steel must live.

Alongside bending, a beam also carries shear - the tendency for one part of the beam to slide vertically past the next, greatest near the supports where the full reaction is being transferred. Shear is resisted in RC beams by stirrups (vertical links), which is why you see closely spaced links near the ends of a beam and wider spacing in the middle. So a well-detailed beam is a map of its own internal forces: main bars tracing the tension, stirrups clustered where the shear is fiercest. All of this is codified in India by IS 456 for concrete and IS 800 for steel, but the intuition comes first: find the tension, and put the strong stuff there.

Because depth governs a beam's strength and stiffness so powerfully, architects carry a family of span-to-depth ratios - rough proportions that let you guess a beam's depth from its span before an engineer designs it. These are among the most practical numbers in early design, because beam depth drives floor-to-floor height, which drives building height, cost and daylight.

The idea is simple: a beam's depth is typically some fraction of its span. As very rough, sketch-stage starting points for ordinary loads: a simply supported reinforced-concrete beam is often around a twelfth to a fifteenth of its span deep; a continuous RC beam, being more efficient, can be shallower, perhaps a fifteenth to an eighteenth; a steel beam, using a far stronger material, is typically much shallower still, often around a twentieth of its span; and a cantilever, which works hard, needs to be deep for its reach - roughly a sixth to a tenth of its projection. A shorthand many designers keep for RC is depth in millimetres is roughly the span in millimetres divided by twelve to fifteen. So a six-metre RC beam is plausibly 400 to 500 mm deep; a cantilever projecting two metres might need 250 to 350 mm at its root.

These ratios are not laws - they shift with load, material grade, and whether deflection or strength governs - and they must never replace the engineer's calculation. But they are invaluable for a first sketch: they let you reserve realistic beam zones in a section, test whether a desired floor-to-floor height can accommodate the spans you have drawn, and catch impossible ambitions early (a very long span at a very shallow depth is a warning sign, not a detail to be sorted out later). When a structural layout feels right in section, it is usually because the beam depths obey these proportions.

How a beam is supported changes its behaviour as much as its size does. There are three archetypes, and knowing how each bends is knowing how each must be built.

A simply supported beam rests on a support at each end, free to rotate. It sags into a single curve, its maximum bending moment (and greatest tension, on the bottom) sits at midspan, and it is the simplest, most predictable case. But it is not the most efficient, because the whole beam has to carry the full midspan moment with no help from its neighbours.

A continuous beam runs over three or more supports in one unbroken length, and this changes everything. Instead of each span working alone, the spans lean on each other across the internal supports. The result is that the peak bending moments are lower than in a series of simply supported beams, and they are shared: each span still sags (tension bottom) in its middle, but over each internal support the beam hogs (tension on top). Continuity therefore makes beams more efficient - shallower for the same span - which is one reason cast-in-place concrete frames, which are naturally continuous, are so economical. The price is complexity: the reinforcement must switch from bottom to top and back, and the beam is sensitive to how the supports settle.

A cantilever is supported at only one end and reaches out into space with its far end free - a balcony, a canopy, a projecting floor. It is the most demanding of the three. The entire cantilever hogs: its top face is in tension along its whole length, and the bending moment is greatest not at the tip but at the root, where it meets its support - which is precisely where a cantilever must be deepest and most heavily reinforced, in the top. Cantilevers also deflect much more than backed-up spans and are unforgiving of error, because there is no second support to catch the load if the first is inadequate. They are thrilling architecture and a favourite of the misconception that follows.

Ask a non-engineer what limits a beam and they will say breaking. But for a great many ordinary beams, the beam is nowhere near breaking - it is limited instead by how much it deflects: how far it sags under load. This is the difference between two ways of judging a structure. Strength (or ultimate limit state) asks: will it collapse? Serviceability (or serviceability limit state) asks: will it work, look right and feel right in daily use, without cracking finishes, ponding water, jamming doors, or bouncing alarmingly underfoot? For beams, serviceability very often means deflection, and deflection frequently governs the size long before strength would.

Deflection matters because a beam that is safely strong can still sag enough to crack the plaster ceiling beneath it, tilt a floor so a marble rolls across it, bind a door in its frame, or flex enough that people feel the floor bounce when they walk - a serviceability failure even though nothing is close to breaking. Codes therefore cap deflection to fractions of the span - commonly on the order of span over 250 for total deflection, or tighter (span over 350 or 500) where brittle finishes or sensitive equipment are involved. Long-span and cantilevered floors, and floors carrying stiff partitions or stone, are the usual places where deflection, not strength, sets the depth.

The practical upshot reinforces the span-to-depth rule: the reason those ratios keep beams reasonably deep is largely to control deflection, not just to prevent collapse. A shallow beam might pass a strength check yet sag or bounce unacceptably, so the depth is dictated by stiffness. This is also why creep matters in concrete (Module 3): an RC beam keeps deflecting slowly for years under sustained load, so long-term deflection, not just the deflection on the day of loading, must be kept within limits. When you hold a beam to a sensible span-to-depth ratio, you are, more than anything, keeping it stiff enough to stay comfortable and uncracked in service.