Beam Deflection by Superposition
The deflections from separate loads simply add.
Before this Lesson
- Beam Deflection · Mechanics of Materials
A steel floor beam spans 24 ft. It carries its slab as a uniform load, and a 12-kip air handler sits on it 8 ft from one support. How far does the beam sag under the unit? Look through the Handbook's beam tables and you won't find this beam. Every row holds exactly one load, and no table could print a row for every mix of loads a real beam carries. Two of those one-load rows, read at the right point, are all you need.
Why separate loads add
Put both loads on the beam and cut it anywhere. The moment at the cut is the uniform load's moment plus the point load's, because statics just adds forces. From Beam Deflection, each piece of a beam bends with curvature , so the curvature is a sum too. Integrate a sum and you get a sum: the slopes add, and so do the deflections. Each load's curve is zero at both supports, so their sum is too. The curve under both loads is the two one-load curves, added point by point.
The Handbook calls this the Principle of Superposition and asks two things: each effect is linear in its load, and the deformations are too small to change how the other loads act. A steel beam in service stays elastic and sags a fraction of an inch over many feet, so it qualifies. Beams with One Redundant uses the same idea on a beam with one support too many.
That's the one rule that matters: add at the same point. Each curve is lowest somewhere else, and adding the lowest values gives a deflection the beam has nowhere. The question asks about the point under the load, , so read every curve there. The tables measure upward, so a sag prints with a minus sign. Every load in this Lesson pushes the same way, so drop the signs and add sizes.
From Beam Deflection: in the Handbook's tables Simply Supported Beam Slopes and Deflections and Cantilevered Beam Slopes and Deflections, is the slope in radians, and is the largest deflection. The tables print no units. Use kips, inches, and ksi, with a load in kip/in., which is kip/ft divided by 12, or N, mm, and MPa. The Handbook's Table 1 - Typical Material Properties gives = 29.0 Mpsi (200 GPa) for steel and 10.0 Mpsi (69.0 GPa) for aluminum.
Reading a curve at the point you need
A row's is the deflection at one point. For a full-span uniform load, is the sag at midspan and nowhere else. For any other point, the row's Elastic Curve column gives the whole curve, with measured from the left support:
Put in and you get the back. Put in , and the sag is , about 13% less.
The point load has its own row, the off-center point load, with a distance from the left support and from the right. Its Deflection column gives the sag under the load itself:
Since , the bracket is , so this is , the same from either end. The row's elastic curve is fussier: it holds only for . For a point past the load, measure from the right support and swap and .
The same point, not the lowest point
Now load only half a beam. The Handbook has a row for that too, a uniform load over the left half, and it prints two deflections. One is , at : the lopsided load pulls the lowest point a little toward the loaded side. The other is the deflection at midspan:
Pair this load with a point load at midspan, and the point load's sag is a midspan sag, so the half-span load's has to be too. Take the second number.
That is exactly half of the full span's , and superposition says why. Load the right half instead, and you get the mirror image of the curve, with the same sag at midspan. Now load both halves. That's the whole span, so the two midspan sags add to . Two equal shares of it are each.
Past the last load, a cantilever is straight
Put a point load on a cantilever a distance from the wall, short of the free end. Between the wall and the load, the beam carries moment and bends. Past the load, nothing acts on it, the moment is zero, and that stretch doesn't bend at all. The tail just goes along for the ride: it leaves the load at whatever tilt the beam has there and runs straight to the tip.
So the tip deflection comes in two pieces. Up to the load, the beam is a cantilever of length loaded at its end, so the Handbook's point-load row with the load at the end gives:
The straight tail, long and tilted at radians, drops another :
Multiplied out, that's the row's own , so either works. A uniform load with a point load at the free end needs no such move: both deflections are at the tip, so add and .
Stand the cantilever on end
Fence posts and the stems of retaining walls are cantilevers too: fixed at the bottom, free at the top, and pushed from the side. Soil and water push harder the deeper you go, so the load on a wall stem grows from zero at the top to a peak at the base. Turn the page sideways, and that's the Handbook's cantilever row for a triangular load, which peaks at the wall at and falls to zero at the free end. The tables don't care which way the beam points.
Right panel adapted from the FE Reference Handbook, Cantilevered Beam Slopes and Deflections
Add a point load at the top, from a rail bolted to the wall, and superposition finishes the job:
That is well under the uniform load's . The triangle carries half the total of a uniform , and most of that presses low on the post, on short lever arms, so it makes less moment everywhere. Check which end has the peak: the row puts it at the fixed end, which is where soil puts it.
Slopes add too
Slope is the first integral of the curvature, so slopes add just as deflections do, at the same point. The exam asks for the ones the tables print: at a simple beam's supports and a cantilever's free end. Three things to watch:
- is in radians. A slope of 0.005 rad is about 0.29°. Multiply by for degrees.
- A centered load or a full-span uniform load tilts both ends of a simple beam equally, by or . An off-center load doesn't. The Handbook gives at the left end and at the right, so the end nearer the load tilts more.
- On a cantilever, the straight tail has the slope at the load, so a load short of the free end tilts the tip by , the row's .
Spread out or piled up
Gather a simple beam's uniform load into one point load of the same total at midspan. What happens to its sag? Write the total as , and compare:
The ratio is : piled at midspan, the load sags the beam 1.6 times as far. The pile sits as far from both supports as a load can get, so it makes more moment everywhere, and more moment means more curvature. Its peak moment is twice the uniform load's, but its sag isn't, because near the supports the two diagrams nearly agree.
Keep the totals equal: the point load that matches is , not . With the total fixed, the sag goes as either way, so, as in Beam Deflection, it grows 8 times when the span doubles and halves when doubles.