Shear Stress in Beams
The layers of a beam try to slide past each other, hardest at the neutral axis.
Before this Lesson
- Bending Stress
- Centroids of Areas · Statics
Lay two planks one on the other across a ditch and walk out to the middle. They sag, and their ends no longer line up: the top plank now pokes out past the bottom one. Glue them into one block, and the same wood is twice as strong and four times as stiff. All of that rides on one thin glue line at mid-depth, which keeps the planks from sliding, and which turns out to sit in the worst possible place.
Layers that can't slide
Each loose plank bends about its own neutral axis, so where they touch, the bottom of the top plank stretches while the top of the bottom plank shortens. Two surfaces pressed together, one growing and one shrinking, have to slide. Glue stops the slide by gripping both, and that takes a horizontal shear stress along the glue line.
A solid beam is a stack glued by the material itself, so wherever the beam carries shear, its horizontal planes carry horizontal shear stress.
That stress has a vertical twin. A tiny square in the side of the beam carries the internal shear as a stress , down one vertical face and up the other. That pair is a couple that would spin the square, so its top and bottom faces must carry an equal to stop it. At any point, the horizontal and vertical shear stresses are equal. At the beam's top and bottom faces, nothing outside pushes sideways, so both are zero there.
Where the horizontal push comes from
Why does a layer push at all? Cut a slice long, with moment on its left face and on its right. Then cut it again at the level you care about, and keep the piece above.
From Bending Stress, the bending stress a distance from the neutral axis is . Over the piece's left face, whose area is , it adds up to a push of:
is the first moment of the piece's area about the neutral axis, as in Centroids of Areas. The right face pushes back harder, with . Only the shear on the cut, over a strip wide and long, can hold the difference:
The last step uses , from Shear and Moment Diagrams. So shear stress lives where the moment changes. Under a constant moment, is zero, the two faces push equally, and nothing slides.
The Handbook prints this under Stresses in Beams as the transverse shear stress, , with . is the area above the layer, and runs from the neutral axis to its centroid. The Handbook calls the width of the cross section, and the derivation says where: across the cut.
Q: how much area, and how far out
bundles two things: how much area lies beyond the level, and how far from the neutral axis it sits.
Measure from the neutral axis, not the top face, because the push grew with distance from the axis. And is the width at the level, not the widest part of the section.
Slide the level down from the top face and watch . At the face, nothing lies beyond, so . Every strip the level passes on the way down lies above the axis and adds to , though less and less as it nears the axis, where its is small. So grows until the neutral axis, then shrinks as strips below the axis subtract.
Below the axis, use the part below the level instead. The whole section's first moment about the neutral axis is zero, which is how Bending Stress located the axis, so the parts on either side of a level have the same . Use whichever is simpler. That zero is also why never uses the whole section.
A rectangle's parabola
In a rectangle wide and deep, the area beyond a level above the axis is a strip tall, with its centroid halfway up it:
That's a parabola in , and with and fixed, so is . At the neutral axis, :
The Handbook doesn't print this shortcut, but the exam leans on it. The peak is 50% above the average, , because the faces carry none of the shear.
Beside the bending stress, it's the mirror image. Bending stress is largest at the faces and zero at the neutral axis, and shear stress is the reverse. So two planks glued face to face put the glue right on the peak.
The largest shear stress in a rectangular beam
The statics half takes where the shear is largest, not where the moment is. Under a uniform load, the moment peaks at midspan, where .
From Shear and Moment Diagrams: for a simple beam under downward loads, the largest shear is the larger reaction, under a full uniform load and under a center point load. For a cantilever, it's the whole load, at the fixed end. The Handbook's beam tables print moments but not these shears.
Timber is where this check bites. Wood splits easily along its grain, on exactly the planes where the layers slide. The Handbook's table of Average Mechanical Properties of Typical Engineering Materials gives select structural Douglas fir an ultimate shear strength parallel to the grain of only 0.90 ksi, and it lists no allowable stresses for design, so a question states its own.
I- and T-sections: the web does the work
In an I or a T, changes with the level. Just above and just below the web-flange junction, the area beyond is the same flange, so is the same. But drops from the flange width to the web thickness , so the stress jumps by a factor of . Inside the web, it keeps rising to the neutral axis, but only a little, because the flange, far from the axis, has already supplied most of . So the flanges carry little shear stress, and the web carries most of .
Anywhere in the web, is the web thickness. And is the rectangle's alone, so for an I or a T, use the formula. Design codes check steel beams for shear differently, with a formula built on the web's area, in Selecting Steel Beams.
Working back to a depth
For a rectangle, shear design is one line: keep the peak within the allowable shear stress and solve for the depth.
It needs only the largest shear, and no moment. A real beam must pass bending too, but a shear question asks only for this.