Triaxial Tests and Failure Planes
At failure, Mohr's circle just touches the strength envelope, and that one point fixes everything else.
Before this Lesson
- Shear Strength
- Principal Stresses and Mohr's Circle · Mechanics of Materials
A vacuum-packed brick of ground coffee is hard. You can stand it on end and lean on it. Snip a corner, and with a hiss it goes soft and the grounds pour like sand. The coffee didn't change. The squeeze did: while the bag was sealed, the air outside pressed on it at 14.7 psi with almost nothing pushing back, and that held the grains together hard enough for their friction to make a solid.
A triaxial test turns that brick into a measurement. Squeeze a cylinder of soil all around, then push on its ends until it gives way. This Lesson works out how hard you have to push.
Squeezing a cylinder
The specimen is a soil cylinder about twice as tall as it is wide, sealed in a rubber membrane inside a chamber of water. Pressurize the water, and it presses on every face with the cell pressure . A piston then pushes on the ends only. That extra is the deviator stress, , and the total vertical stress is . A valve on the drainage line decides whether water can leave the soil.
Water and rubber can't grip, and the end caps are treated as smooth, so no shear acts on the specimen's faces, which makes them principal planes: acts on horizontal planes and on vertical ones. Soil almost never carries tension, so soil mechanics takes compression as positive, the reverse of the convention in Principal Stresses and Mohr's Circle. Compressive stresses plot to the right.
Every plane through a point is a point on one circle, with its normal stress across and its shear up. The circle's ends on the axis are the principal stresses. Turning a plane by moves its point around the circle.
On the circle, and are the ends, so the center and radius are:
These are the and of Principal Stresses and Mohr's Circle. The Handbook's Mohr-Coulomb diagram, in its Geotechnical section, labels them , the mean normal stress, and , the maximum shear stress. The radius is half the deviator stress.
Failure is the circle touching the line
Now hold and push the piston harder. The left end of the circle stays put while the right end slides out, so the circle swells. Every plane in the specimen is a point on it, and the envelope from Shear Strength marks the most shear each plane can carry. While the circle sits below the line, every plane holds. When it touches the line, one plane has reached its strength, and the specimen slips along it.
For a clean sand, , and the envelope runs through the origin at the angle . The radius to the tangent point meets the envelope at a right angle, so , the center, and make a right triangle, with hypotenuse and the side opposite :
Adapted from the FE Reference Handbook, Geotechnical section, Mohr-Coulomb failure diagram
Only the triangle's angle matters, not its size. Double , and the triangle scales up, so doubles too: for a sand, fixes the ratio . Solve for it:
The last step is a trig identity. Check it at : , and . That sand fails when reaches three times the cell pressure, whatever the pressure. That's the coffee brick's secret: squeeze it harder, and it holds proportionally more.
The Handbook prints neither form above, but under Horizontal Stress Profiles and Forces, its Rankine passive coefficient is exactly at failure, and the active coefficient is its inverse. Lateral Earth Pressure shows why.
The plane that fails
The tangent point is the plane that failed, and its coordinates are that plane's stresses, which the Handbook's diagram marks . To find which way the plane faces, start at the end of the circle. It stands for the plane acts on, the major principal plane, which is horizontal in a triaxial test. The angle at the center from there around to is . Inside the triangle, the angle at the center is , so:
The failure plane is tilted from the major principal plane. The Handbook's diagram marks but doesn't print this result. For , that's 60° from the horizontal. Measured from the vertical, the same plane is at : the right plane from the wrong reference. Its stresses come off the triangle: and .
Clay with the drains shut
Now shear a saturated clay with the valve shut the whole time. Raising the cell pressure then raises the pore pressure by the same amount, just as the water took a quick fill's weight in Shear Strength, so the effective stress and the strength don't change. Every circle has the same diameter and touches the same horizontal envelope, the line at . A test with the drains shut throughout is unconsolidated-undrained, or UU.
So the cheapest version skips the cell. In the unconfined compression test, a clay cylinder stands in the open, , and the axial load at failure over the area is the unconfined compressive strength:
The circle runs from the origin to , and a horizontal envelope touches it at the top, where the shear is the radius:
is the diameter, and the radius. The Handbook prints neither, so carry this one in. The top is 90° around from the end, so the failure plane is at 45°, which is with . And unlike sand, which would slump like the opened coffee, a saturated clay stands with no cell, because its pore water went into suction when the sample was unloaded.
Matching the test to the field
A triaxial test's name records two stages. The first word covers the cell pressure going on: consolidated means the valve is open and the clay settles under that pressure, as clay in the ground has under the soil above it, and unconsolidated means it stays shut. The second covers shearing: drained, with the valve open and the loading slow enough that no excess pore pressure builds up, or undrained, with it shut. Match the test's drainage to the field's.
| Test | Under cell pressure | While shearing | Use for |
|---|---|---|---|
| UU | Drains shut | Drains shut | End of construction on clay |
| CU | Drains open | Drains shut, u measured | Long term, from σ′ = σ − u |
| CD | Drains open | Drains open, slowly | Long term, directly |
An embankment built in a few months loads a soft clay faster than it can drain, so at the end of construction the field clay hasn't drained either. A UU or unconfined test gives its , with . That's often the critical moment, since the clay gains strength as it drains.
The long term runs on and . A consolidated-drained (CD) test measures them directly, but shearing a clay slowly enough can take days. A consolidated-undrained (CU) test shears quickly with the valve shut and measures the pore pressure . Subtracting from both principal stresses slides the circle left without changing its size, onto the same envelope a CD test finds. Without , a CU test gives only total stresses. And a soil with cohesion needs at least two tests, with the envelope drawn tangent to both circles.