# Stability

Weights and uplift on a retaining wall, sliding on the base, the eccentricity that takes the place of overturning, and the checks under an earthquake.

Together with the [earth pressure](https://rischio.io/docs/retaining-wall/theory/earth-pressure/), the weights and the uplift decide whether the wall slides on its base, and where the resultant crosses the base.

## Weights and uplift

The soil between the wall and the virtual back, wedge above the crest included, weighs $\gamma$ above the water table and $\gamma_{sat}$ below. Concrete weighs the same under water; a [gabion](https://rischio.io/docs/retaining-wall/theory/gabions/#weight-under-water), whose voids fill with water, weighs $\gamma + n\gamma_w$ there, with $n$ the porosity of its rock fill, 0.40 by default. Buoyancy enters as uplift on the base: hydrostatic at the heel and at the toe, from the water on each side, and linear in between. A gabion’s voids fill to that same line, so they hold the water the uplift pushes with.

## Sliding

The base slides on the stratum it rests on, and resists by AASHTO’s 10.6.3.4, the rule free mode takes too (11.6.3.6 sends a wall there). With $\Sigma V$ the vertical force net of uplift and $B$ the base:

-   **On a drained soil**, friction: $R_\tau = \Sigma V \tan\phi$, concrete cast against the soil ($C = 1.0$, Eq. 10.6.3.4-2). A drained soil’s $c'$ isn’t counted.
-   **On a clay**, its undrained strength: $R_\tau = c_u B$.
-   **On a drained stratum over a clay**: at least 0.15 m of it below the base (AASHTO’s 6 in. of compacted granular material), the area under the lesser of $c_u$ and half the normal stress across the base, $R_\tau = \int_0^B \min(c_u, \sigma'_v/2)\,dx$ (Fig. 10.6.3.4-1); thinner, the clay’s $c_u B$. The base takes the lesser of this and its friction. The normal stress $\sigma'_v$ varies linearly from toe to heel with the resultant’s eccentricity $e$, and is zero where the base lifts, beyond the middle third: $c_u$ holds over the width $B_{s_u}$ where $\sigma'_v$ passes $2c_u$, half the stress over the rest. A centred resultant gives $\min(c_u, 0.5\,\Sigma V/B)\,B$; an eccentric one, less.

$FS = R_\tau/\Sigma H$. A line load over the heel, within the base’s width, adds nothing to $\Sigma V$ here nor in the eccentricity, which it would hold: FHWA NHI-06-089 leaves it out of both (§10.5.3), and its weight bears only in the [bearing](https://rischio.io/docs/retaining-wall/theory/bearing-capacity/#the-base-as-a-strip-footing). A [gabion level](https://rischio.io/docs/retaining-wall/theory/gabions/) slides on the one below with $\Sigma V \tan\phi$ of its stone. The soil in front isn’t counted, the safe side AASHTO allows (11.6.3.5). Under AASHTO the check is $\phi_\tau R_\tau$ against the factored horizontal load, with $\phi_\tau = 1.0$, the soil shear component of Table 11.5.7-1.

## Eccentricity

$e$ is how far from the middle of the base the resultant crosses it. Within the middle third, $e \le B/6$, the whole base is in compression; beyond it, $q_{min}$ turns negative and that edge lifts off. The eccentricity takes the place of a factor of safety against overturning, in free mode as under AASHTO (11.6.3.3), and the report gives the overturning moment about the toe. Beside it, as a reference with no verdict of its own, Results gives the resisting moment about the toe over the overturning one, $\Sigma M_R/\Sigma M_O$, the same two moments the eccentricity takes, $e = B/2 - (\Sigma M_R - \Sigma M_O)/\Sigma V$: under AASHTO, the governing combination’s factored moments; on each gabion level and under an earthquake, their own.

## Under an earthquake

With an earthquake, the wall is checked again on the forces of the [seismic case](https://rischio.io/docs/retaining-wall/theory/earth-pressure/#under-an-earthquake), by AASHTO’s Extreme Event I (11.6.5): sliding, the eccentricity and bearing, nothing in front counted. The eccentricity’s limit there depends on the share of the live surcharge in the combination, $\gamma_{EQ}$: the middle two thirds of the base at 0, eight tenths at 1 (11.6.5.1). Under AASHTO a dead surcharge over the heel holds the wall at 1.00 in all three, as the permanent loads do at Extreme Event I (C11.5.6), and a live one bears at $\gamma_{EQ}$ in the bearing only; both, as every weight under the earthquake, times $(1 - k_v)$. A line load within the base’s width bears in the bearing only, in both modes: a dead one at 1.00 and a live one at $\gamma_{EQ}$, in or out on its own, out with its inertia. In free mode the factors of safety are unfactored, with USACE EM 1110-2-2502’s limits for an earthquake beside them (Table 4-1, case R3: sliding 1.1, bearing above 1.0), a surcharge’s weight bears in the bearing only, never holding the wall, and the resultant has to stay within the base, $e \le B/2$.

---

Source: https://rischio.io/docs/retaining-wall/theory/stability/
