Last chapter told you how hard the soil pushes. This one is about surviving that push. A retaining wall has exactly three ways to lose — tip over, slide out, or sink — and design is nothing more than proving, with numbers, that none of them happen.
A retaining wall is a structure that holds back soil (the backfill) and resists the lateral earth pressure it generates. It has no bracing, no floor slab, nothing tying it to another structure. It stands up purely by being heavy enough and wide enough. So the whole design reduces to one question asked three different ways: is it heavy and wide enough?
Rotation about the toe, translation along the base, and crushing of the foundation soil. Every number you compute in this chapter serves one of these three.
All four do the same job. They differ in where the resisting weight comes from — and that single difference drives cost, buildability, and the height each type can economically reach.
Plain concrete or stone masonry, thick and heavy. No reinforcement — the section is made big enough that it never goes into tension. Simple to build, but concrete volume grows fast with height, so it stops being economical past roughly 3 m.
A gravity wall with light reinforcement added so the section can be slimmed down. A middle ground: less concrete than pure gravity, less steel detailing than a cantilever.
A reinforced-concrete T or L: a thin vertical stem on a wide base slab. The clever part is the heel — the slab extends back under the backfill, so the soil sitting on it becomes stabilizing weight. The most common type, economical to about 6–8 m.
For tall walls, the stem's bending moment becomes brutal. Counterforts — thin concrete ribs tying the stem back to the heel at intervals — turn the stem into a slab spanning horizontally between supports. Used above roughly 7–8 m.
A gravity wall is a sumo wrestler — it wins on mass alone. A cantilever wall is a judo throw: it makes the soil's own weight do the work of holding the soil back. That's why it needs so much less concrete for the same height.
Stability checks tell you whether a wall works — they can't tell you what to try first. So design starts with proportioning rules of thumb, then iterates: size it, check it, adjust, re-check. These proportions get most walls close on the first pass.
Three of these belong to the wall. Two belong to the ground the wall happens to be standing on — and those two are the ones that surprise people, because a perfectly designed wall can still ride a failing hillside down.
The wall rotates forward about its front bottom edge. Resisted by the weight of the wall and the backfill on the heel.
The whole wall translates outward. Resisted by friction and adhesion under the base slab, plus passive pressure at the toe.
The foundation soil is crushed by the pressure under the base — worst at the toe, because the load is eccentric.
A slip surface passes beneath the entire wall. This is a slope-stability problem, not a wall problem — covered in the slope stability chapter.
The wall doesn't collapse, it just sinks or tilts enough to crack, misalign, or lose serviceability. A settlement problem, checked separately.
Modes 1–3 are strength checks with factors of safety. Modes 4–5 ask a different question: is the site stable, and can it tolerate the movement?
The earth pressure pushes the wall's upper body outward, trying to spin it about the front bottom edge. Everything heavy sitting on the base — concrete and backfill alike — pushes down at some distance behind that pivot and spins it the other way.
Take moments about the toe. The driving moment is the horizontal active force times its lever arm (H'/3 above the base). The resisting moment is each weight times its horizontal distance from the toe. The heel soil dominates — it's usually the single biggest term.
FS ≥ 2 for granular backfill, ≥ 3 for cohesive.
Widening the heel is almost always the cheapest fix for a failing overturning check — it adds a large soil weight at a long lever arm, and costs only a bit more slab.
Nothing is bolted down. The only thing stopping the wall from being shoved outward is friction between the underside of the base slab and the soil — and friction is proportional to how hard the wall presses down. Heavier wall, more grip.
The interface is concrete cast against soil, not soil against itself, so it can't mobilize the soil's full φ' and c'. Practice uses reduction factors k₁, k₂ ≈ ½ to ⅔. Passive resistance at the toe may be added, but conservatively it's often ignored.
Here's the subtlety that makes this check different from an ordinary footing: the resultant force under a retaining wall is not centered. The earth pressure is constantly trying to rotate the wall forward, which shifts the load toward the toe. So the pressure under the base is a trapezoid, not a uniform block — and the toe carries the peak.
qult from Meyerhof, using the effective width B' = B − 2e.
If e > B/6, the formula returns a negative qmin — soil "pulling down" on the heel, which is impossible. The heel has physically lifted off. Keep the resultant inside the middle third of the base and the whole base stays in compression.
A complete cantilever wall check. Change any dimension and watch which check fails first — try shrinking the base width B, or widening the heel, and see how differently each factor of safety responds.
Problem: A cantilever retaining wall has stem height Hs = 6 m, base thickness D = 0.7 m, base width B = 4.0 m, toe projection 1.0 m, stem tapering from 0.7 m at the bottom to 0.3 m at the top. Backfill: γ = 18 kN/m³, φ' = 32°, horizontal surface. Foundation soil: φ'₂ = 28°, c'₂ = 30 kPa, qult = 450 kPa. Concrete γc = 23.6 kN/m³. Compute all three factors of safety.
| Component | W (kN/m) | x (m) | M (kN·m/m) |
|---|---|---|---|
| Stem (rectangle) | 42.5 | 1.55 | 65.8 |
| Stem (taper) | 28.3 | 1.27 | 35.9 |
| Base slab | 66.1 | 2.00 | 132.2 |
| Soil on heel | 248.4 | 2.85 | 707.9 |
| Σ | 385.3 | 941.8 |
1. In a cantilever wall, what usually provides the largest share of the resisting moment?
2. Your calculation returns e = 0.9 m for a base width B = 4.0 m. What does this tell you?
3. Why are the base friction parameters reduced by k₁, k₂ ≈ ⅔?
4. A wall fails its sliding check. Which fix targets that specific failure mode most directly?