One of my first research chores at the University of Illinois, back in 1985, was to investigate residential basement walls that were buckling inward. This was common—almost universal—in 8” block basement walls of depression-era and post-war housing, with as much as 7 feet of unbalanced fill. I was invited to work on this with Paul duMontelle of the Illinois State Geological Survey. He and I went to one 5-year old house, with a concrete foundation that showed no buckling. We did an excavation about 3’ deep, exposing the profile of soil in contact with the foundation. Here is a photograph of that excavation.

You can’t make much of that photo. But a geologist can, especially by taking samples to the lab. The A horizon is the topsoil, rich in organic materials, usually just a few inches thick around houses. It is very easily distinguished from the subsoil—the B horizon—using analysis of particle size and composition. What was clear to him, and perhaps to you if you squinch hard enough, is that the A horizon takes a dive at the foundation, extending in a thin downward layer that is in contact with the concrete.

Of course. The soil here contains clay, which shrinks and swells with changes in moisture content. During dry spells, the soil appears cracked. And by far the biggest crack occurs where the soil meets the foundation. This crack may persist for weeks or more. It is common sense that wind, or brief rains, or traffic, or little mouse feet scurrying along the foundation, will cause particles from the A horizon to fall into the crack which may be inches or a foot or more deep. Then the rains will come and the soil will swell back to its original dimension, plus this increment that came during the dry spell. This applies a considerable lateral force to the foundation wall, primarily where the A horizon soil accumulates, a foot or so below grade.

A very simplified loading diagram of lateral forces against foundation walls is shown in the drawing below. The drawing is from the Building Foundation Design Handbook, from the US Department of Energy 1988. This book is sadly out of print, and I’m unable to find it on the web, even from OSTI.gov, the library for US DOE. The principal author, Ken Labs, sadly died not long after publication. The other colleague-co-authors have gone on to illustrious careers in building science. A loading diagram like this might make sense if soil was a liquid. Soil mechanics is a field far more developed than we see here, but this was considered sufficient. In structures class, I heard this called Rankine loading. William Rankine, a Scottish engineer in the mid to late 1800s, did work on soils and dams and structures (and thermodynamics, and…), but I’ve not encountered anything this elementary in his work. As is evident, it did not consider shrink-swell ratchet loading.

Soil is not a liquid. If this lateral loading theory was correct the buckling would be at the bottom third of the wall. Instead, it’s usually at the top third of the wall.

I encountered an interesting confirmation of this at Frank Lloyd Wright’s Taliesen home and studio in Wisconsin. In the yard at a high point of the site is a stone retaining wall a few feet high. And the condition of the retaining wall is very good, on the right hand side. The left side shows a fair amount of deterioration. The difference between these two ends is this: on the right side the soil extends up and over the stone so no crack occurs during dry spells. On the left side the retaining wall is higher than the soil it retains, so a crack during dry spells is what is to be expected.

Paul duMontelle and I learned something important from this investigation, and it has been confirmed in all I’ve seen in the subsequent almost 40 years. We wrote an internal report but nothing more. We never published the results. We didn’t have journals back then interested in why low-income housing has a basement problem. I would love to point you toward an authoritative peer-reviewed article with impact, however…no can do. You’ll have to trust me on this: The cyclic shrink and swell of soil in contact with retaining walls (including basement walls) will cause A horizon soil to accumulate in a vertical layer against the wall, leading to elevated lateral forces and buckling or other deformation of the structure.

There is a sizeable industry of professionals who solve residential basement foundation problems. I have been disappointed in their understanding of the forces acting on the walls they repair. They are warmly invited to set me straight in the comments below. Shouldn’t everyone know these forces? If Paul and I had published it, yes. We didn’t so I guess any misunderstanding should be on us, not on them.

In future posts, I will be giving other examples of cyclic effects giving rise to directional ratcheting that distresses buildings. And I may discuss periodic and monotonic mathematical functions that take an interesting turn with complex numbers. Later.

You may be interested in how to prevent this from happening. The principal strategy here is to “flash” the building into the ground, preventing crud from falling into the gap created by shrinking soil. I will discuss how to do this soon, in an upcoming post. But here’s a hint, again from the Building Foundation Design Handbook.