By the time a foundation dispute reaches an expert, the building has usually been looked at several times — by a contractor, an adjuster, sometimes a repair company selling a solution. What is rarely present is a record capable of distinguishing movement that finished years ago from movement still underway, or of separating either from the way the building was originally built. The investigation producing that record follows a fairly fixed sequence, and each step is worth understanding for what it can and cannot establish alone.
Documentation comes before repair
Crack widths, elevation differences and the visible pattern are the evidence, and patching, regrading or underpinning removes it. This is not an argument against repair — a distressed building sometimes has to be stabilised quickly — but the sequence matters.
Where repairs have already happened the investigation is not over, but it is harder. The analysis shifts to pre-repair photographs, invoices describing what was done, and whatever movement can be measured from that point forward.
Crack mapping is a measurement, not a photograph
Useful crack documentation records location, orientation, width, whether the crack passes through or around masonry units, whether the edges are offset out of plane, and the condition of the faces. Photographs alone rarely capture width or offset; a scale in the frame and a recorded dimension do.
Orientation carries information. Diagonal and stair-step cracking generally reflects shear from differential vertical movement; vertical cracking at openings reflects tension; horizontal cracking in a basement wall usually reflects lateral earth pressure rather than settlement.
The floor-level survey
A relative elevation survey across the floor plate, taken on a grid rather than at a few convenient points, produces the surface everything else is checked against. It shows where the low and high areas are, whether the shape is a bowl or a ridge, and how abrupt the transitions are.
Two cautions apply. Floors are not built perfectly flat, so part of any deviation is original construction tolerance rather than movement. And a relative survey establishes shape, not direction — knowing which points moved requires a benchmark outside the structure or a repeat survey.
Borings, sampling and in-situ testing
Below the surface the question is what the foundation bears on and how that material behaves. Standard penetration testing under ASTM D1586 gives resistance and a disturbed sample at intervals; thin-walled tube sampling recovers relatively undisturbed specimens; cone penetration testing gives a continuous profile and, with pore-pressure measurement, information about drainage.
Depth and spacing matter as much as method. Borings must reach through the compressible or expansive zone and be distributed across the footprint, because variation is the question. A single boring at the least-distressed corner can support almost any conclusion.
Laboratory testing on recovered samples
Classification under ASTM D2487 and Atterberg limits under ASTM D4318 establish soil type and plasticity. One-dimensional consolidation testing under ASTM D2435 gives compressibility and stress history, supporting an independent estimate of how much settlement the structure's loads should have caused and over what period. ASTM D4546 measures swell or collapse on wetting directly.
The laboratory work is valuable because it is independent of the building. It predicts what should have happened; the survey records what did. Agreement between the two is a finding, and so is disagreement.
Only monitoring establishes rate
A single set of measurements is a snapshot. Whether movement is ongoing is answered only by repeating the survey and the crack measurements over an interval long enough to span a full wet and dry cycle; anything shorter can read seasonal movement as progressive or miss it.
Consolidation theory supplies a second, independent check: for a layer of known thickness and properties, the expected time course of settlement can be estimated and compared against the structure's age and loading history. Where monitoring and theory agree, the conclusion is hard to dislodge.
Pre-existing, historic or progressive
Much of what a survey finds is old. Cracks patched and not reopened, painted-over separations, weathered and dirt-filled crack faces, and doors planed to fit a distortion that stabilised long ago all indicate finished movement.
Clean crack faces, spalled edges, cracks cutting through prior repairs and measurable change between visits indicate the opposite. Repair records, listing photographs, permit history and earlier inspection reports fix dates the building cannot supply itself.
What the geotechnical report did and did not commit to
The original geotechnical report is usually the most important document in the file, and it is often read too loosely. What matters is what it stated: the borings performed and where, the bearing value recommended, the anticipated total and differential settlement, the soil-preparation and fill-compaction requirements, and any conditions attached.
Equally important is what it did not address — a portion of the site never explored, a moisture condition assumed rather than measured, a recommendation conditioned on drainage being maintained afterward. IBC Chapter 18 sets the code framework for soils and foundations and ASCE 7 the design loads; the report is where site-specific conditions enter.
The construction record, and the predictable challenges
Compaction test results, special-inspection reports, as-built foundation drawings and change orders establish whether what was recommended is what got built. Their absence, where the code or the specification called for them, is itself a finding.
The challenges are worth anticipating: no stable datum for the survey, construction tolerance never separated from movement, borings too few or too shallow for a variable site, a monitoring interval too short to see a season. An investigation stating its methods, precision and limits absorbs those. A conclusion offered without them does not.
This article is general technical orientation, not a failure analysis, an engineering opinion, or advice on any specific matter. Determining the cause of a particular incident requires hands-on examination by a credentialed expert.