Building Failure Analysis: A Field Guide to Reading the Evidence

What "Building Failure Analysis" Actually Covers

Attorneys and adjusters use the phrase "building failure analysis" as a catch-all, but the term covers at least five distinct categories of distress, each with its own diagnostic signature, testing protocol, and cost structure. A foundation heave claim and a curtain wall leak claim both get labeled "structural failure" in the file, but the evidence, the applicable codes, and the forensic timeline are nothing alike. Before you retain an engineer, it helps to know which category you're probably dealing with — it changes who you call, what documents you preserve, and how long the investigation will take.

I've reviewed failures across 25 states over 15 years, and the pattern holds everywhere: the failure mode usually announces itself through observable physical evidence well before anyone calls a structural engineer. Learning to read that evidence — not to replace an engineer's opinion, but to triage the claim correctly — saves clients real money and protects evidence that gets destroyed by rain, demolition crews, or well-meaning maintenance staff. For the full investigative methodology we use once retained, see our building failure analysis overview; this piece focuses on what the evidence looks like before that process starts.

Settlement and Foundation Failures: Crack Geometry Tells the Story

Foundation-related failures show up as diagonal, stair-step cracking through masonry joints, typically at 30 to 60 degrees from horizontal, widest at the top and tapering toward the foundation. In drywall and stucco, you'll see cracking radiating from door and window corners — the stress risers in any wall system. The critical measurement isn't crack width alone; it's differential settlement, expressed as angular distortion (delta/L). Skempton and MacDonald's classic criteria put cosmetic cracking onset around 1/500 angular distortion and structural damage risk above 1/150. A building settling uniformly by three inches is a non-event structurally; a building settling two inches on one corner and a quarter-inch on the opposite corner over a 40-foot span is a serious problem even though the total movement is smaller.

Soil reports matter more than most adjusters expect. Expansive clay soils (high plasticity index, PI above 25) swell and shrink seasonally, producing cyclical cracking that opens and closes with moisture content — a pattern distinct from the one-directional, ever-widening cracks typical of consolidation settlement under new fill or poor compaction. We document crack width with a comparator card, log it against a date and weather event, and in disputed cases install vibrating-wire crack monitors to capture movement over 60-90 days. See our foundation cracks guide for the thresholds we use to distinguish cosmetic from structural, and differential settlement and foundation failure in our failure library for mechanism detail.

Structural Collapse and Load-Path Failures

Collapse failures are the ones that generate headlines, but the physical evidence that precedes them is often visible for months. Punching shear failures around slab columns show a characteristic cone-shaped fracture surface and radial cracking on the slab soffit before the connection fully fails — ACI 318's shear provisions exist precisely because this failure mode is brittle and gives little warning compared to flexural failure. Connection failures in steel and wood framing show up as elongated bolt holes, sheared welds, or split members at fastener locations, almost always at a point where the load path was altered — a wall removed without an engineered header, a column relocated during a tenant fit-out, added rooftop equipment that was never checked against the original design loads.

The forensic question in a collapse case is rarely "did it fail" — it's "why did this specific element fail first." That requires reconstructing the as-built load path, which frequently doesn't match the permitted drawings. On a parking structure collapse I investigated, field verification found three structural columns had been relocated roughly 18 inches from the stamped structural drawings during construction, shifting tributary load onto a transfer beam never designed for it. See structural collapse and parking garage failure for more on how these load-path failures develop.

Envelope and Water Intrusion Failures — the Slow-Motion Failure

Unlike collapse, envelope failures develop over years and the damage is mostly invisible until it isn't. Moisture meter readings above 19% in wood framing indicate active decay risk; readings in the high 20s to 30s mean decay is likely already underway. We routinely find EIFS and stucco systems with no functioning drainage plane — water gets behind the cladding through a sealant joint or window flashing defect and has nowhere to go, so it wicks into sheathing and framing. By the time staining or bulging appears on the interior finish, the structural sheathing behind it has often lost 30-50% of its fastener holding capacity to decay.

The forensic distinction that matters for claims purposes is proximate cause: was this a sudden storm event that overwhelmed an otherwise functioning envelope, or was it a long-term design or installation defect that a storm merely exposed? Insurance policies treat these very differently, and the physical evidence — growth rings in decayed wood, staining patterns that predate the loss date, efflorescence buildup — usually answers the question even when nobody kept maintenance records. See building envelope failures, water intrusion, and wood deterioration for the specific defect patterns we see most often.

Corrosion-Driven Failures in Steel and Reinforced Concrete

Corrosion failures in reinforced concrete follow a predictable sequence: chloride ions or carbonation reach the rebar, the passive oxide layer breaks down, the steel corrodes and expands up to six to ten times its original volume, and that expansion cracks and eventually spalls the concrete cover. We test for this with half-cell potential surveys per ASTM C876 — readings more negative than -350 mV indicate a greater than 90% probability of active corrosion at that location — and carbonation depth testing with a phenolphthalein indicator sprayed on a fresh concrete break. A carbonation front that has reached within 1/4 inch of the rebar on a structure with only 3/4 inch of clear cover is a near-term failure, not a long-term risk.

Structural steel corrosion is more visually obvious but harder to quantify without ultrasonic thickness testing. A beam flange that looks heavily rusted might have lost 5% of its section or 40%; you can't tell by eye, and the difference determines whether the member needs reinforcement or wholesale replacement. On industrial and coastal properties we regularly find connection plates and base plates with measured section loss exceeding 25%, which materially changes the capacity calculations in the original design. See steel corrosion and concrete cracking for how we document and quantify these losses.

Case Example: Post-Tensioned Parking Deck, Charlotte, NC

A commercial property owner retained our firm after tenants reported standing water and visible cracking on the second-level deck of a seven-year-old parking structure. Field survey found transverse cracks averaging 0.012 to 0.015 inches at the deck surface, widening to 0.030 inches directly over several post-tensioning tendon profiles, with efflorescence staining consistent with chronic water intrusion through the cracks rather than a single storm event. Deflection surveys showed 1.4 inches of downward deflection over a 20-foot bay — against an L/360 live-load deflection limit of roughly 0.67 inches for that span, the measured deflection was more than double the design allowance.

Core samples at four locations showed chloride content at the rebar depth averaging 4.2 lbs per cubic yard, well above the ACI 318 corrosion-initiation threshold of roughly 1.0 to 2.0 lbs/cy depending on exposure class. Half-cell testing returned readings between -400 and -480 mV at the worst locations. Exposed tendon inspection at two spalled areas found strand section loss of approximately 22%, concentrated at low points in the tendon profile where the original waterproofing membrane had failed at construction joints — a design and installation defect, not deferred maintenance. The finding shifted the claim from a maintenance dispute between the owner and property manager into a construction defect claim against the original contractor and waterproofing subcontractor, and the matter settled for $1.8 million covering tendon replacement, membrane reconstruction, and deck resurfacing before the case reached trial. The lesson for claims handlers: deflection and crack-width measurements, taken early and compared against code-based thresholds, are what moved this from a disputed maintenance issue to a defensible construction defect finding.

What Attorneys and Adjusters Should Document Before the Engineer Arrives

The single biggest avoidable loss in these cases is evidence destroyed before anyone measured it — demolition starts, repairs get made, rain washes out a staining pattern. None of the following requires engineering training, and all of it materially improves what a forensic engineer can later conclude.

• Photograph every crack and area of distress with a ruler or coin in frame for scale, and repeat from the same vantage points if the condition is monitored over time.
• Record the date, recent weather (especially the 72 hours before a reported loss), and any recent construction, renovation, or equipment changes nearby.
• Preserve original design drawings, geotechnical reports, and any prior inspection or repair records — these are often the fastest way to establish a baseline condition.
• Avoid immediate demolition or 'temporary' repairs in the failure area until an engineer has documented conditions, even if that means temporary shoring instead of removal.
• Note any prior complaints, work orders, or tenant reports referencing the same area — these establish a timeline that matters for both causation and notice arguments.

When Field Observations Aren't Enough: Instrumented Testing and Formal Reports

Visual survey gets you a hypothesis, not a defensible conclusion. Once a case has litigation exposure or claim value above roughly $50,000, we typically move to instrumented testing — crack monitors, tilt meters, ground-penetrating radar to locate reinforcement and verify as-built conditions against drawings, and in some cases short-duration load testing on suspect members. The deliverable also changes: a property condition assessment summarizes current condition for a transaction, while a failure analysis report has to establish mechanism, timeline, and proximate cause to a reasonable degree of engineering certainty, because it will likely be read by opposing counsel and tested under Daubert or a state equivalent.

If you're uncertain which category your situation falls into — or whether the observed distress is cosmetic or structural — that's the right moment to call a forensic engineer rather than guess. Our structural inspections and engineering due diligence services both start with the same field triage described above; the scope narrows once we know what type of failure we're actually looking at. For claims and litigation support specifically, see insurance claim support and litigation expert witness, and browse prior matters for examples closer to your fact pattern.

Frequently Asked Questions

What does a building failure analysis actually involve?

It's a structured investigation to determine the mechanism, cause, and timeline of a structural or building system failure — not just a description of damage. It typically includes field survey, material testing (concrete cores, steel thickness, moisture content), review of original design documents, and a written report with findings tied to applicable codes and standards, suitable for use in litigation or claims.

How long does a building failure analysis take?

A straightforward investigation with clear physical evidence can be completed in 2-4 weeks from site visit to draft report. Cases requiring lab testing (chloride analysis, metallurgical testing), instrumented monitoring over weeks or months, or coordination with multiple experts commonly run 8-16 weeks. Litigation deadlines and discovery schedules often drive the actual timeline more than the engineering work itself.

Can I tell the difference between cosmetic and structural cracking myself?

You can screen for it, but you shouldn't rely on your own judgment for a claim decision. Diagonal stair-step cracks wider than 1/4 inch, cracks that are wider at one end than the other, or cracking accompanied by sticking doors, uneven floors, or visible separation at the roofline are red flags that warrant an engineer's evaluation rather than a maintenance work order.

Who typically pays for the forensic engineer in a property damage claim?

It depends on the policy and the posture of the claim. On first-party property claims, the insurer frequently retains its own engineer, and the policyholder's counsel or public adjuster may retain a second expert if findings are disputed. In third-party or construction defect litigation, each side typically retains its own expert, with costs allocated through the litigation or settlement.

Is a building failure analysis report enough to support litigation, or do I need something else?

A properly scoped report can serve as the foundation for expert testimony, but it needs to meet admissibility standards — documented methodology, reproducible testing, and conclusions tied to recognized codes and engineering principles rather than opinion alone. We build every report assuming it may be challenged under Daubert or a state equivalent, which is a different bar than a report written only for an insurance file.