Visual Inspection in a Structural Audit: What Engineers Look For, Member by Member

'Visual' undersells it. The survey is visual the way a radiologist's reading is visual — the looking is trained, systematic and recorded. Its two products are a defect record precise enough that another engineer could find every observation, and a testing plan: the survey raises the questions that NDT will answer. The order is deliberate — testing without a survey is measurement without a hypothesis.

Stilt and ground-floor columns: the priority stop

The survey begins where consequence concentrates. Ground and stilt-level columns carry every floor above them — the bottom of the load path — and simultaneously live in the building's worst microclimate: splash from parking wash-downs, leaking plumbing stacks, poor ventilation. The engineer examines each column face for vertical cracks tracing corner bars, spalled or hollow-sounding cover, rust staining, exposed stirrups, and past repairs (a patch is a record of an old problem — and its edges show whether the problem stayed solved). Column bases get particular attention where they meet floors that hold water. Tap-testing with a light hammer finds the delaminated cover that looks intact but sounds hollow — cover that has already separated and simply not fallen yet.

Beams and slab soffits

Overhead, the engineer reads bending members through the crack grammar of the cracks chapter: fine vertical cracks near mid-span are flexure behaving as designed; diagonal cracks near supports are shear and rank among the most urgent visual findings; straight soffit cracks along a beam's length, often with rust bleeding, are the corrosion signature. Slab soffits show their own patterns — corrosion lines following bar spacing, damp patches and leaching (white mineral deposits marking long-term water passage), and sagging visible against door heads and tile lines. Wherever a bathroom or terrace sits above, the soffit below is examined as a matter of routine: wet areas write their history on the ceiling underneath.

The envelope: façades, terraces, tanks and shafts

The building's outside is where weather does its work. On façades: cracked or hollow plaster, vegetation rooted in joints (roots follow water — a fig sprouting from a crack marks a moisture path, not a gardening problem), corroding chajjas and balconies, failed sealant at expansion joints, and staining that maps water runoff. Upper levels that binoculars once served are now flown: drone survey brings every square metre of façade to inspection distance, and photogrammetry turns the imagery into a measurable record. The terrace is surveyed for waterproofing condition, ponding zones, and cracked china mosaic or blistered membrane; overhead tanks for leakage and the chronic dampness around their supports; plumbing shafts — the building's wettest vertical zones — for the corroded members that hide behind access panels no one opens between audits.

A useful mental model: the survey is water-led. Nearly every defect trail the engineer follows — stains, leaching, vegetation, hollow plaster, corrosion lines — is the record of water moving through the building. Find the water paths and you have found most of next decade's repair scope.

Classification: turning observations into information

Raw observations become an audit through classification. Each defect is graded — typically from cosmetic through moderate to severe/safety-significant — and mapped, so that patterns emerge: corrosion clustered on the weather face, distress concentrated at one stilt bay, leakage tracking one plumbing line. The classified map decides where instruments are deployed (half-cell where corrosion signs cluster, UPV where concrete quality is doubtful) and later becomes the skeleton of the report's priority schedule. In our practice the record feeds a digital twin, so every observation is pinned to its exact position on the model — and the next audit can compare like with like, member by member.

For a society, the practical value of understanding the survey is knowing what to expect and what to provide: access to stilt areas, shafts, terraces and a sample of flats (especially those below terraces and beside shafts), records of past repairs and persistent complaints, and time — systematic coverage of a large building takes days, not hours. A survey that finished suspiciously fast almost certainly sampled instead of covered, and everything downstream — testing, priorities, budget — inherits the gaps.