Why Ground Floor Problems Usually Start Below the Slab

Much of the confusion here is vocabulary. The component entries draw the boundaries carefully, and the distinctions are the whole subject: a formation is a surface, a subgrade is a body, engineered fill is ground that was built, a capping answers weakness below, a sub-base answers loading above, and a blinding carries nothing at all.

Mixing them up has consequences on site, because the material that suits one position does not necessarily suit another and the tests applied to each are different.

The formation entry is unusually direct about this: two excavations to the same depth can arrive at very different formations, and a formation can be spoiled by weather or trafficking while the subgrade beneath it is unchanged. It is a made surface rather than a natural one, and it is at its best at the moment it is first exposed.

The sequence between forming the level and covering it is where formations are commonly lost, because the surface degrades while other work is being organised. An unprotected formation can deteriorate to the point where the level has to be taken lower, which is a decision somebody has to be authorised to make.

What makes a fill engineered is the control over placement and compaction, not the material itself, and a fill is only as reliable as its worst-compacted layer. The entry is blunt about the common misreading: imported material does not become engineered fill simply because it has been rolled.

The same applies to the sub-base above it. An under-compacted granular layer settles under the load it was built to spread, and any softness or unevenness left in it tends to reappear in the slab as differential movement. Placement and compaction records are the substance of the claim that a layer was built to a regime rather than tipped into a hole.

A capping sits between a weak subgrade and the sub-base above it, introduced when the ground will not support construction of the sub-base directly, either because it is too soft to compact against or because loads would reach it too concentrated. It is a statement about the ground rather than about the pavement, and a build-up may contain both layers.

It is also not a cure. A capping reduces how concentrated loading reaches weak ground; it does not make weak ground strong, and the ground behaviour beneath remains a separate question for the geotechnical designer.

A filter and separation layer sits between materials of different grading and does two things at once: it lets water pass and it stops fine material working its way up into the voids of the coarse layer above. Filtration is a relationship between the two soils rather than a property of the layer alone.

Its failure is gradual and hidden. A drainage layer that has silted up looks unchanged from the surface but no longer drains, and a granular layer contaminated with fines no longer spreads load the way it did. Damage during placement of the layers above is the most common way separation is lost.

Two of the entries draw the line explicitly. The strip foundation entry says concrete strength is not a substitute for a sound formation. The blinding entry says a blinding is not structural, is not designed or reinforced as though it were, and that treating it as though it were misreads its purpose.

The sub-base entry adds a third: deeper is not automatically a remedy. What the ground beneath can accept still governs, and additional depth is a design decision rather than a default response to soft ground.

Settlement showing at a surface often originates in the layers beneath it rather than in the surface itself, and loading that exceeds what a build-up was constructed for tends to appear as progressive deformation rather than sudden failure. That is why the evidence arrives slowly and in the finished floor.

Movement in the concrete itself is a separate mechanism running alongside it. A slab shrinks as it cures and moves with temperature afterwards, and where joints are placed, and whether the perimeter lets the slab move at all, determines whether that movement appears in planned lines or as random cracking through the finish. Telling the two apart is assessment work rather than something to read off a crack.

Barrier continuity, insulation joints and service positions cannot be revised once concrete is placed, which puts unusual weight on inspection beforehand rather than remediation afterwards. The same is true of every layer discussed above.

That makes the record the only route back. What was found at formation, what was placed, how it was compacted and what testing was carried out are the evidence a later engineer will have to work from, and their absence is itself a constraint on the building.