The Master Guide to Pouring a Professional Concrete Slab

A week out: the subgrade you will never see again

A slab fails from the bottom up, and the failure is always older than the crack that reveals it. Before any concrete is ordered, the ground beneath has to be brought to a uniform bearing condition, and uniform matters more than hard. A subgrade that is consistently firm across the whole footprint carries a slab better than one that is rock under three quarters of it and spongy in one corner. Concrete does not object to settling. It objects to settling unevenly.

Compaction happens in lifts or it does not happen at all. Granular fill dumped to full depth in a single drop will not densify through its thickness however many passes the plate makes across the top, so fill goes in at the lift thickness the geotechnical report or the local authority specifies, each lift compacted before the next arrives. Moisture governs how well the material responds; bone-dry sand and saturated clay both shrug off a compactor, for opposite reasons.

Proof-roll before the formwork goes up. Once forms and steel are set, the subgrade becomes invisible and stays invisible for the life of the building. Walk every square metre of it, run a loaded barrow or a truck wheel across, and watch for pumping, rutting, or a patch that simply feels different underfoot.

Softness found now costs an hour with a mattock and a barrow of compacted stone. Softness found later is found by the slab itself: the concrete bridges the void, hogs over the firm ground either side, and cracks, often within the first month, sometimes not until the first hard winter, and always in a position that reads as random until somebody cores it. No repair available at that stage is cheaper than the hour would have been.

1. Strip topsoil and organics down to competent material across the full footprint, not just under the walls.
2. Place fill in lifts at the thickness the geotechnical report or local authority specifies, compacting each lift before the next.
3. Check fill moisture; dry material will not bond and saturated material will not densify.
4. Proof-roll with a loaded barrow or truck wheel and mark anything that pumps, ruts, or feels soft.
5. Dig out and replace soft spots with compacted stone before a single form board is set.

The day before: forms, grade, and the last honest measurement

Forms get set to a datum, never to the existing ground surface. Ground undulates; a laser or a dumpy level does not. Shoot the high point of the subgrade first, because that point sets the minimum top-of-slab elevation unless there is appetite to dig, and everything else falls out of it: form height, screed rail height, the fall to the door threshold or the gully.

Edge form stiffness is worth more attention than it usually gets. Stakes at generous spacing let the form belly outward under the weight of wet concrete, and a bellied form is not a straight edge that can be corrected later; it is a permanent line that every eye entering the building follows. Brace externally, kick the stakes, and check the string line after the last stake is driven rather than before.

Depth checks come last and they cost real money. Over-excavation is a silent invoice: twenty-five millimetres of extra depth across a fifty square metre slab is an extra one and a quarter cubic metres of concrete, paid for at full rate and never seen again. Take gauge stick readings on a grid, not just at the corners, and bring low spots up with compacted stone rather than with the concrete order.

1. Establish a datum and shoot the high point of the subgrade before setting any form height.
2. Set and brace edge forms, then re-check the string line and the diagonals after the last stake goes in.
3. Gauge slab depth on a grid and correct low areas with compacted stone.
4. Set screed rails or establish wet-screed positions and confirm falls to thresholds and drainage.

Under the steel: the vapour retarder and the argument about it

The vapour retarder goes down after the subgrade is signed off and before any steel is placed. Sheets are lapped generously, laps taped, penetrations sealed around pipes and column boxes, and the membrane turned up at the perimeter so that the slab sits in a tray rather than on a mat. Material classes for these membranes are set out in the ASTM specification covering plastic water vapour retarders used under concrete slabs, and the specified class is not a detail to substitute on site.

Placing concrete directly on the retarder remains the live argument on many jobs. With a membrane immediately beneath, no bleed water can escape downward, so every drop of it has to rise; the slab bleeds longer, the finishing crew waits longer, and the moisture gradient through the slab thickness raises the risk of edge curling. Interposing a granular blotter layer between the membrane and the concrete shortens the bleeding and reduces curl.

Against that, a blotter layer is a reservoir. Rain on it before the pour, or a wet fill placed in autumn, and it holds water directly under a slab that will later be closed off with a floor covering, feeding vapour upward for months and lifting adhesives. Where anything moisture-sensitive is going down over the slab, current guidance in the ACI floor and slab construction document points toward placing directly on the retarder and managing the finishing consequences. Where the slab stays bare and curl control dominates, the blotter argument gets made. The specification decides; the crew does not decide it on the morning of the pour.

Reinforcement: chairs, laps, and why hooked-up mesh does not stay up

Steel does not stop a slab cracking. It holds the crack tight once it happens, and it can only do that from the correct position through the slab thickness. A bar or a sheet lying on the membrane is doing almost nothing, and a bar riding too high has lost its cover and will rust its way out through the surface within a few winters.

Chairs carry the steel, at spacing close enough that the bar does not sag between them under the weight of a boot. Sheet mesh sits on chairs; rolled mesh fights back and never lies flat enough to trust. The old habit of pulling mesh up with a hook as the concrete goes past is the one to abandon. It lifts the sheet at the point of the hook and nowhere else, the crew walks it straight back down two minutes later, and the finished slab has reinforcement that wanders between the top and the bottom of the section with no record of where it went.

Cover and lap length are code matters. Minimum cover varies with exposure, with the member, and with the edition of the code in force, and ACI 318 governs it for structural work; the drawing takes precedence over any yard rule of thumb. Development and lap lengths depend on bar size, concrete strength, cover, and bar spacing, so a single memorised multiplier is not a substitute for the schedule on the drawing. Mesh sheets are lapped and tied so that the lap acts as a lap and not as two sheets sitting near each other.

Booking the truck: volume, slump, and the weather window

Volume gets calculated from the formed dimensions and then adjusted upward. Subgrade tolerance, form deflection, and spillage all eat concrete, and running short mid-pour is far more expensive than a small over-order, because a cold joint through the middle of a slab is permanent. A waste allowance in the region of five to ten per cent is common practice, tightened when the subgrade has been laser-screeded and loosened when it has not.

Slump belongs on the ticket and gets verified on site. The test method is ASTM C143, and the reason for running it is not paperwork. Water added at the truck to make placing easier raises the water-cement ratio, and that single act reduces compressive strength, increases drying shrinkage, weakens the wearing surface, and widens every crack the slab was going to get anyway. Any water addition on site is governed by the ready-mixed concrete specification and the mix design, and it is not a decision for whoever happens to be holding the chute.

Weather is the other half of the booking. Concrete temperature, ambient temperature, relative humidity, and wind speed together set the rate at which the surface loses water, and it is wind that catches crews out; a warm still day is far more forgiving than a cool day with a stiff breeze across an open slab. High evaporation rates produce plastic shrinkage cracking in the first hours, before the concrete has any strength to resist it. Hot conditions push the start time earlier; cold conditions demand that the ground is not frozen and that the slab is protected before nightfall.

Crew size decides the pour rate, not the pump. Concrete arriving faster than the finishing crew can close it is the most common scheduling error on a slab, and it turns the back half of the pour into a scramble against a surface that is already stiffening.

Pour day, first hour: placing, consolidating, screeding

Discharge as close to final position as the access allows. Concrete dragged any distance with a rake segregates: the mortar travels, the coarse aggregate stays, and the slab ends up with a weak, sandy zone that will craze and dust later. Move the truck or move the pump line rather than moving the concrete.

Placing runs against a live edge. Each load goes against the face of the last, in a continuous advancing front, so nothing sits long enough to start setting before its neighbour arrives. A slab placed in scattered heaps and joined afterwards has cold joints running through it that nobody will find until they show as cracks.

Consolidation gets the air out and the concrete around the steel and into the form corners. Vibrate deliberately rather than dragging a poker sideways to move material, which is another route to segregation. Screed immediately behind placing, off rails or wet screeds, and screed once; repeated passes bring fines and water to the surface and build the weak layer that finishing problems come from.

Bull float straight after the screed, while the surface is still plastic and before bleed water has appeared. One or two passes, and then the crew stops and waits. Stopping is the hard part.

1. Discharge close to final position; never rake concrete across the slab.
2. Keep a live edge and place continuously against the previous load.
3. Consolidate around steel, edges, and penetrations without using the poker to move material.
4. Screed once, immediately behind placing.
5. Bull float straight away, before any bleed water shows, then stop.

Bleed water and the finishing window that moves

Troweling too early is the single most common finishing error on a concrete slab, and it is the one with the most expensive consequences. While bleed water is on the surface or a wet sheen still shows, any tool that touches it works that water back down into the top few millimetres. The result is a surface layer with a far higher water-cement ratio than the slab beneath, which then dusts under traffic, scales in frost, crazes into a mosaic of fine cracks, or blisters where a sealed skin has trapped rising air and water. Sprinkling dry cement to dry the surface, or a hose to lubricate the trowel, does the same damage by another route.

Bleed water goes away by evaporating or by being dragged off with a hose or a squeegee, never by being troweled in. The classic readiness check still works: when the sheen has gone and a boot leaves an impression of a few millimetres rather than sinking, the surface will take a float. Beyond that point the operations follow the concrete rather than the clock.

Temperature and wind move the whole window, and they can move it in both directions. On a hot, breezy, low-humidity day the surface may show no bleed water at all and crust within minutes while the body of the slab is still soft, which produces blisters and a delaminated skin if the crew reads the dry surface as readiness. On a cool damp day the same mix will bleed for hours, the crew stands around, and finishing runs past dark under lights. Neither situation is unusual, and neither is solved by working to the time the last slab took.

Broom finishes close earlier than hard-troweled ones and are the safer choice for external slabs, where a burnished surface is a slip hazard and holds less air entrainment near the top. Hard troweling is a sequence of passes with progressively tighter blade angles, each one waiting on the concrete, not on the shift ending.

Four to twelve hours: cutting the joints before the slab cuts its own

The saw is the last operation of the pour, and its timing has almost no slack in it. Cutting is possible once the surface has hardened enough that the blade does not ravel the edges of the cut, and it must happen before drying shrinkage stress exceeds the tensile strength of the young concrete. With a conventional wet saw that window commonly falls somewhere in the first four to twelve hours after finishing; early-entry equipment is designed to go in far sooner, sometimes within an hour or two. Set time, mix, and ambient temperature move it, so somebody has to be on site watching rather than booking a slot.

Depth matters as much as timing. A joint cut too shallow does not create a plane of weakness strong enough to attract the crack, and the slab cracks somewhere else instead, usually a few hundred millimetres away and running at an angle. The conventional working rule for a wet-sawn joint is a depth of about one quarter of the slab thickness; early-entry systems are cut shallower by design, and the project specification governs which applies. Spacing follows the rule of thumb of roughly twenty-four to thirty-six times the slab thickness in consistent units, adjusted for mix shrinkage and reinforcement, with panels kept close to square. Long thin panels and re-entrant corners crack regardless of how well the rest is laid out, so those get jointed deliberately.

Cutting late is not a partial success; it is a total failure of the operation. Concrete that has already relieved its shrinkage stress through a random crack has nothing left to give the saw cut, and the cut becomes decoration around a crack that now runs across the floor and through the joint line. Curing continues around the sawing: the cure is applied as soon as the finish closes, interrupted only for the length of the cut, and restored immediately after.

Days one to seven: curing, the step most often skipped

Curing is the step most often skipped, and it gets skipped because nothing visible happens on the day it is skipped. Cement hydration needs water. A slab that dries out at the surface stops gaining strength exactly in the top layer that carries abrasion, impact, and freeze-thaw, so an uncured slab can test acceptably on cylinders taken from the truck and still dust, scale, and wear badly within a year.

Methods vary and any of them beat nothing: ponding or continuous wetting, wet hessian kept genuinely wet rather than allowed to dry and act as a wick, plastic sheeting lapped and weighted down at the edges, or a membrane-forming curing compound applied at the specified coverage rate. Sheeting can mottle the surface where it touches unevenly, which matters on exposed floors. Curing compounds have to be compatible with whatever coating or adhesive comes later, and a compound applied under a floor covering that needs a clean substrate is a problem bought for the tiling contractor.

Duration is set by code and by the mix rather than by convenience. Minimum curing periods are given in ACI 318 and set out in detail in the ACI specification for curing concrete, and they are commonly on the order of a week for ordinary mixes and shorter for high-early-strength concrete, with the applicable edition and the local building control having the final say. Protection from freezing during that period is separate from and additional to curing. Leaving side forms in place for the first days protects the arrises, which are the first thing to chip and the last thing anyone wants to patch.

1. Start curing as soon as the finish is closed, not at the end of the shift.
2. Choose a method compatible with the finish and with any covering that follows.
3. Keep wet coverings genuinely wet; a dry hessian sheet draws water out of the slab.
4. Maintain the cure for the period the specification and the code in force require, and protect against freezing separately.
5. Leave side forms on through the early days to protect the edges.

The month after: strength, movement, and what the slab tells you

Twenty-eight days is the conventional age for the specified compressive strength, and cylinders or cubes made and cured in accordance with the relevant ASTM field practice are what settle any argument about the concrete supplied. Delivery tickets, slump and temperature results, and a note of when curing started and stopped are worth keeping in one folder; they are the only evidence that exists if a floor is questioned two years later.

Loading matters through this period. A slab reaching design strength at twenty-eight days is not at design strength when the scaffold tower goes up on day four, and racking, wheeled plant, and stacked material placed early leave permanent marks or cracks in a slab that would have carried them comfortably a month on.

Movement shows up in the first weeks and then slows. Joints open as the slab shrinks, edges curl slightly upward at panel corners where the top dries faster than the bottom, and fine surface crazing may appear on floated finishes. Filling joints with a semi-rigid material is usually deferred until most drying shrinkage has occurred, which is a matter of months rather than weeks, because a joint filled while it is still opening will simply split or debond. Cracks that appear in this window are worth recording with a date and a width; the pattern tells you whether you are looking at plastic shrinkage from the first hours, drying shrinkage that the joints failed to control, or a subgrade problem that started the week before the pour.