The argument starts at about the third floor of most mid-rise sites in Metro Manila. The schedule says forms off at 7 days and props out at 14, because it said so on the last three projects. The design engineer walks two floors below the pour, sees a 14-day-old slab sagging the morning after its props came out, and asks who checked it. The answer is usually that the fresh slab's weight "goes down the props to the columns". It does not. Props bear on slabs, and the slab working hardest is rarely the one directly under the pour. This is a construction-stage load-path check, not member design, argued from arithmetic with every fraction shown rather than from any code clause.
A fresh slab's weight passes through the shores into the slab below and, if that one is still shored, into the next slab down. The interconnected slabs carry it together, and it reaches the columns only through their own bending. Because props come out from the bottom up, the most heavily loaded slab is typically two levels below the pour, at the moment its own shores have just been pulled and a new floor lands on top. Calendar days are not strength; 7 and 14 are proxies that depend on cement, temperature and curing. And laboratory-cured cylinders describe the mix, not the slab baking on the fourth floor; field-cured cylinders or a maturity meter describe that.
In words, without section numbers: NSCP 2015 and ACI 318 require that forms and shores be removed without damaging the structure, and that the concrete exposed by removal be able to carry its own weight and the loads placed on it. ACI 318 also expects the contractor to have a removal and reshoring procedure and schedule, with the structural analysis and strength data available on request. Neither code fixes a calendar stripping age for slabs (some project specifications do); 7 and 14 days are common practice, not code mandates. Neither code prescribes how that analysis is done, but a trace like the one below is the core of what the engineer of record will want to see, alongside strength data from the actual slab.
The procedure traces back to Grundy and Kabaila (1963) and is carried in ACI's shoring and reshoring guide for multistory buildings (ACI 347.2R) and in ACI SP-4, Formwork for Concrete. Assumptions: shores and reshores are rigid relative to the slabs; every slab has the same stiffness regardless of age; the ground is rigid; and load applied through props is shared equally among the slabs they interconnect. Reshores are installed snug, carry nothing at installation, and pick up load only from later operations. It is an approximation: methods that model the lower stiffness of young concrete (Liu, Chen and Bowman, for example) redistribute load between slabs, so the simplified answer is not conservative for every slab, and it ignores column shortening and formwork stiffness.
Six-storey residential building in Quezon City, 150 mm reinforced concrete slab, f'c = 28 MPa, 600 m² floor plate, one floor every 7 days. Slab self-weight = 0.15 m × 24 kN/m³ = 3.6 kPa. Formwork is taken at 0.5 kPa, a common-practice allowance. D = 3.6 + 0.5 = 4.1 kPa; keeping the formwork inside D for every level, even stripped ones, is a conservative simplification. Construction live load on the deck being cast is 2.4 kPa, the minimum ACI 347R recommends for horizontal formwork where motorized carts are not used, with a combined dead-plus-live minimum of 4.8 kPa for that case. These are guide recommendations, not NSCP mandates, and they govern the formwork and its supporting shores rather than the strength check of the permanent slabs underneath; carrying 2.4 kPa into the slab check is a deliberate borrowing. Construction live load ought to be considered on the lower levels too, which this example applies nowhere.
Loads in units of D. S1 is cast at day 0, S2 at day 7, S3 at day 14, S4 at day 21. The shores under a slab come out just before the slab two levels above it is cast.
The governing case is S2 = 2.25 D at the casting of S4, when S2 is 14 days old: 2.25 × 4.1 = 9.225 kPa. The 2.4 kPa live load on S4's deck passes through the shores into S3 and is shared with S2, so S2 takes 2.4 ÷ 2 = 1.2 kPa. Peak construction load on S2 = 9.225 + 1.2 = 10.425 kPa.
Superimposed dead load 1.5 kPa (finish, ceiling, services; illustrative). Live load 1.9 kPa, the NSCP 2015 minimum live load table value for residential basic floor area. Service load = 3.6 + 1.5 + 1.9 = 7.0 kPa. Factored load = 1.2 × (3.6 + 1.5) + 1.6 × 1.9 = 1.2 × 5.1 + 3.04 = 6.12 + 3.04 = 9.16 kPa.
Screening: 10.425 ÷ 7.0 = 1.489 of the service load and 10.425 ÷ 9.16 = 1.138 of the factored design load, on a 14-day-old slab. With an illustrative field-cured cylinder result of 88 percent of f'c at 14 days, 0.88 × 28 = 24.6 MPa. Flexural capacity barely moves with f'c, but the concrete terms for one-way and punching shear scale roughly with √f'c, so √0.88 = 0.938, about 6 percent down; the modulus of elasticity drops similarly, which is why the slab sags. An unfactored construction load above the factored design load fails screening outright, and the early deflection is locked in by creep.
The order inside each cycle matters, so state it: (1) pull the reshores under the lowest reshored slab; (2) strip the shores under the slab cast one cycle earlier; (3) reshore that slab snug with the props recovered in (1); (4) cast the next slab. Crews follow this bottom-up order because they need the recovered props.
Peak = 1.75 × 4.1 = 7.175 kPa. The slab above carries stripping crews and stacked forms during that operation, so keeping the same 1.2 kPa share is a conservative pairing: 7.175 + 1.2 = 8.375 kPa. Factored like the design load, 1.2 × 7.175 + 1.6 × 1.2 = 8.61 + 1.92 = 10.53 kPa, so 10.53 ÷ 9.16 = 1.150. That is the Scheme 2 result: the reshores cut the peak hard, but the slab still needs roughly 15 percent of flexural reserve, so Scheme 2 does not clear the check outright. The unfactored 8.375 ÷ 9.16 = 0.914 is a screening number only and must not be read as a pass, since it sets an unfactored construction load against a factored one. If the crew strips the shores above before pulling the reshores below, the same idealization gives only 1.5 D, which is why the written sequence matters as much as the number of props.
Comparing an unfactored construction load with a factored design load is only a screening step. The proper check factors the construction load and compares it with the design strength, phi times nominal capacity, computed with f'c at that age. So be clear about the 1.150 above: 9.16 kPa is a factored demand, not a design strength, so the ratio is a proxy for that check and assumes zero reserve. Whether Scheme 2 passes depends on the reserve in the actual bar schedule; a 150 mm residential slab governed by minimum steel or deflection usually has some, but the engineer of record must confirm it against phi times nominal capacity at the age strength. Note too that 1.2 and 1.6 are the in-service combination factors; where the project specifies the construction-loads standard ASCE/SEI 37, use its combinations, which assign their own factors to construction dead and variable loads and are not 1.2 and 1.6 across the board.
If the reserve is not there, add a second level of reshores and re-trace, because that is the only one of the three usual fixes that changes the trace. The Grundy-Kabaila ratios contain no time variable, so a 10-day cycle returns the identical 2.25 and 1.75; a longer cycle moves the capacity side alone, since the slab is stronger at the instant of peak load. Backpropping moves neither peak.
The numbers above belong to this cycle, and to assumptions the arithmetic never states.
One extra level of reshores for a 600 m² plate at an illustrative rental of ₱45 per m² per month costs 45 × 600 = ₱27,000 per month, or 2 × 27,000 = ₱54,000 over a two-month superstructure. Stretching the cycle from 7 to 10 days on five typical floors instead adds 5 × 3 = 15 days of site overhead; at an illustrative ₱25,000 per day, 15 × 25,000 = ₱375,000. On cost alone the reshores win by a wide margin, at roughly a seventh of the price of the slower cycle, and spare the finishing trades a cracked ceiling. Cheaper is not adequate, though: Scheme 2 still needs about 15 percent of flexural reserve, and only the capacity check at the age strength settles that. Both figures multiply the same floor-plate area you take off for the deck forms; the formwork take-off in RHCES Estimator gives that area per floor, so prop density and reshore rental come off the same quantity sheet as the formwork.
Vertical faces carry no load once the concrete has stiffened, so early side stripping is common practice and mainly a surface-damage question. The forms are also a curing aid, though: the concrete has to be kept moist and above roughly 10 degrees C for at least 7 days, or 3 days for high-early-strength concrete, so stripping beam sides at 3 days is acceptable only if curing by another method starts immediately on the exposed faces. A beam soffit at 7 days only makes sense if the beam is reshored immediately or the trace shows it can carry what sits above it at 7-day strength; beams are part of the same interconnected stack as the slabs.
Substantially. A post-tensioned slab is normally stripped after stressing, when the tendons balance much of its self-weight, so it often supports the floor above with fewer levels of shores. But stressing needs a minimum concrete strength at the anchorages, and the stressing sequence becomes the critical item instead of the stripping age. Rewrite D and the sequence for the stressed condition, and take the stripping logic from the PT designer.
Backpropping keeps the rigid-ground assumption alive for those floors: a slab still propped to ground carries nothing extra. Once the lowest props are pulled, the trace above restarts from that floor and the peaks are the same. Backpropping delays the problem; reshores reduce it.