Why Carry Out a Floor Load Test?

It is done when calculation alone cannot show whether an existing floor can carry a new load. In renovation work or new machinery installation, the most common question is "Can the existing floor carry the machinery or the new use?" But the data needed for calculation is often incomplete, for example:

  • No original construction (as-built) drawings, or incomplete drawings
  • The actual concrete strength and amount of reinforcement are unknown
  • The building is old, and materials may have deteriorated from their design condition
  • The new use imposes more load than the original design

When there are many unknowns, analytical evaluation alone may not be enough. A Load Test tests the real structure under real conditions, so the result already reflects the materials, construction quality and deterioration. If you are not sure what floor system you have or where to start checking, read the overview at how much load can a factory floor carry

📌 Tip

In practice it is often done together with a Concrete Core Test (concrete coring per ASTM C42 / C39): the Core Test confirms material quality, while the Load Test confirms the behavior of the whole slab. Both methods are part of structural inspection (NDE)

Which Standard Does a Floor Load Test Follow?

It follows ACI 318-19 Chapter 27, Strength Evaluation of Existing Structures by load test, which divides the test into 2 methods:

← Scroll to see the full table →

Test methodCharacteristicsBest for
Monotonic Load Test (27.5)Load is increased in one direction up to the maximum, then held ≥ 24 hrsWhere shear failure or development of reinforcement is a concern — the sustained load allows cracking/creep to show
Cyclic Load Test (27.6, per ACI 437.2)Load is applied and removed in cycles at increasing levelsWhen you need to check linear-elastic behavior and the permanency of deflection in each cycle

For typical factory floors, the Monotonic method with water ponding is the most popular, because water gives a uniform load that is easy to control, safe, and can be calculated directly from the water level: 10 cm of water equals 100 kg/m²

⚠️ Key requirement (27.4.4)

The structure to be tested must be at least 56 days old, unless the owner, contractor and design engineer jointly agree to test earlier.

How Is the Total Test Load (Tt) Calculated?

Use the largest value from the 3 equations in ACI 318-19 Section 27.4.6.2; the total test load Tt includes the self-weight of the existing floor.

(a)   Tt = 1.0Dw + 1.1Ds + 1.6L + 0.5(Lr or S or R)
(b)   Tt = 1.0Dw + 1.1Ds + 1.0L + 1.6(Lr or S or R)
(c)   Tt = 1.3(Dw + Ds) Dw = dead load (slab self-weight) · Ds = superimposed dead load · L = live load

Worked example — RC slab 120 mm thick, service live load L = 300 kg/m²

← Scroll to see the full table →

ItemValueSource
Dead load Dw288 kg/m²0.12 m × 2,400 kg/m³
Tt (governing eq. a)768 kg/m²1.0(288) + 1.6(300)
Load actually applied480 kg/m²Tt − Dw (the slab already carries itself)
Converted to water depth48 cm480 ÷ 1,000 kg/m³
📌 A common mistake

Load actually applied = Tt − load already on the floor, not the full Tt, because the slab itself already counts as part of the test load. The value of L and the load items used in the calculation must be set by the responsible engineer for each project.

How Is a Monotonic Water Ponding Test Carried Out?

The test has 5 steps, from preparing the water bund to measuring recovery. Site time is typically 2–3 days per test location.

1. Prepare the water bund

Build the bund with plywood walls braced by scaffolding or steel joists to resist water pressure, lined with a waterproof membrane. The bund should cover the critical region of the slab panel to be tested. If water ponding is not possible on site, for example on an upper floor or where there is no drainage, see alternatives and how to arrange the load so results are not distorted at Floor Load Test loading methods: water ponding vs IBC tanks vs sandbags

3D model of a 4 by 4 meter water bund for a Floor Load Test
Model of a 4×4 m water bund with plywood walls braced by steel joists against water pressure and lined with a waterproof membrane
⚠️ Place the load on the critical region (27.4.6.1)

The load must be arranged to produce the maximum load effect in the critical region. For a single-span slab, the critical point is the midspan moment, so a patch load centered on the span gives a result very close to loading the full span (in the case study below, a 4.0 m patch on a 4.75 m clear span gives ~97.5% of the full-load midspan moment).

2. Install the instruments

  • Deflection gauges (dial gauges) under the slab soffit at no fewer than 3 points, with midspan as the critical point
  • Gauge stands must be independent of the structure carrying the test load, otherwise readings will be distorted
  • Take the datum (zero reference) reading no more than 1 hr before applying the first load increment (27.5.2.2)
Dial gauges installed under the slab soffit to measure deflection in a Floor Load Test
Dial gauges installed under the slab soffit to measure deflection, with stands independent of the structure carrying the test load

The 3D view below cuts the slab panel in half so both sides can be seen at once: on top is the water bund straddling midspan, and below are the dial gauges on stands set up from the floor below, not touching the slab under test.

3D model of a water-ponding Floor Load Test, with the slab panel cut in half to show the water bund on top and the deflection gauges under the slab soffit A conceptual 3D view of one reinforced concrete slab panel supported on beams and columns, cut in half along midspan to show the cross-section. On top of the slab is a temporary 4.0 by 4.0 meter water bund with plywood walls braced by steel joists and lined with a waterproof membrane, holding water 48 centimeters deep. The bund wall has four water-level marks at 12, 24, 36 and 48 centimeters, equal to uniform loads of 120, 240, 360 and 480 kilograms per square meter. The section shows a slab 120 millimeters thick with a clear span of 4.75 meters. Under the slab soffit are three dial deflection gauges, with midspan as the critical point; each is mounted on a stand set up from the floor below, independent of the slab carrying the test load. An engineer stands on the floor below reading the gauges. An orange dashed line under the slab shows the exaggerated deflected shape. The frame on the right summarizes five elements of the test: the water bund, test load from the water level, deflection gauges at no fewer than three points, stands independent of the tested structure, and visual crack inspection throughout the test. The figures in the image are from the case study in this article. 3D model: water bund on slab + gauges below Not to scale Bund 4.0 × 4.0 m Plywood + steel joists + membrane 4 water levels: 12 · 24 · 36 · 48 cm 10 cm water = 100 kg/m² RC slab 120 mm thick Deflection Exaggerated Gauges under slab ≥ 3 points · midspan critical Free stand From floor below Not touching test slab Cordoned area Entry only for readings Engineer reads each step and checks cracks Slab cut in half to show soffit · conceptual view Test elements 1 Bund straddles critical zone Max moment at slab midspan 2 Load from water level ≥ 4 equal increments then hold full load ≥ 24 hrs 3 Deflection at ≥ 3 points Gauges under slab 4 Stands independent otherwise readings are off 5 Check cracks throughout Any distress: stop at once Case study figures Slab 120 mm · span 4.75 m Max water 48 cm = 480 kg/m² Max deflection 1.53 mm (limit 9.40) Result: passes ACI 318-19
3D model — the slab panel is cut in half to show the 4.0 × 4.0 m water bund on top and three dial gauges on independent stands under the slab soffit. Figures are from the case study in this article (conceptual view, not to scale).

3. Apply the load in increments

Per 27.5.1.1, the load must be applied in at least 4 roughly equal increments, for example water at 12 / 24 / 36 / 48 cm (25% of the test load each). At each increment, wait for readings to stabilize, record every deflection gauge and visually inspect for cracks.

4. Hold the full load for 24 hours

Once the full load is reached, hold it for no less than 24 hrs (27.5.1.3). Record readings at +30 min, +60 min, then every 1 hr, along with the temperature and actual water level.

5. Remove the load and measure recovery

After 24 hrs, remove the load promptly (27.5.1.4), then measure the residual deflection 24 hrs after load removal (27.5.2.4), unless the exception in 27.5.3.6 applies (see the acceptance criteria section).

What Are the Pass/Fail Criteria for a Load Test? (27.5.3)

The floor passes when no damage is found during the test and the deflection meets Section 27.5.3.5 or 27.5.3.6, where Δl is the maximum deflection and Δr is the residual deflection after load removal.

← Scroll to see the full table →

No.CriteriaMeaning
27.5.3.1No spalling / crushing of concreteVisual inspection throughout the test
27.5.3.2No cracks indicating impending shear failureDiagonal cracks that keep lengthening + widening = danger sign
27.5.3.5Δr ≤ Δl / 4Residual deflection must not exceed 1/4 of the maximum deflection
27.5.3.6If Δl ≤ the greater of 0.05 in. or ℓt²/(20,000h)27.5.3.5 can be waived outright, because the structure is so stiff that recovery cannot be distinguished from field measurement error
Formula most often used in reports (27.5.3.6):   Δl ≤ ℓt² / ( 20,000 × h ) ℓt = clear span, h = slab thickness (same units) — if the maximum deflection is below this value = immediate pass, with no need to wait to measure recovery

If the criteria are not met, a retest is still allowed under 27.5.3.7, starting no sooner than 72 hrs after the first load removal and using the criterion Δr ≤ Δ2/5 (27.5.3.8). The diagram below brings the 5-step loading sequence and the decision path together in one image.

2D diagram of the Monotonic floor load test loading sequence per ACI 318-19 Chapter 27 and the decision chart based on deflection The diagram has two parts. The upper part is a stepped graph of test load versus time. It starts with a zero-reference reading no more than one hour before the first load increment, then load is applied in at least four equal increments — in the example, 25, 50, 75 and 100 percent, equal to water levels of 12, 24, 36 and 48 centimeters. Once the full load is reached it is held for no less than 24 hours, with readings at 30 minutes, 60 minutes and then every hour. The load is then removed promptly, and the residual deflection is read 24 hours after removal. Below the load graph is a schematic deflection line showing the maximum deflection delta-l at the end of the hold period and the residual deflection delta-r after load removal. The lower part is the decision chart. Throughout the test there must be no spalling or crushing of concrete and no cracks indicating shear failure; if any are found, stop the test and treat it as a fail. The first question is whether the maximum deflection does not exceed the greater of 1.3 millimeters or the span squared divided by twenty thousand times the slab thickness. If yes, it passes without waiting to measure recovery. If no, check the second question: whether the residual deflection does not exceed one quarter of the maximum. If yes, it passes. If no, a retest is still allowed, starting no sooner than 72 hours after the first load removal, using the criterion that residual deflection does not exceed one fifth of the second-cycle maximum. Floor Load Test loading sequence and decision chart (2D) Not to scale A · Loading sequence Monotonic · ACI 318-19 Sec. 27.5 · time axis not to scale 25% · 12 cm 50% · 24 cm 75% · 36 cm 100% · 48 cm Applied load ≥ 4 equal increments Read + crack check (27.5.1.1) Hold full load ≥ 24 hrs Read at +30 min, +60 min then every 1 hr (27.5.1.3) Remove load promptly (27.5.1.4) Wait 24 hrs (27.5.2.4) Time Deflection Schematic Not actual Δl = max deflection (end of hold) Δr = residual deflection Zero reading ≤ 1 hr before start (27.5.2.2) B · Decision ACI 318-19 Sec. 27.5.3 · Δl, Δr at critical point Visual check throughout No concrete spalling / crushing No shear cracks (27.5.3.1–2) Δl ≤ greater of? 1.3 mm (0.05 in.) or ℓt² / (20,000 h) (27.5.3.6) Pass So stiff, no need to wait for recovery Fail Stop test at once Do not reuse even if derated Δr ≤ Δl / 4 ? Residual ≤ 1/4 of maximum (27.5.3.5) Pass Good recovery Still elastic Retest allowed (27.5.3.7) Start ≥ 72 hrs after first unloading · 2nd-cycle criterion Δr ≤ Δ2 / 5 (27.5.3.8) Yes Yes No No Damage found All numeric criteria as stated in article (ACI 318-19)
2D diagram — the upper part is the Monotonic loading sequence (zero reading, 4 increments, full load held 24 hrs, load removal, then residual reading); the lower part is the decision chart per ACI 318-19 Section 27.5.3. The deflection line is schematic, not actual values.

Case Study: Factory Basement Floor, Lat Krabang District

A factory in Lat Krabang District wanted to install utility systems for an injection machine on its existing 120 mm RC basement floor slab. SPN was engaged to evaluate it with a Core Test and a Floor Load Test.

Core Test results: compressive strength of 2 concrete cores ≈ 212–214 ksc (≈ 21 MPa), consistent quality, within a good range.

Floor Load Test: test load Tt = 768 kg/m² (governed by 1.0Dw + 1.6L, L = 300 kg/m²), water ponded in a 4.0 × 4.0 m bund to a maximum of 48 cm, applied in 4 increments, full load held 24 hrs, with 3 dial gauges under the slab soffit.

Actual water bund on a factory basement floor during a Floor Load Test
Actual water bund on the basement floor — a maximum water depth of 48 cm gives a uniform load of 480 kg/m², held for 24 hours
1.53 mm
Max deflection Δl
9.40 mm
Criterion 27.5.3.6
Pass
1.53 ≤ 9.40 mm

← Scroll to see the full table →

ItemValue
Max midspan deflection Δl1.53 mm (at full load +24 hrs)
Criterion 27.5.3.6 = ℓt²/(20,000h) = 4,750²/(20,000×120)9.40 mm
Result1.53 ≤ 9.40 mm → pass (recovery measurement can be waived)

The deflection behavior is also telling: from +6 to +24 hrs the value increased by only ~0.16 mm, showing that deflection had stabilized. Behavior was mainly elastic with limited creep, and no cracks were found throughout the test (Δl ≈ ℓ/3,100, very small compared with the common limit of ℓ/360).

✅ Conclusion

The floor has sufficient capacity to carry a live load of 300 kg/m² for the planned machinery installation.

See another case study with readings at every increment and a deflection–time graph at Case Study: Floor Load Test of a building floor before alteration, Thung Khru District

Summary

A Floor Load Test to ACI 318-19 Chapter 27 proves the strength of an actual structural floor with real load, and suits buildings without original drawings or undergoing a change of use. The essentials are calculating the test load Tt correctly, applying load in at least 4 increments, holding full load for 24 hours, and comparing deflection with the 27.5.3 criteria. Testing the bearing capacity of the "soil" beneath a slab-on-ground is a different test; use a Plate Bearing Test instead.

SPN Soil Engineering provides Floor Load Tests together with Core Tests nationwide, with complete reports including deflection graphs and photos. Reports are certified by a civil engineer licensed by the Council of Engineers Thailand (COE), with signature and registration number. Contact us for a free quotation within 24 hrs.