Testing standard
Standard Test Method for Splitting Tensile Strength of Cylindrical Concrete Specimens
Written and technically reviewed by Dak System Inc. engineeringLast reviewed
ASTM C496 loads a concrete cylinder across its diameter until it splits, giving an indirect measure of tensile strength. Note its status before citing it: C496/C496M-17 was WITHDRAWN in 2026 and the ASTM catalogue names no replacement. IS 5816 and EN 12390-6 cover the same test and remain current.
A concrete cylinder — a moulded specimen or a drilled core — is laid on its side between the platens of a compression machine and loaded along a diametral line down its length. Narrow bearing strips run along the top and bottom contact lines so the load is distributed rather than concentrated on a knife edge.
Compressing a cylinder across its diameter puts the material in the middle of the cylinder into tension at right angles to the load. The specimen splits down that diametral plane, and the load at which it does gives the splitting tensile strength.
The load is applied continuously and without shock at a constant rate within the range 100 to 200 psi/min, which is 0.7 to 1.4 MPa/min of splitting tensile stress, until the specimen fails.
Concrete is roughly ten times stronger in compression than in tension, and its tensile strength is difficult to measure directly because gripping a concrete specimen to pull it apart is impractical. The splitting test is the standard indirect route: it produces a tensile failure using a compression machine.
The result is used in the design of structural lightweight concrete members to evaluate the shear resistance provided by the concrete and to determine the development length of reinforcement. It is also used to compare mixes, to assess aggregate quality and to check curing.
Splitting tensile strength is not direct tensile strength and is not modulus of rupture. All three describe the tensile behaviour of concrete and all three give different numbers; a design that names one is not satisfied by another.
The bearing strips are as much a part of the specimen as the concrete is. They are what turns a line contact into a distributed load.
2P / (π L D)
The classical Brazilian-test expression. It assumes an elastic, homogeneous cylinder loaded along a line.
Diametral compression produces transverse tension along the loaded plane
Concrete is roughly ten times stronger in compression than in tension, so it fails in tension across the diameter long before it crushes.
Direct tension, splitting tension and modulus of rupture all differ
A design naming one is not satisfied by another. Modulus of rupture is the highest of the three and direct tension the lowest.
Flat, parallel platens with an alignment arrangement that holds the cylinder and both bearing strips square while the platen closes. Most of the operator variability in this test comes from alignment, and a jig removes it.
SpecificationsA 150 by 300 mm cylinder splits at around 200 kN and higher-strength mixes take more, so the frame is sized for that. What matters as much is that the machine can hold the slow, steady ramp the stress rate demands.
SpecificationsA 150 mm by 300 mm cylinder splitting at typical concrete strengths demands on the order of 200 kN, and higher-strength mixes take more, so a compression frame of 300 kN upwards with force accuracy to ASTM E4 is the practical requirement.
Platen parallelism and a means of aligning the cylinder along the platen centreline matter more than headline capacity. An alignment jig that holds the cylinder and the bearing strips square while the platen closes removes most of the operator variability from this test.
Loading has to be controllable at a low, steady rate: the specified stress rate corresponds to a slow force ramp, and a machine that can only be driven quickly at the bottom of its range cannot meet it. No extensometer or strain measurement is involved; the result is a failure load.
Misalignment is the dominant fault and it lowers the result. Reused or wrongly sized bearing strips concentrate the load. Loading faster than the specified rate raises the apparent strength, as it does on every concrete test. And a specimen that fails on a plane other than the marked diametral one has not measured what the method defines.
All three are in use, all three give different numbers, and none of them is the direct tensile strength of the material.
| Splitting (C496) | Flexural (C78) | Compressive (C39) | |
|---|---|---|---|
| What is measured | Indirect tension across a diameter | Extreme fibre stress in bending | Compression |
| Specimen | A cylinder on its side | A beam | A cylinder upright |
| Relative magnitude | Lower than modulus of rupture | The highest of the tensile measures | About ten times the tensile values |
| Used for | Shear resistance, development length | Pavement design | Almost everything else |
| Status | Withdrawn 2026 | Current | Current |
Empirical relationships between these three exist and are mix-specific. Converting a compressive strength into an assumed splitting strength is a rough estimate, not a substitute for the test.
No. C496/C496M-17 was withdrawn in 2026 and the ASTM catalogue status line reads Standard Withdrawn, No replacement. The entry is kept here because the designation is written into a great many specifications, mix designs and quality manuals, and because the test itself remains in routine use under other documents.
The same test is covered by IS 5816 in Indian practice and by EN 12390-6 in Europe, and both remain current. Where a contract specification still names ASTM C496 the sensible course is to run the test to the withdrawn edition, state clearly in the report that the standard is withdrawn and which edition was followed, and raise the substitution with the specifier rather than making it unilaterally.
Because concrete cannot practically be gripped and pulled. Loading a cylinder across its diameter puts the material along that diametral plane into transverse tension, and since concrete is roughly ten times stronger in compression than in tension, it splits along that plane long before it crushes. It is an indirect measurement, and the strength it gives is not the direct tensile strength — but it is repeatable, it uses an ordinary compression machine, and it uses the same cylinders already being cast.
To spread the load along the contact line. Without them the platen bears on the cylinder along a knife edge, the local stress is enormous and the specimen crushes at the contact instead of splitting. The strips are specified in material and dimensions and are single-use: a strip that has already been indented by one test no longer distributes evenly, and the crack starts wherever it has thinned.
They are different numbers for a related property, and modulus of rupture is the higher of the two. Modulus of rupture comes from beam theory applied to a flexural test and assumes concrete behaves elastically to failure, which it does not, so it overstates the true tensile strength by more than the splitting test does. A design that specifies one is not satisfied by the other, and pavement work in particular is specified on flexural strength.
The apparent strength goes up. This is true of every strength test on concrete and it is why the rate band is specified narrowly — 0.7 to 1.4 MPa/min of splitting tensile stress. On a large compression frame that is a slow force ramp, and a machine whose control is poor at the bottom of its range will run faster than the method allows without anyone noticing.
A compression frame of 300 kN upwards with force accuracy to ASTM E4. A 150 by 300 mm cylinder splits at roughly 200 kN at ordinary strengths and higher-strength mixes take more. Platen parallelism and a means of aligning the cylinder and its bearing strips along the platen centreline matter more than extra capacity, because misalignment is the dominant source of low results.
Dak verifies against whichever standard the method names, and where a class applies our frames sit a class tighter than it asks.
| The method asks for | Dak supplies | |
|---|---|---|
| Capacity | High — around 200 kN for a 150 by 300 mm cylinder at ordinary strengths, more on high-strength mixes | Load cells from 1 kg to 60 ton on the Series 7200, and 0.5 to 100 kN on the Series 9000 |
| Force accuracy | ASTM E4 | Verified to ASTM E4, and to ISO 7500-1 Class 0.5 |
| Gripping | Flat parallel compression platens with an alignment arrangement and single-use bearing strips | Our compression anvils, built to the specimen |
| Environment | Cured as the project requires; the test itself is at ambient laboratory conditions | 3009 series chambers, −150 °C to +400 °C — temperature only |
This page describes the method as practised. The governing text is the current edition from the issuing body. Tell us what you are testing and we will answer with the machine, the fixture and a quotation.
