5 min read
Why the break pattern matters as much as the number
Compressive strength methods such as EN 12390-3, ASTM C39 and SANS 5863 describe not just how to apply load but also how the specimen should look when it fails. The pattern is your evidence that the cube was loaded the way the method intends. A cube that fails through an edge, or splits apart at low load, has usually measured a preparation or machine defect rather than the strength of the concrete.
The names used for the patterns differ slightly between methods, and some methods publish correction factors for particular modes. Treat those factors as a laboratory policy decision to be justified and documented, never as an automatic adjustment applied quietly to a result.
Satisfactory patterns
The two cones: the classic hourglass, with a cone forming from each loaded face and spalling around the middle. This is the pattern most methods treat as a normal, well-behaved failure and it is what a properly capped, square cube typically does.
Cone and shear: one face shows a cone, the other a diagonal shear plane. Also treated as a normal failure and common where the two load faces are not perfectly parallel — the cube finds the weaker path.
Columnar, or explosive, vertical splitting: the cube breaks into vertical columns with little or no spalling. This appears in high-strength concrete and in very stiff testing machines, and is generally accepted as a valid failure of a strong specimen.
Unsatisfactory patterns
Edge or corner splitting: the cube fails through a vertical plane near an edge, often at a load well below the expected strength. It usually points at a cap or a face that is not flat, a cube that was not square, or a platen that loaded one edge first.
Horizontal splitting or a cap that detaches: the top or bottom of the cube separates cleanly. Look immediately at the capping material, its thickness, its curing and the flatness of the face underneath — and at whether the cap was applied to a damp or dusty surface.
Failures that start at the platen contact with visible crushing or a ring of broken material around the loaded face suggest a platen that is worn, dished, or holding grit. A machine whose spherical seat is stiff will often produce the same signature across several cubes in a row, which is exactly why the pattern should be recorded per cube rather than summarised.
What to do when you see an unsatisfactory pattern
Do not silently discard the result and do not silently correct it. First, check the specimen: faces clean and flat, cap sound and within the method's permitted thickness, dimensions as recorded. Then check the machine: platen condition, seat freedom, alignment of the specimen under the centre of the platen.
If a defect is confirmed, the result is normally void and a replacement specimen — from the same set if one is available, or a repeat sample agreed with the client — is the honest route. Document what you found, what you did and who authorised it. If the result is reported anyway, it should carry the pattern and a note, because a reviewer or an auditor will ask about it later.
Record it, don't guess it
The failure pattern belongs in the record next to the failure load, the dimensions and the age. Photograph the broken cube, note which pattern you judged it to be, and keep the photo with the result. Patterns only become useful evidence when they are captured at the time; a description reconstructed a month later, after a report has been queried, carries very little weight.
Software helps here for one reason: it makes the pattern a first-class field, so trends become visible. A lab that records patterns can see that a particular platen or a particular capping batch is producing edge failures long before anyone thinks to investigate.
Try Simplecrete free for 1 month
The full version, on your own specimens and presses: QR labels, automatic kN capture, calculated MPa and density, pass/fail against the mix target, and a client-ready report.
Full version for 1 month. No card needed.