The new SANS 51992 1 1 code for structural concrete changes the way the material is mixed and placed in order to ensure structures fail in a controlle
The new SANS 51992 1 1 code for structural concrete changes the way the material is mixed and placed in order to ensure structures fail in a controlled, ductile manner rather than suddenly and catastrophically.
At the Concrete Society of Southern Africa (CSSA) seminar on the SANS 51992 1 1 code, Dr Kim Timm of Stellenbosch University guided delegates through the critical subject of ultimate limit states in the new code. She also highlighted new terminology and models which will constitute the new principles of reinforced concrete design.
She began by reminding the audience of the basic assumption that has long underpinned reinforced concrete design: plain sections remain plain. Even under bending the strain distribution across a section is assumed to remain linear allowing engineers to balance compression in the concrete against tension in the steel. This principle is carried over from SANS 10100 and remains the foundation of the new code. The difference lies in how stress–strain models are expressed and how neutral axis depth and design values are calculated.
The SANS 51992 1 1 is based on the Eurocode equivalent and introduces slightly different parameters for the stress block used in flexural design. Where the old South African code applied a factor of 0.67 to compressive strength the new code uses 0.85 and reflects the shift from cube to cylinder strengths and the incorporation of long term effects. Similarly the neutral axis depth is limited to 0.8 of the effective depth compared with 0.9 in the old code. These refinements may appear minor but they ensure consistency with international practice and more accurate modelling of material behaviour.
“Over reinforcing compression zones may produce brittle failures which the code explicitly seeks to avoid. The design philosophy is to ensure that steel yields before concrete crushes which gives warning and adds resilience. “We need structures to fail in a ductile manner so that if failure ever occurs it does so gradually not suddenly.
“There is continuity between the old and new systems which I illustrated by comparing the familiar K value method from SANS 10100 with its new counterpart. Engineers still calculate a dimensionless K value to determine whether compression reinforcement is required. The process is identical though the limiting values differ slightly. If the design falls below the threshold only tensile reinforcement is needed. If it exceeds it both tension and compression reinforcement must be provided so the equations may look different but the logic is unchanged.
“When it comes to shear design the changes are more obvious because the old code relied on empirical British data while SANS 51992 1 1 adopts a variable strut inclination model. This approach treats shear resistance as a truss mechanism within the beam with concrete carrying compression and stirrups carrying tension. Engineers can select strut angles between 21.8° and 45° and so balancing the contribution of concrete and steel. Shallow angles reduce steel requirements but demand more from the concrete whereas steeper angles increase steel demand but ease concrete stresses. The key is to ensure that steel yields before concrete fails in order to maintain ductility and safety,” Timm said.
She noted that in this regard the SANS 51992 1 1 code’s shear provisions are more sophisticated and show the improved understanding of concrete behaviour that has been gained over the years since the old code was adopted. It is also worth noting that minimum shear reinforcement is still required but the code clarifies that ribbed slabs such as waffle slabs do not need stirrups in their ribs unless specifically required. This distinction avoids unnecessary detailing while maintaining safety.
Examples provided also showed that while calculations may produce slightly different reinforcement quantities compared with the old code the orders of magnitude remain consistent. Engineers should not expect to see designs suddenly requiring double the steel and if they do it is a sign of misapplied equations rather than a flaw in the code.
