Edward G. Sherwood, Evan C. Bentz, Michael P. Collins
4 min
In reinforced concrete design, predicting the shear strength of members without stirrups is a critical safety concern. Traditional design codes often rely on empirical formulas that may not adequately account for the 'size effect'—the observation that larger structural members often exhibit lower shear strength per unit area than smaller, geometrically similar ones. This study investigates how the maximum aggregate size influences the beam-shear capacity of thick slabs, aiming to determine whether current design standards are sufficient for large-scale construction.
The researchers conducted an experimental program involving 20 reinforced concrete slab-strip specimens. The set consisted of 10 large-scale specimens and 10 geometrically similar small-scale specimens. By loading these to failure, the authors were able to isolate the role of aggregate interlock as a primary mechanism for shear transfer. The experimental results were then used to evaluate the predictive accuracy of the standard ACI design method against a simplified approach derived from the Modified Compression Field Theory (MCFT).
The study demonstrates that aggregate interlock is the dominant mechanism for shear transfer in these elements. Consequently, the maximum aggregate size is a vital parameter in determining shear capacity. The results reveal that the standard ACI design method becomes unconservative when applied to thick slabs or wide beams lacking stirrups, as it fails to properly account for the reduction in shear strength as member size increases. In contrast, the simplified MCFT-based method proved both safe and accurate, successfully predicting how changes in aggregate size affect the behavior of lightly reinforced members. The authors conclude that future shear design codes must explicitly incorporate the role of aggregate interlock to ensure structural safety in large-scale concrete elements.
First Name is required invalid characters Last Name is required invalid characters Email Address is required Invalid Email Address Invalid Email Address
Sam: [quiet confidence] That's one of the notable implications in the data. Increasing maximum aggregate size can restore shear capacity in these deep members, because larger stones keep the interlock mechanism active even as cracks widen. It's a way to recover strength without necessarily adding more steel — though that's a design lever the paper flags rather than fully validates across a range of mixes.
Alex: [reflective, slower pace] So instead of throwing more steel at a shear problem, you're treating the concrete mix itself as a structural variable. There has to be a limit to that. [[RP_SECTION:scope-and-limitations|Scope and limitations]]
Sam: [measured, acknowledging the point] There is. The study is scoped to slabs without stirrups — the specimens most vulnerable to this size effect. MCFT holds up as accurate and conservative there, but the authors don't extend it to stirrup-reinforced members or high-strength concrete. In high-strength mixes the aggregate can fracture before it slides, which changes the failure mode entirely — a different mechanism the paper doesn't test.
Alex: [deliberate, checking understanding] So MCFT is a more physically grounded model, but a targeted one — it's specifically better at capturing aggregate interlock in these deep, unreinforced sections, not a universal replacement for the code.
Sam: [nodding] That's the right scope to put on it. The ACI code's k-factor doesn't capture how cracks propagate in large volumes, and treating aggregate size as a primary input lets engineers anticipate the brittle, sudden shear failures that the empirical method can miss in exactly this class of member.
Alex: [sitting back, broader view] It sounds like the harder problem for future codes isn't this one equation — it's the general move from empirical formulas to mechanistic models that actually track how concrete behaves as a composite material at different scales.
Sam: [warm, professional] If you want the figures and the method choices we skipped, you can generate a deep dive of this paper. The paper has the rest either way.
Alex: Thanks for listening.