Rebar Development Length and Lap Splice Design
Reinforcing steel doesn't do anything for a concrete member unless the concrete around the bar can actually transfer force into it, and that transfer happens entirely through bond, friction and mechanical interlock between the bar's deformed surface and the surrounding concrete, along the length the bar is embedded. A bar that's yielding at one section but only embedded a short distance beyond that point can simply pull out of the concrete before it ever reaches its full tensile capacity, no matter how strong the steel itself is. Development length is the embedment distance required to make that pullout impossible before the bar yields.
What Development Length Actually Depends On
Required development length scales up with bar diameter, since a fatter bar has proportionally less surface area per unit of cross-sectional force it needs to transfer, and scales down with concrete compressive strength, since higher-strength concrete provides more bond stress capacity per unit length. Bar spacing and cover both matter too: bars packed close together or with thin concrete cover are prone to a splitting failure mode, where the concrete around the bar cracks radially and releases the bar's grip, rather than a clean pullout, and code development length equations penalize tight spacing and shallow cover with longer required lengths to guard against that splitting mode specifically.
Top bars, horizontal reinforcement with more than 12 inches of fresh concrete cast beneath them, need meaningfully longer development length than bottom bars in the same member, because water and air trapped rising through the fresh concrete during placement tend to collect against the underside of top bars, weakening the bond there compared to bars near the bottom of the pour where the concrete consolidates more fully around them. This top-bar effect is one of the few development length factors that has nothing to do with the bar or the load and everything to do with which way up the member was cast.
Hooked bars, where the bar terminates in a 90-degree or 180-degree bend rather than running straight, need much shorter development length than a straight bar of the same size because the hook itself provides substantial mechanical anchorage independent of straight-length bond. This is why beam-column joints and other tight, congested regions of a member almost always use hooked bar details rather than trying to fit a full straight development length into a space that can't physically accommodate it.
Lap Splices: Two Bars Sharing the Load Through the Concrete Between Them
Where a continuous bar isn't practical, at a construction joint, or simply because mill lengths of rebar are finite, two bars are lap spliced by running them parallel and overlapping for a specified length, transferring force from one bar to the other indirectly through the concrete surrounding both bars rather than through any direct bar-to-bar contact. Lap splice length is generally longer than straight development length for the same bar, because the splice has to develop force through a load path that goes bar, into concrete, into the second bar, a less direct and less efficient transfer than developing a bar into a continuous, uninterrupted concrete section.
Lap splices are explicitly prohibited by code in certain locations of seismically detailed members, particularly within the plastic hinge region of a beam or column expected to yield repeatedly in a design earthquake, because a splice in that zone would be subjected to the same repeated inelastic strain reversal that governs boundary element detailing in shear wall coupling beams, and a lap splice's bond-dependent load transfer degrades under exactly that kind of cyclic reversal far more readily than the continuous bar itself does.
Mechanical and Welded Splices as Alternatives
Where space doesn't allow the lap length a straight splice would need, or where code prohibits lap splicing outright in a critical zone, mechanical couplers or welded splices transfer force directly bar-to-bar without depending on concrete bond at all, at a materially higher cost per splice than simply overlapping two bars, a tradeoff similar to choosing precast concrete connections that transfer force directly through steel-to-steel hardware rather than through a cast-in-place, bond-dependent detail.
ACI 318's chapter on development and splices of reinforcement is the primary reference most engineers design against directly, published by the American Concrete Institute, and it remains one of the more frequently misapplied chapters in practice precisely because the required length depends on so many interacting factors rather than a single lookup table value.