By: Stanford Gibson, Ph.D. and Alex Sánchez, Ph.D.

Most river engineers and scientists learn early on that the highest velocities in a bend occur near the outside bank. But standard two-dimensional (2D) hydraulic models often predict the opposite result. Without accounting for the secondary flow - the three-dimensional (3D) helical motion that develops through a bend - 2D models tend to place the maximum velocity along the shortest flow path, near the inside bank. That is hydraulically incorrect, and it has important consequences: morphological models struggle to form point bars, outer-bank shear stresses are underrepresented, and sediment is often over-deposited behind navigation structures.

Figure 1: Without secondary currents a 2D model erroneously predicts maximum velocity and scour along the inside of a bend (left).  The new secondary current algorithms push that mas velocity to the outside of the bend allowing the model to form point bars.

Figure 1: Without secondary currents a 2D model erroneously predicts maximum velocity and scour along the inside of a bend (left).  The new secondary current algorithms push that mas velocity to the outside of the bend allowing the model to form point bars.

To address this limitation, HEC added quasi-3D secondary flow algorithms to HEC-RAS 7.0 (HEC, 2025). Using channel curvature from a flow centerline, these capabilities compute separate lateral velocities near the water surface and near the bed, improving how mass and momentum are redistributed within each 2D cell. The result is a practical quasi-3D enhancement to the 2D equations that better represents bend flow physics.

With these tools, HEC-RAS can now reproduce deposition on the inside of river bends, as shown in the idealized example in Figure 2.  HEC-RAS can also simulate sediment dynamics around navigation structures. In a Mississippi River Geomorphology and Potamology (MRG&P) initiative, the Memphis District (Holt et al, 2006) applied these tools (as well as RAS2025 meshing) to simulate navigation structures on the Mississippi River and found that the results matched prototype behavior well (Figure 2).

These new secondary flow capabilities expand the applicability of 2D modeling, especially for river erosion, deposition, and navigation studies.

Figure 2: Observed (left) bed change on a Mississippi River bend with navigation structures and the bed change simulated with the HEC-RAS sediment model using the new secondary current features from Holt et al. (2026).

Figure 2: Observed (left) bed change on a Mississippi River bend with navigation structures and the bed change simulated with the HEC-RAS sediment model using the new secondary current features from Holt et al. (2026).

HEC (2025) "Secondary Flow", HEC-RAS 2D User's Manual https://www.hec.usace.army.mil/confluence/rasdocs/r2dum/latest/development-of-a-2d-or-combined-1d-2d-model/secondary-flow

Holt, G., Huffman, J., Sanchez, A., Gibson, S., Brown, G., McAlpin, T. (2026-in press) Comparison of Adaptive Hydraulics (AdH) and HEC-RAS 2D Hydrodynamics and Sediment Transport – Dike Effects on Channel Morphology: Hickman, KY to New Madrid, MO reach, MRG&P Report.