Non-erodible surfaces (also known as hard bottoms) may be simulated in HEC-RAS and are specified using a minimum bed elevation,z_{b,\text{min}} , or maximum depth, h_{\text{max}} hmax. In the 2D model the non-erodible surfaces are specified at both cells and faces independently. Non-erodible surfaces are modeled at the subarea scale by limiting the erosion rate so that the minimum bed elevation is preserved (i.e. z_{bi}≥ z_{b,\text{min}}). The time stepping scheme is given by z_{bi}^{n+1} = z_{bi}^n+\Delta z_{bi} in which z_{bi}^n is the previous bed elevation, z_{bi}^{n+1} is the new bed elevation, and \Delta z_b is the bed change. Assuming that the bed elevation at the current time is higher than the hard bottom (i.e. z_{bi}^n≥z_{b,\text{min}} ). Inserting the above equation into the bed change equation leads to the hard-bottom limited erosion rate
1) |
E_{tki,hb} = D_{tki} + \Psi_{na,i} \frac{f_{1ki}\rho_{d1}}{f_M\Delta t}\left(z_{bi}^n - z_{bi,hb}\right) |
where
E_{tki,hb} : hard-bottom limited fractional erosion rate [M/L2/T]
D_{tki} : fractional deposition rate [M/L2/T]
\Psi_{na,i} : non-alluvial erosion limiter [-]
ρ_{d1} : dry density of active layer [M/L3]
f_{1ki} : grain fractions by weight [-]
f_{M} : morphologic acceleration factor [-]
\Delta t : computational time step [T]
The non-alluvial erosion limiter \Psi_{na,i} is function which limits the maximum erosion rate based on the proximity of the mobile bed surface to the non-erodible surface and is computed here as
2) |
\Psi_{na,i} = \frac{\delta_{1i}^n}{\delta_{1i}^*} |
where
\delta_{1i}^n : active layer thickness [L]
\delta_{1i}^* : potential or maximum active layer thickness if there was no non-erodible surface present [L]
When \delta_{1i}^n = 0, no erosion is possible and the maximum erosion rate becomes equal to the deposition rate (i.e. E_{tki,hb}= D_{tki} ) . Likewise when \delta_{1i} = \delta_{1i}^*, the maximum erosion rate is equal to the amount of material in the active layer available to erode in a timestep plus the deposition rate.
The hard-bottom limited erosion rate is therefore
3) |
E_{tki}' = \min\left(E_{tki,hb}, E_{tki}\right) |
The bed-slope term and avalanching algorithm are also modified so that only deposition may occur over non-erodible surfaces following an approach similar to that described above.
When simulating multiple grain classes, the bed gradation can significantly change at non-erodible surfaces during a time step. This means that the model may require at least one to two iterations to converge when simulating non-erodible surfaces with multiple grain classes.