Dynamic wave routing – batch-oriented St. Venant equations. More...

#include "DynamicWave.hpp"
#include "Node.hpp"
#include "Outfall.hpp"
#include "XSectBatch.hpp"
#include "ForceMain.hpp"
#include "Culvert.hpp"
#include "../core/Constants.hpp"
#include "../core/SimulationContext.hpp"
#include "../core/UnitConversion.hpp"
#include "../math/SIMD.hpp"
#include <cmath>
#include <algorithm>
#include <cstdio>
#include <cstdlib>
#include <numeric>

Include dependency graph for DynamicWave.cpp:

Namespaces
namespace	openswmm

namespace	openswmm::dynwave

Detailed Description

Dynamic wave routing – batch-oriented St. Venant equations.

The solver follows this pattern per Picard iteration:

computeLinkGeometry() – batch call to XSectGroups for ALL links:
- depth1/depth2 from node heads and link offsets
- Preissmann slot geometry applied when depth > y_full (closed conduits)
- XSectGroups::computeAreas/HydRad/Widths for free-surface depths
- Slot area/width/hrad overrides for surcharged depths
solveMomentumBatch() – pure array arithmetic over ALL links:
- velocity = old_flow / area_mid (vectorisable)
- Froude from velocity and hydraulic depth (vectorisable)
- sigma (inertial damping) from Froude (vectorisable)
- full inertial damping when closed conduit is surcharged
- friction slope dq1 = dt * roughFactor / rMid^(4/3) * |v| (vectorisable)
- head gradient dq2 = dt * g * aMid * (h2-h1)/L (vectorisable)
- momentum update: q = (qOld - dq2 + dq3 + dq4) / (1 + dq1)
- under-relaxation (vectorisable)
updateNodeFlows() – scatter link flows to node inflow/outflow
updateNodeDepths() – per-node Picard convergence check
- EXTRAN surcharge: dQ/dH with smooth transition near crown
- Separate convergence tracking for surcharged vs free-surface nodes

Note: Legacy reference: src/legacy/engine/dwflow.c, dynwave.c

Author: Caleb Buahin caleb.nosp@m..bua.nosp@m.hin@g.nosp@m.mail.nosp@m..com

License\n MIT License

Namespaces

Detailed Description