I'll admit, I had to look that up!
adendritic: Not having a branching structure, not tree-like.
While it's true that most HydroCAD drainage models consist of converging branches, they can also have diverging flows (commonly overflows) which can recombine at a point downstream, or even remain separate. You can also have multiple networks with no interconnections, or even loop flows. Most arrangements are permissible.
There's no trick to setting up the model. Just route the outflows where you want them to go.
"No trick" is being a tiny bit disingenuous. Having worked with these in HydroCAD and other software several times, you may in fact need some tricks depending on your setup.
In HydroCAD, you likely want to do these with dynamic storage indication method or simultaneous routing instead of straight storage indication method if they come back together downstream, since your geometry for such arrangements is probably somewhat tailwater driven. But then when you do so, you can sometimes get oscillations in your model that will cause it to go haywire. That's the same for any software. Peter will tell you to crank down the time scale discretization of your modeling effort to get rid of these (so will the error message) but sometimes that doesn't work. Sometimes it even makes them worse. I recommend first trying multiple different time increments, both shorter and longer, to get a clean run. I learned back in my SWMM days that another good way to dampen oscillations and try to smooth things out is often to put small amounts of dummy storage in each node, which acts like sandpaper on the sharp edges of your hydrographs.
HEC-HMS can (used to?) choke up a bit if your calculation time increment isn't a factor (doesn't divide evenly) with the time increment of your unit hydrograph table. I'm not sure whether that's an issue with HydroCAD at all, given that it's building its hydrographs out of TR-20. I expect not.
Hydrology, Drainage Analysis, Flood Studies, and Complex Stormwater Litigation for Atlanta and the South East -
Sorry fellows, I mistakenly omitted a space in the word "adenditric"...I meant to say "a dendritic", with many branches combining and flowing into other branches to a common POI (Point of Investigation). Sorry for the confusion.
Beej67 - I was referring to the ease of setting up the routing diagram: Just drag the outflows where you want them to go - even for diverging and loop flows it's easy. Getting a stable routing can be challenging if there are tailwater effects - that's the nature of a tailwater-sensitive analysis with rapidly varying flows - everything is in flux. But most of these can be resolved by appropriate fine-tuning of the model. The warning messages will guide you to the appropriate "tricks"!
Katracho - Did no get your question answered? "Dendritic" networks are the norm.
My question has to do with the following: What is the proper way to combine subareas into other subareas downstream? Do I use Reaches, Links or Ponds with no significant storage?
Why can't I route an upstream node through a subcatchment?
A subcatchment contains only the information needed to perform a runoff calculation. Use a reach if you want to route another hydrograph through this land area. For shallow overland flow, use a wide channel with a suitable Manning's number. For details read about overland flow.
How do I add two hydrographs?
Hydrographs are automatically added whenever two or more outflows are routed to the same node. If you want to sum two hydrographs without doing any further routing, create a reach, pond, or link with no description. When the node is calculated the inflows will be added and passed through unchanged.
