culverts connected in series

[benoit]

Hello everybody,
Please excuse me for my poor english.
Does anybody experienced modelling of various culverts connected in series, taking into account a change in slope, radius etc. ?
Is ther any possibility to do that with Hec-Ras ?

[Anonymous]

If a culvert changes slope (the culvert has a vertical bend), or if it changes size/shape, then this should be modeled as two separate culverts (break the culvert into two parts where the culvert changes slope). Each culvert can then have a different slope. Between the two culverts, there should be two regular cross sections with lids (the contraction and expansion loss coefficient for these two cross sections should be set to zero). The cross sections with lids can be tightly spaced. For instance, maybe ~1 foot (~0.1 meters) of distance between cross sections. The cross sections with lids should be the same basic size and shape as the culvert.

If you do not wish to model minor losses, the exit loss coefficient on the upstream culvert should equal zero and the inlet loss coefficient on the downstream culvert should also be zero. For minor losses, set the inlet loss coefficient on the downstream culvert to the “k” minor loss coefficient. The Solution Criteria for the downstream culvert should also probably be set to Outlet.

Be careful with reach lengths. If a 2000 foot culvert is broke into two, 1000 foot sections, the reach length for the cross section just upstream of the upstream culvert should be around 1000 feet. Likewise, the downstream of the two cross sections with lids should also have a reach length of around 1000 feet.

To add a lid to a cross section: From the cross section data editor, go to Options and select Add a lid to XS. The lid is like a road deck on a bridge or culvert. The High EL is the top of the lid (that is, the road deck) and Low EL is the underside of the lid (the bottom of the road deck, or, in this case, the top of the culvert). The High EL should be set higher than the energy grade line can ever get (or else the program will put flow up there). So if a 10′ culvert has an invert of 100′, the High EL might be set to 200′ and the Low EL would be set to 110′ for a box culvert. A circular culvert would require several points to create the curve. A plot of the cross section should be enough to see if the lid has been entered correctly.

[Anonymous]

Does anyone have an idea of how I can put an opening below the lid. I am trying to model a canal with inverted siphons. I would like to model the siphons as lids but after setting up the lids and run simulation, I get the error message saying. there is no opening below the deck/lid.

I am looking for a procedure to solve this issue.

[Anonymous]

A cross section with a lid is like a bridge opening. You have to enter the bottom of the opening (the ground/channel) and you have to enter the underside of the bridge deck. There needs to be space between the two. For instance the underside of the bridge deck might be at elevation 130′ and the invert of the channel might be at 100′.

From the error message, it sounds like the data isn’t entered correctly.

[ldguethle]

It’s a bit tedious to create xsecs w/ lids… I suggest setting up a template in Excel where you can copy-paste the xsec station-elevation data and then modify it. Create a chart that you can see the result in Excel as you go. Then copy-paste the data back to RAS and confirm it looks good. Then you have an Excel file you can always use if you ever need to do this again…

Jarvus – they could also do the entire series as xsec with lids, no culverts. What do you think about that? better/worse?

[Anonymous]

Yes, modeling it entirely with cross sections with lids is a viable option. There are a few trade-offs although for most situations they may not matter.

If culverts are used, the chart and scale number can be specified for the culvert opening at the upstream end. This will usually compute a little more accurate entrance losses. Also, the program can check to see if the flow is under inlet control which can be sensitive to the shape of the opening (which is entered with the chart and scale). Most culverts, most of the time, do not operate under inlet control. The difference in inlet loss between using a culvert and a cross section will usually be pretty small.

If using entirely cross sections with lids, to model minor losses in steady flow, there is a “k loss” option under Set Changes in WSE and EG that can be used. In this case, contraction and expansion losses should be set to zero at the minor loss locations so as to not double account the losses.

For Unsteady flow, modeling it entirely with cross sections with lids makes it trickier to incorporate minor losses, but this may not matter that much. For Unsteady flow choosing whichever method is more stable might be the over ridding concern. I haven’t had to model anything like that in Unsteady flow for a while so I don’t have much advice to give for that.

[Author]

Posts