WSE drops in front of bridges

[rsbe]

what may cause this WSE drop in front of a bridge in my unsteady model (with constant discharge)? I cannot figure out why there is a drop since there is no change in geometry or slope with respect to the XS. In the very same model without the bridge the WSE looks fine (no drop, continuous WSE)…

[cameron]

Is the wse touching the bridge at all? Bridges/culverts tend to add losses/changes to wse if they impact the flow at all.

[rsbe]

no, it is not touching the bridge at all. the WSE is way below the low chord. maybe I should mention I have an unsteady model with constant discharge (to generate a hot start file for a flood event afterwards). it is really strange because looking at the HTAB curves of the bridge reveals that for 6.9m3/s discharge the WSE should be at 459.37, and 459.37 is the actual WSE before it drops. so everything would be ok if there wasn’t the post processor (with a different solution scheme somehow, have a look at page 8-26 in the manual) which produces a WSE of 459.29 (by calculating backwards from the downstream bounding XS to get detailed WSE info within the bridge) instead and therefor I have this strange drop for “no reason”. I would like to get rid of this drop, any ideas?

[cameron]

Did you modify the internal bridge cross-sections at all? When you add a bridge two internal cross-sections are generated which are just copies of the upstream and downstream cross-sections including the same invert. You can go in and adjust the inverts in the internal cross-sections to follow the same slope which will probably fix your problem.

[Anonymous]

What bridge modeling approach are you using? It if is momentum, try switching to energy.

Also, have you looked at the bridge curves that unsteady created? There could be a problem with them.

[rsbe]

Hi. Yes, I have already tried keeping the same slope by changing the internal XS and get rid of any abrupt change in the slope. I only had to add/subtract about max. 1cm to/from the internal XS elevations, so this “XS change” should not have such a huge impact i my opinion. But unfortunately, it didn’t solve my problem… i still have these drops and they tend to make my model unstable (seen in unsteady state with constant discharge because these drops are not fix in the sense of they move up and down a little bit)… any other thoughts? However, in flat areas i could manage to get rid of these drops or let’s say at least shorten them to a couple of millimeters which i don’t care about. but in steeper areas the problem remains unsolved…

[rsbe]

Hi. Yes I use the energy bridge modelling approach. An yes I have had a couple of looks at the HTAB curves of the bridges and they look fine (without any jumps really smooth). I have also put the max of points there and limited the discharge to maximize the number of points in the “discharge-area of interest”. Nothing helped…

[Anonymous]

I recommend you post the model.

You mention this issue doesn’t exist for flat areas; perhaps the XS spacing (as shown in the screenshot) is insufficient for this particular reach in an unsteady simulation.

Also, although it sounds like you’ve exhausted most avenues to address the issue, it’s not clear from the provided profile geometry, or the discussion, what may be going on.

[Anonymous]

XS 4 (flow beginning to contract) is very close to XS3 and the bridge, while XS 1 (flow fully expanded) is downstream of the screenshot. Assuming the bridge abutments do not obstruct flow as mentioned, I suppose XS 1-4 all ought to be closer together (i.e. XS1 closer to the bridge).

The ‘bridge’ could also be modeled by using XS’s instead (no structure in RAS), if you’re confident there should be no flow obstruction to warrant the jump.

[rsbe]

Hi. Here https://polybox.ethz.ch/index.php/s/HaM8NS8ILEiOe5i you can find and download my model (on the top right of the screen where it says “Herunterladen”). It has a main river with bridges and a tributary. The unsteady flow file has a constant discharge in order to generate a restart file for a flood event afterwards. The bridge shown in the screenshot above is 15’676.9. In HECRAS lingo I often only have cross section 3 and 2 with respect to bridges. So you would recommend to interpolate the other two XS “4” and “1” to allow the flow to contract and then to expand again after the bridge? To be mentioned: For example through the bridge 14718 the flow is fine and there is almost no jump. But for example through bridge 13599.9 there is this sudden drop from the bounding XS number 3 to the internal XS of the bridge. I cannot remove this bridge (13599.9) because it really affects flow and has a significant backwater effect in case of a flood event (the water leaves the river on the left side due to pressurized flow). The idea of modelling unsteady is that this project is in connection with a retention basin upstream so one would like to see what happens over time (where and when do we have capacity problems etc. since the outflow is varying over time). One was not happy with the steady state because here HECRAS defaulted to critical depth at a couple of XS’s as it couldn’t solve the energy balance. However, in unsteady state without any bridges everything seems to be ok. I would really appreciate if you had a look at my model. Thank you in advance.

[Anonymous]

the bridge at RS 13600 ‘really affects flow and has a significant backwater effect’; that is, the bridge at RS 13600 is an obstruction to flow. The XS geometry at the up- and downstream faces suggests the same. The top width about doubles moving upstream. Why wouldn’t you expect headloss here?

[rsbe]

I fully agree on a significant head loss between XS 13757 and XS 13600 due to contraction, BUT I would actually suppose this head loss to happen some kind of gradually from XS 13757 (XS 4 in HEC lingo) to XS 13600 (XS 3 in HEC lingo), wouldn’t you? However, this head loss seems to happen in the form of a sudden drop between XS 13600 and the internal upstream bridge XS, even though the internal XS is a simple copy of XS 13600 so there shouldn’t be any major head loss between them, right? Or am I missing something?

[Author]

Posts