User Forum of Software BASEMENT

davidz

User

Registered: 2016-02-01

Posts: 16

Sediment stays in the first few elements

Hi

I tried to modell a steep mountain stream with fixed bed to see how much sediment stays in the lower part, where the slope is much lower than in the upper part. But the sediment stays in the first few cells and builds up a sediment-mountain of about 500 m.

Here is what i chose:
- 2D hydraulic modell which leads into a lake (fixed zhydrograph as outflow boundary)
- single-grain transport with smartjaeggi
- I use the same scenario just with water as inital-condition so it's not wetted.
- I tried several grain-sizes (original as d90=310mm, d30=35mm and d_m=140mm). For all the same with no transport at all.
- the slope is in the beginning 6.8% steep and there is an inflow of 6.5 m3/s of water and 0.1 m3/s of sediment.

in the end the slope gets bigger and bigger and the modell warns me that there are some mass conservation problems in the elements where the sediment is staying.

How can i get the sediment transported downstream?

Kind regards and thanks for the help.

David

Matthias

Advanced_User

Registered: 2014-12-16

Posts: 29

Re: Sediment stays in the first few elements

Hi David

For me, it seems like there is a lot of sediment input in your model boundary. What is the sediment boundary condition? Did you choose Transport Capacity?

I would recommend you to use an artificial inlet channel of a few hundred meters upstream of your river you want to analyze and add water at the top and a bit further down you add sediment to the artificial channel. You can choose transport capacity but try to use the "factor" to reduce the amount of incoming sediment. The artificial channel helps you to reduce mixing effects due to the boundary conditions etc. and gives you useful results already at the top of your model.

Another option could be to adjust the critical shear stress with the influence of the slope (see documentation of smart&jaeggi formula in the literature). This could increase the transport capacity of your channel.

Best regards and good luck
Matthias

davidz

User

Registered: 2016-02-01

Posts: 16

Re: Sediment stays in the first few elements

Hi Matthias

Thanks for your help.

I had the sediment input modelled as external_source for the hydraulic boundary nodes.
I didn't realize the boundary-options for the sediment. I tried now as well Transport Capacity and the boundary condition sediment_discharge (for my original sedigraph and for just 10% of the amount with extreamly reduced grain-sizes) but then the modell stopps after a few seconds, saying "At the hydrograph boundary all edges are set inactive (not wetted), but the inflow is bigger than 0.0. Either the cells in the boundary cross section are too big or the inflow is too small to ensure that at least one element is wetted."

I also changed the critical shear stress to 0.005 as it is recomended in the Flussbauskript from Bezzola for fixed bed channels.

I tried as well putting the sediment not directly at the hydraulic boundary nodes (similar to the recommendation of the artificial channel which i will try as well soon.). Here I could observe something i didn't understand. When I  just choose one node_id for the sediment input the calculation work fine. When i want to distribute the sediment over 3 node_ids the same error-message as above is coming. But for both cases the sediment stays at the same spot.
(By the way: this tree nodes are equivalent to the ground of the river (approx. 6 m wide). This means the whole river is 5 elements wide. Is this a too coarse grid for sediment transport?)

Best regards
David

Last edited by davidz (2016-02-02 11:06:24)

Matthias

Advanced_User

Registered: 2014-12-16

Posts: 29

Re: Sediment stays in the first few elements

Hi David

Have you found a solution to your problem yet? Just some check-ups to your model:
- Have you activated the Morphology Block? I guess you have because you added grain sizes as Soil_def - Parameters
- Have you added roughness (kst) values to all your elements/zones?
- Maybe check the input reference (http://www.basement.ethz.ch/docs/doc_BASEMENT.html) for the keyword "inactive" - maybe there is something missing in your model.
- Is a hydraulic-only simulation working fine?
- Are you using triangles or four sided elements (squares)?

Best regards
Matthias

PS: Your grid resolution should be fine.

davidz

User

Registered: 2016-02-01

Posts: 16

Re: Sediment stays in the first few elements

Hi Matthias

I just found a solution. I heard that at the User-Meeting this year (January) this  problem was discussed as well in one project (Hydromorphologische Analyse, Flussaufweitung Sandey.) I got their presentation from the meeting and i discovered, that I forget to set a gravitational transport. With GRAVITATIONAL_TRANSPORT i get much better results. Now if i enter the sediment, all the sediment is tranported and deposited where it is "expected" to be and i can calibrate it now.

At the moment I've set following parameters:
angle_failure_dry: 35
angle_failure_wetted: 15
angle_failure_deposited: 5
angle_wetted_criterion: partially_wetted

To your questions:
- I've activated the Morphology Block.
- I use an overall roughness. so all my Elements have a strickler-value.
- hydraulic-only simulation is working fine
- I use triangles for my grid.

And thanks again for your help. Maybe this can help others as well.

David

dalmo

User

Registered: 2016-05-13

Posts: 1

Re: Sediment stays in the first few elements

@davidz could you please upload the presentation you are talking about?
Thanks.

maarnold

Advanced_User

Registered: 2014-09-13

Posts: 28

Re: Sediment stays in the first few elements

Hi Dalmo

You can find the presentation under the following link:
Hydromorphologische Analyse Flussaufweitung Hasliaare

Mat