PDA

View Full Version : 2D Meshing - (Baseline vs. Development Scenario) - Changes to the wider 2D domain



ICM User
June 21, 2016, 01:41 PM
Evaluating the flood risk to a development site within a large 2D Zone (domain). Several mesh zones and mesh level zones have been created for the development scenario. Including the representation of river reaches to represent swale structures. The problem encountered, is the 2D zone elements change once re-meshed and a direct comparison between baseline and development scenarios for the 2D zone elements outside the area of interest cannot be made. Element sizes have changed.
Any idea’s how to keep the 2D zone the same, so a comparative analysis of results can be made in wider 2D domain i.e. outside the area of interest.

Hope this is self-explanatory?

Thanking you in advance.

Kristian Ravnkilde
June 22, 2016, 12:19 AM
If there are specific points or areas you are interested in, you can use 2D results points, lines and zones to extract consistent results. Alternatively, and if it's worth the effort, you can split up your base 2d zone into mesh zones that fit the scenario mesh zones in such a way that the areas outside the scenario mesh zones are always unaffected - except possibly at the mutual boundaries if the meshing inside the scenario zone/s is VERY different. Even that can be prevented by having lots of points along your mesh zones boundaries so that the meshing is forced to follow those.

ICM User
June 22, 2016, 01:00 AM
Thanks for the response. Please see my comments in red

If there are specific points or areas you are interested in, you can use 2D results points, lines and zones to extract consistent results. Acknowledged. This is not the problem.

Alternatively, and if it's worth the effort, you can split up your base 2d zone into mesh zones that fit the scenario mesh zones in such a way that the areas outside the scenario mesh zones are always unaffected - except possibly at the mutual boundaries if the meshing inside the scenario zone/s is VERY different. The red line boundary outline of the development does contain mesh zones.

Even that can be prevented by having lots of points along your mesh zones boundaries so that the meshing is forced to follow those. Re-meshing the river reaches within the 2D zone, will change the mesh and this will have an effect on the wider 2D meshing.

Please see attached are two snapshots images of the same element ID, upstream of the site over 1 km away.


292293

The problem is changes to an isolated boundary. In this particular case a development red line boundary, where site levels are raised and SUDS features are introduced. Re-meshing the 2D zone for the entire model in is current set up, will change the element sizes in the larger 2D domain. Is there a option to re-mesh for just the site boundary - without making changes to the wider 2D baseline for a direct comparison of elements IDs?

Duncan Kitts
June 22, 2016, 01:06 AM
You will need consistent mesh polygons in each mesh to ensure the mesh is identical. You can add the mesh zones and mesh levels zones to the base scenario and the development scenario. The geometry of these features will be used to create the 2D mesh. In the base scenario, the mesh zones and mesh levels zones can be set not to modify the mesh ground levels, whereas in the development scenario they can be set to adjust the elevations as necessary.

There is no option to mesh a specific area of the 2D zone.

ICM User
June 22, 2016, 01:16 AM
Thanks for your response. I will double check, however I have been through this process of backwards engineering the models so they are identical where possible. I was under the impression the meshing of the river reaches were causing the problem...?

Duncan Kitts
June 22, 2016, 01:20 AM
Are the river reaches consistent between the 2 scenarios?

Kristian Ravnkilde
June 22, 2016, 01:21 AM
Maybe I didn't express myself clearly. The point is to introduce mesh zones with the "scenario" boundaries into the BASE, so that the areas outside these mesh zones are unaffected when you change things within the scenario zones. This may get complicated, and you will need to define the boundaries carefully. Is it really worth the effort? The difference in elevation is 2cm, which is slightly odd if it's 1km away unless it's very flat, so there may be more at work than the meshing? You can't only re-mesh one part of the 2D zone.

ICM User
June 22, 2016, 01:26 AM
No.

The baseline model site boundary is subject to flood risk. The development scenario is to raise the site and use swales / channel, etc to reduce the flood risk.
Therefore by introducing rivers reaches into the development scenario to represent swale features, etc , you are forced to re-mesh and this appears to create smaller mesh elements, which has a knock on effect to the wider 2D domain. Perhaps extending the mesh zone boundary may solve this, they are close to the mesh zone boundary.

Any suggestions welcome?

Kristian Ravnkilde
June 22, 2016, 01:32 AM
So wouldn't doing what I suggested introduce the smaller elements in the base as well?

ICM User
June 22, 2016, 02:34 AM
Comments in red.

So wouldn't doing what I suggested introduce the smaller elements in the base as well?

Already tried this prior to the posting the message.

Kristian Ravnkilde
June 22, 2016, 02:39 AM
So is it something else? Is one (or more) of your banks unstable, leading to an unrealistic maximum level? They are evil creatures and conspire against us!

Duncan Kitts
June 22, 2016, 05:01 AM
If I understand correctly, you have River Reaches in the scenario but not the base network. Therefore, you're fundamentally changing the geometry that requires meshing. This would result in different meshes. This would mean it would not be possible to do a direct comparison between element depths. However, an alternative approach would be to interpolate the results from the mesh into an ascii grid or similar is GIS and then do the comparison based on this.

ICM User
June 22, 2016, 05:11 AM
The alternative approach may have to be adopted on this occasion.
Cheers Duncan.

Martin Spiers
June 27, 2016, 02:10 AM
Adding swales will create different and probably smaller triangles in the area where the swales are. You can't really avoid that.
But you can
a) specify several mesh zones and only remesh those which have changed. You'll need to change the baseline scenario as well, but from then onwards you should have consistency in the wider area
b) generalise the results by interpolating into a consistent (say 1m) grid, which may smooth out any effects from changes to the mesh between scenarios


No.

The baseline model site boundary is subject to flood risk. The development scenario is to raise the site and use swales / channel, etc to reduce the flood risk.
Therefore by introducing rivers reaches into the development scenario to represent swale features, etc , you are forced to re-mesh and this appears to create smaller mesh elements, which has a knock on effect to the wider 2D domain. Perhaps extending the mesh zone boundary may solve this, they are close to the mesh zone boundary.

Any suggestions welcome?