PDA

View Full Version : Initial Conditions 2D - ICMv6.5



BrettCawley
April 4, 2016, 07:53 AM
Hi

I am trying to utilise an IC Zone and Initial conditions in order to represent a pond water level, however this seems to be causing issues within the 1D part of the network and initialisation is failing (see screenshot). The elevation specified on the mesh is 60.34 however during initialisation of the 1D network the model is assuming a water level of 61.322 all the way down to the outfall.

Any thoughts?

Cheers
Brett
228

Kristian Ravnkilde
April 5, 2016, 01:38 AM
It's not clear where the pond is, but it sounds like the initial level is leading to either a constant inflow into the system or if the pond is at the outfall, then the inflow from upstream is maybe trying (and failing in the specified time?) to fill the pond to an outlet weir level? Hard to tell without more info.

BrettCawley
April 5, 2016, 02:00 AM
This appears to be due to a manhole within the IC Zone, the ground level of this node is set higher than the level of the 2D element it sits within. Although this is correct in reality this causes instability in the 1D system when there is flow on the mesh. Solution was to reset the node ground level to match the element level.

Duncan Kitts
April 5, 2016, 02:28 AM
This appears to be due to a manhole within the IC Zone, the ground level of this node is set higher than the level of the 2D element it sits within. Although this is correct in reality this causes instability in the 1D system when there is flow on the mesh. Solution was to reset the node ground level to match the element level.

This is because of the way the default transfer of flow between 1D and 2D is based on depth and the fact that you have a mismatch between the element elevation and the manhole level. In most cases the mismatch between node ground level and mesh element elevation is due to different data sources and mesh averaging. Therefore transferring depth, means the mismatch does not affect the simulation. However, at ponds, there is quite commonly a deliberate mismatch of levels. The depth in the element is then applied to the outfall invert level. There is an option to change the transfer globally within the simulation parameters using the 'Use 2D elevations instead of depths' option. In 6.5, it is also possible to specify this on a node by node basis. As you have found, removing the mismatch addressed the issue but use of either of the above options would also help without the need for removing the mismatch in levels (which in this case is deliberate).

BrettCawley
April 5, 2016, 03:27 AM
Thanks Duncan, this is exactly what I required.