InOut Boundary and Conservation of Mass
I am in the process of trying to calibrate a coupling between DualSPHysics and a 2D wave model. For a reason that I am not quite sure, the InOut boundary is not removing particles as I would expect, resulting in (based on SWL elevation gauges) surface elevation showing a generally positive trend, omitting stages of significant outflow.
I am using an InOut boundary using velocity as an input. See timeseries of input below:
For reference, although not used as an input, this is the time series for surface elevation, taken from the same 2D model at the point where velocity is also extracted:
I have set up the InOut boundary within the *_Def.xml file as follows:
NOTE: I realise that the z limits for the InOut zone may look a little weird, however, the basin I am working with has a thick base to it.
When I run this case, it would seem as though the boundary only provides positive flow into the domain. Below is a series of plots comparing DualSPHysics elevation output (red) to approximate expected elevation taken from the 2D model (grey). Each of these plots comes from a wave gauge at an increasing distance from the InOut boundary - the bottom plot being closest.
Especially clear in the top two plots, you can see the equilibrium seems to stop returning to zero. When looking at the first velocity input plot, there should be a significant decrease in the water level just before the 1-hour point (first vertical line from the left), yet even the elevation gauge closest to the InOut boundary doesn't show much of a dip below zero at all.
I wondered if anyone had any suggestions or had experienced similar issues when coupling another model with DualSPHysics? Any ideas would be much appreciated.
Thanks in Advance!
Comments
Perhaps @iarba27 can help you