Most particles get excluded - I can't work out why!

MikeHersee

Hi, I'm trying to generate my own simple test cases for the first time using the DPI. I have created a case, as below, which is just a ball of water falling on a slanting cylinder. But when I run it most of the particles get excluded towards the end of the simulation. Some even get chopped off the side of the fluid particle sphere right at the beginning, despite apparently being an adequate distance from any boundaries. I have looked for what I think are some obvious potential causes, such as minimum and maximum density and have tried changing parameters to see what changed, but I still don't understand.

Can anyone spot my stupid mistake? Appreciative of any feedback - even if to point to where I might find out the exact reasons those particles got excluded.

Mike

<?xml version="1.0" encoding="UTF-8" standalone="no"?>





















actual | dp | fluid | bound | void



all

MikeHersee

Ok, it seemed to interpret the XML file rather than simply displaying it. Let me see...

MikeHersee

MikeHersee

Sorry, I had to replace the '<' with '!' and it seemed to remove the tabs as well when displayed.

!?xml version="1.0" encoding="UTF-8" standalone="no"?>
!case application="DualSPHysics Pre-processing Interface" date="31/01/2016 13:11:54">
!casedef>
!constantsdef>
!lattice bound="1" fluid="1"/>
!gravity x="0.0" y="0.0" z="-9.81"/>
!cflnumber value="0.2"/>
!hswl auto="true" value="0"/>
!coefsound value="10.0"/>
!coefficient value="0.6"/>
!gamma value="7.0"/>
!rhop0 value="1000.0"/>
!eps value="0.5"/>
!/constantsdef>
!mkconfig boundcount="1" fluidcount="1"/>
!geometry>
!definition dp="0.05">
!pointmin x="0.0" y="0.0" z="0.0"/>
!pointmax x="2.1999998" y="2.1999998" z="2.1999998"/>
!/definition>
!commands>
!mainlist>
!setshapemode>actual | dp | fluid | bound | void!/setshapemode>
!setdrawmode mode="full"/>
!setmkvoid B="0" G="0" R="255" name="fluid space"/>
!drawbox>
!boxfill>all!/boxfill>
!point x="-0.3" y="-0.3" z="-0.3"/>
!size x="2.9" y="2.9" z="2.9"/>
!/drawbox>
!shapeout file="fluid space"/>
!setmkfluid B="153" G="153" R="0" mk="0" name="fluid block"/>
!drawsphere radius="0.5">
!point x="1.4" y="1.4" z="1.4"/>
!/drawsphere>
!shapeout file="fluid block"/>
!setmkbound B="0" G="255" R="0" mk="0" name="pyramid"/>
!drawcylinder mask="0.0" radius="0.4">
!point x="1.0" y="1.0" z="0.0"/>
!point x="1.5" y="1.5" z="0.5"/>
!/drawcylinder>
!shapeout file="pyramid"/>
!/mainlist>
!/commands>
!/geometry>
!/casedef>
!execution>
!parameters>
!parameter comment="Time-stepping algorithm. 1: Velocity-Verlet; 2:Symplectic" key="StepAlgorithm" value="1"/>
!parameter comment="Frequency that Eulerian equations are applied at when Velocity-Verlet in use" key="VerletSteps" value="40"/>
!parameter comment="Interaction Kernel. 1: Cubic Spline; 2: Wendland" key="Kernel" value="1"/>
!parameter comment="Viscosity Formulation Method. 1: Artificial; 2: Laminar+SPS" key="ViscoTreatment" value="1"/>
!parameter comment="Viscosity Value" key="Visco" value="0.05"/>
!parameter comment="Frequency that the Shepard density filter is applied. 0: Not used" key="ShepardSteps" value="0"/>
!parameter comment="delta-SPH coefficient, 0: Not used" key="DeltaSPH" value="0"/>
!parameter comment="The size of the initial time-step" key="DtIni" value="1.0e-4.0"/>
!parameter comment="The minimum allowed size of a time-step" key="DtMin" value="1.0e-6.0"/>
!parameter comment="The length of time to simulate" key="TimeMax" value="1.0"/>
!parameter comment="The time that elapses between data being output" key="TimeOut" value="0.01"/>
!parameter comment="The allowed extra space in the Z dimension" key="IncZ" value="0.01"/>
!parameter comment="Allowed percentage of fluid particles out the domain, 1: 100%" key="PartsOutMax" value="1"/>
!parameter comment="Maximum density value allowed before particles are excluded" key="RhopOutMax" value="2000.0"/>
!parameter comment="Minimum density value allowed before particles are excluded" key="RhopOutMin" value="0.0"/>
!/parameters>
!/execution>
!/case>

Alex

Particles are excluded due to density?? In Run.out it says the number of excluded particles and the number of particles excluded by density.

You can send the XML to us and we can run it and check it.

Two suggestions first:
DO not use DPI, try to learn from other XML examples
coefficient value="0.6" is toooo low so very low number of neighbours are included in the interaction.

MikeHersee

Damn, I was using DPI specifically to try to make sure I had working examples. The XML is hard.