How can I fix a local coordinate system to the inlet

phys-zephyr · May 4, 2014, 10:07

Dear all,

I am working with a time- and spatially dependent velocity inlet profile together with a time-dependent geometry. The velocity profile is defined in a local coordinate system which lies on the inlet boundary surface. The time-dependent geometry is realized with the morphing feature. So far, everything works fine.
I just have the problem, that as my geometry moves and deforms in time, the inlet part also moves but the local inlet coordinate system stays at its origin. As a consequence, the velocity profile does not stay at the inlet boundary and therefore after a certain time, no velocity is defined at the inlet anymore.
Is there a way to attach the local coordinate system to the inlet surface in a way that it will move together with the surface instead of staying at its origin? Did anyone ever found a way to solve such an issue?
A possible solution would be to morph the geometry in a way that the inlet always stays at its position and only the rest of the geometry moves. But this unfortunately can not be easily realized, as I only have the displacement for the surface and not the inlet...

Thanks for any suggestions/comments in advance

Cheers

me3840 · May 6, 2014, 09:31

Is the inlet moving physical? Can you change your BCs to a location such that they don't move?

ping · May 7, 2014, 06:41

you can define point probes and enable the follow motion property so that they move with your mesh since they will follow any mesh motion but be wary of remeshing since the cell the points are locked to will disappear

so if you define some near your inlet boundary you will have some useful values like normal or corner locations using a few reports of position of the points

then this information can be used in a macro which runs the case and every few time steps or even every time step update the coordinate system values - might need to do a bit of maths getting there like cross products but this is easy in a macro

ping · May 8, 2014, 05:46

a maybe better idea is to isolate one central cell face on your inlet boundary as a separate boundary but with the same settings as the the rest of the boundary

then do reports on just that boundary for position and normal and use this then is a macro to alter the coordinate system properties

and thinking about it some more if the boundary is planar the same could be done on the whole boundary depending on the report you choose for getting the position eg area weighted average position maybe

May 4, 2014, 10:07	How can I fix a local coordinate system to the inlet	#1
phys-zephyr New Member Join Date: Mar 2014 Location: Switzerland Posts: 10 Rep Power: 12	Dear all, I am working with a time- and spatially dependent velocity inlet profile together with a time-dependent geometry. The velocity profile is defined in a local coordinate system which lies on the inlet boundary surface. The time-dependent geometry is realized with the morphing feature. So far, everything works fine. I just have the problem, that as my geometry moves and deforms in time, the inlet part also moves but the local inlet coordinate system stays at its origin. As a consequence, the velocity profile does not stay at the inlet boundary and therefore after a certain time, no velocity is defined at the inlet anymore. Is there a way to attach the local coordinate system to the inlet surface in a way that it will move together with the surface instead of staying at its origin? Did anyone ever found a way to solve such an issue? A possible solution would be to morph the geometry in a way that the inlet always stays at its position and only the rest of the geometry moves. But this unfortunately can not be easily realized, as I only have the displacement for the surface and not the inlet... Thanks for any suggestions/comments in advance Cheers

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Micro Scale Pore, icoFoam	gooya_kabir	OpenFOAM Running, Solving & CFD	2	November 2, 2013 13:58
[ICEM] Local coordinate system and Standard shapes	Lance	ANSYS Meshing & Geometry	3	May 21, 2013 12:27
Setting the coordinate system for an arbitrary probe	rks171	STAR-CCM+	0	January 3, 2012 13:11
local coordinate system for porousZone	Nico A.	OpenFOAM Programming & Development	0	November 15, 2011 04:32
Defining local coordinate system	bastil	OpenFOAM Pre-Processing	11	August 8, 2011 08:26

May 6, 2014, 09:31		#2
me3840 Senior Member Join Date: Nov 2010 Location: USA Posts: 1,232 Rep Power: 24	Is the inlet moving physical? Can you change your BCs to a location such that they don't move?

May 7, 2014, 06:41		#3
ping Senior Member Ping Join Date: Mar 2009 Posts: 556 Rep Power: 20	you can define point probes and enable the follow motion property so that they move with your mesh since they will follow any mesh motion but be wary of remeshing since the cell the points are locked to will disappear so if you define some near your inlet boundary you will have some useful values like normal or corner locations using a few reports of position of the points then this information can be used in a macro which runs the case and every few time steps or even every time step update the coordinate system values - might need to do a bit of maths getting there like cross products but this is easy in a macro

May 8, 2014, 05:46		#4
ping Senior Member Ping Join Date: Mar 2009 Posts: 556 Rep Power: 20	a maybe better idea is to isolate one central cell face on your inlet boundary as a separate boundary but with the same settings as the the rest of the boundary then do reports on just that boundary for position and normal and use this then is a macro to alter the coordinate system properties and thinking about it some more if the boundary is planar the same could be done on the whole boundary depending on the report you choose for getting the position eg area weighted average position maybe