• Source tree

Commits in savu/codes:master

• Wednesday February 01 2017

  • 00:24

    

    jumijero committed 8c75ade

    cfd.py: save screenshots from all simulations; use different base filename for different simulations

• Monday January 23 2017

  • 13:15

    

    jumijero committed 20d1c04

    cfd_navierstokes_incompressible.py: some final tidying. Seems this method is just bad at dealing with boundaries in flow problems.

  • 13:13

    

    jumijero committed 00423de

    cfd.py: maybe dt=2e-5 also works; better to have fit_order=4 with max_nk=36 (larger neighborhood, also geometrically)

• Tuesday January 17 2017

  • 23:14

    

    jumijero committed 66276bc

    cfd.py: navier_stokes_incompressible(): visualize div(u) to see the problem more clearly

  • 17:26

    

    jumijero committed 586b0e6

    cfd_navierstokes_incompressible.py: attempt at incompressible solver, seems this numerical method doesn't do that well

  • 17:25

    

    jumijero committed fa8f98e

    cfd_navierstokes_compressible.py: comment update; minor code cleanup

  • 17:24

    

    jumijero committed e54a988

    cfd_potentialflow.py: use get_bi()

  • 17:24

    

    jumijero committed f26b264

    cfd_pointclouds_base.py: added get_bi() to get the indices in boundary data corresponding to a set of indices to self.points[i,:] representing boundary points; comment update

  • 17:23

    

    jumijero committed dc90fa4

    cfd_advection.py, cfd_euler.py: comment update

  • 14:19

    

    jumijero committed 731ef8e

    cfd_navierstokes_compressible.py: comment update (document also rho0)

  • 14:17

    

    jumijero committed ec5459e

    cfd_navierstokes_compressible.py: comment update

  • 14:16

    

    jumijero committed 733c8c7

    rename cfd_navierstokes.py --> cfd_navierstokes_compressible.py

  • 14:14

    

    jumijero committed 0cf058a

    cfd: use standard symbol 'mu' for dynamic viscosity (to not confuse it with the kinematic viscosity nu = mu/rho); the molar mass, which is also commonly denoted by 'mu', is now available as 'mu_mass'.

  • 02:15

    

    jumijero committed 7a7707c

    cfd: comment update

  • 02:08

    

    jumijero committed b82d0e8

    navier_stokes_flow3.mp4: example video

  • 02:07

    

    jumijero committed 8e7179a

    encode_video.sh: use libavformat output to produce mp4

  • 02:01

    

    jumijero committed 2953bed

    flow_over_cylinder.geo,msh, potential_flow.pickle: new slightly larger flow domain

  • 02:00

    

    jumijero committed e97a05c

    encode_video.sh: changed baroque encoding options to only slightly antiquated encoding options

  • 01:59

    

    jumijero committed d3dd568

    cfd.py: use timestep subdivision for pressure response; vis improvements in NS (plot also magnitude of u); save all plotted frames as PNG (for making video)

  • 01:58

    

    jumijero committed 6a22448

    cfd_euler.py, cfd_navierstokes.py: support timestep subdivision for pressure response

• Monday January 16 2017

  • 22:49

    

    jumijero committed d8b7f3c

    cfd_navierstokes.py: oops, the 1/rho should also affect the viscosity term

  • 22:48

    

    jumijero committed 7f63ba9

    cfd.py: potential flow vis improvement

  • 20:22

    

    jumijero committed e8fd524

    cfd.py: plotting improvements

  • 18:47

    

    jumijero committed c975d46

    cfd_navierstokes.py: dynamic viscosity for air documented

  • 16:08

    

    jumijero committed feefbd1

    cfd_navierstokes.py: comment update

  • 15:53

    

    jumijero committed 99acfc1

    cfd_euler.py, cfd_navierstokes.py: support also loading rho0 from initial state file

  • 15:52

    

    jumijero committed c0d15d8

    cfd.py: improved state saving/loading; final state plotting in all flow simulation types

  • 15:52

    

    jumijero committed 578086a

    potential_flow.pickle: added simulation_type; now 4000 timesteps

  • 15:05

    

    jumijero committed dd70f28

    cfd: always generate local coordinate systems for all boundary points. This makes indexing much simpler, since then we have only i in self.cloud.points[i,:] and bi in boundary data (which now includes also self.c (cos theta), self.s (sin theta), and self.bpoints (local (tau,n) coordinates of neighbors)).

  • 15:02

    

    jumijero committed 3f6763e

    cfd_pointclouds_base.py: comment clarification