- Max(rho) vs. Time for 3 different runs[JPEG]:
(a) The run that blows up quickly indicates that resolution is not enough for them (h=0.75)
(b) The run in the middle is more stable with increase in central density with oscillation (h=0.375).
(c) The bottome line is with highest resolution tried so far (h=0.1875)
- id file for runs with boundary located at [rhomax=zmax=24]
![[JPEG]](http://lheawww.gsfc.nasa.gov/~choi/axi/GR_BS/max_rho_b24_0.jpg){kind=link}
- Max(rho=phi^2) vs. Time for 4 different runs for about 2 light crossing times, Time=[0--180]. [JPEG]---> parameters for this run was slightly wrong, so I reran: The results are essentailly same [JPEG]
(a) resolutions are h=(1.5 0.75 0.375 0.1875).
(c) results are as expected!-- star is stable and maximum density is almost constant as the resolution is increased!
- Boring evolution-- Nothing happens as supposed!
- id file for runs with boundary located at [rhomax=zmax=48] (this is old one)
- Convergence testing [JPEG] -- This is a Plot of self-convergence factor (defined as usual) of rho for 4 different resolutions. The line goes up first then goes below the other line is computed from the runs of level=[0--2]. The other line is from the runs of level =[1--3]. Note that this is *infinite* norm. LOOKS GOOD!
![[JPEG]](http://lheawww.gsfc.nasa.gov/~choi/axi/GR_BS/maxrho_b48_tmax180_1.jpeg){kind=link}
![[JPEG]](http://lheawww.gsfc.nasa.gov/~choi/axi/GR_BS/max_rho_vs_time_1.jpeg){kind=link}
![[JPEG]](http://lheawww.gsfc.nasa.gov/~choi/axi/GR_BS/cf_infnorm_b48_1.jpeg){kind=link}
- ADM MASS vs. Time for 3 different set up with 5 different initial boost parameter (p_z) [PS] (being updated!)
- Comments 1: With boundary at [0,64]x[-64,64], runs with higher resolution gives smaller changes in ADM MASS before the stars collide---this is sort of what you expect, BUT bigger changes in the ADM MASS after the collistion---this is counter intuitive because you would think that higher resolution woule help. However interaction with boundary is larger for high resolution run-- \rho=\phi^2 at
h=1
h=0.5 
- Comments 2: With the same resolution at h=1, runs with a bigger box (boundary at 96 instead of 64) gives smaller changes in ADM MASS--- this is sort of what you expect.
Convergence factor of AMR runs for a single stationary boson star (although boring) confirms convergence of the solution with convergence factor of about ~3
The two lines in the plot are time history of MAX[ABS(Phi_r_l0-Phi_r_l1)]
and 3 * MAX[ABS(Phi_r_l0-Phi_r_l1)] where Phi_r_l0 is the real part of Phi4 (massive complex scalar field)
for the coarsest AMR run and Phi_r_l1 is real(Phi4) for the higher resolution AMR run, etc.
The fact that the convergence factor is 3 means the whole scheme is convergent between the first
and second... (1.5-order?).
order
![[jpeg]](http://lheawww.gsfc.nasa.gov/~choi/axi/bs_headon_collision/coarse_emap5.jpeg){kind=link}
![[jpeg]](http://lheawww.gsfc.nasa.gov/~choi/axi/bs_headon_collision/psi420_fmr_mult.jpeg){kind=link}
![[jpeg]](http://lheawww.gsfc.nasa.gov/~choi/axi/bs_headon_collision/psi420_fmr_scaled.jpeg){kind=link}