You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Update BackTransformed diagnostics to take into account arbitrary moving window velocity (BLAST-WarpX#5341)
In the `development` branch, the `BackTransformed` diagnostics assume
that the moving window moves exactly at the speed of light. This PR
generalizes the code for arbitrary moving window velocity.
This PR does not add an automated test, but the upcoming PR BLAST-WarpX#5337 will
add a test which features a moving window with a speed different than
`c`.
This is a follow-up of BLAST-WarpX#5226, which modified the transformation of the
simulation box coordinates for arbitrary moving window velocity, but did
not yet update the `BackTransformed` diagnostic code.
Copy file name to clipboardExpand all lines: Docs/source/usage/faq.rst
+2-2Lines changed: 2 additions & 2 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -74,10 +74,10 @@ Several BTD quantities differ slightly from the lab frame domain described in th
74
74
In the following discussion, we will use a subscript input (e.g. :math:`\Delta z_{\rm input}`) to denote properties of the lab frame domain.
75
75
76
76
77
-
- The first back-transformed diagnostic (BTD) snapshot may not occur at :math:`t=0`. Rather, it occurs at :math:`t_0=\frac{z_{max}}c \beta(1+\beta)\gamma^2`. This is the first time when the boosted frame can complete the snapshot.
77
+
- The first back-transformed diagnostic (BTD) snapshot may not occur at :math:`t=0`. Rather, it occurs at :math:`t_0=\frac{z_{max}}c \beta/(1 - \beta\beta_{mw})`, where :math:`\beta_{mw}` represents the speed of the moving window. This is the first time when the boosted frame can complete the snapshot.
78
78
- The grid spacing of the BTD snapshot is different from the grid spacing indicated in the input script. It is given by :math:`\Delta z_{\rm grid,snapshot}=\frac{c\Delta t_{\rm boost}}{\gamma\beta}`. For a CFL-limited time step, :math:`\Delta z_{\rm grid,snapshot}\approx\frac{1+\beta}{\beta} \Delta z_{\rm input}\approx2\Delta z_{\rm input}`. Hence in many common use cases at large boost, it is expected that the BTD snapshot has a grid spacing twice what is expressed in the input script.
79
79
- The effective length of the BTD snapshot may be longer than anticipated from the input script because the grid spacing is different. Additionally, the number of grid points in the BTD snapshot is a multiple of ``<BTD>.buffer_size`` whereas the number of grid cells specified in the input deck may not be.
80
-
- The code may require longer than anticipated to complete a BTD snapshot. The code starts filling the :math:`i^{th}` snapshot around step :math:`j_{\rm BTD start}={\rm ceil}\left( i\gamma(1-\beta)\frac{\Delta t_{\rm snapshot}}{\Delta t_{\rm boost}}\right)`. The code then saves information for one BTD cell every time step in the boosted frame simulation. The :math:`i^{th}` snapshot is completed and saved :math:`n_{z,{\rm snapshot}}=n_{\rm buffers}\cdot ({\rm buffer\ size})` time steps after it begins, which is when the effective snapshot length is covered by the simulation.
80
+
- The code may require longer than anticipated to complete a BTD snapshot. The code starts filling the :math:`i^{th}` snapshot around step :math:`j_{\rm BTD start}={\rm ceil}\left( i\gamma(1-\beta\beta_{mw})\frac{\Delta t_{\rm snapshot}}{\Delta t_{\rm boost}}\right)`. The code then saves information for one BTD cell every time step in the boosted frame simulation. The :math:`i^{th}` snapshot is completed and saved :math:`n_{z,{\rm snapshot}}=n_{\rm buffers}\cdot ({\rm buffer\ size})` time steps after it begins, which is when the effective snapshot length is covered by the simulation.
0 commit comments