Screen diagnostic

[vbrsi]

Hi,
I have some doubts regarding the Screen diagnostic. As far I understand, it counts all particle crossing in the 'direction' the surface of the screen identified by point and the normal direction to point. Then these two diagnostics:
#1
DiagScreen(
shape = "plane",
point = [Lsim[0]/2, Lsim[1]/2],
vector = [1., 0.],
direction = "forward",
deposited_quantity = "weight",
species = ["proton"],
axes = [ ["x", Lsim[0]/2, Lsim[0], 1], # I know this is redundant but I use it for normalization purposes, i.e.
["ekin", 0.0, 200., 1000] # to be equal to the Binning norm.
],
every = 100
)

#2
DiagParticleBinning(
deposited_quantity = "weight",
every = 100,
species = ["proton"],
axes = [ ["x", Lsim[0]/2, Lsim[0], 1],
["ekin", 0.0, 200., 1000]]
)

should give the same spectrum, but that is not the case (yellow spectrum vs blue one(s)). What I am missing here? Are only the particles with strictly px!=0 and py==0 counted maybe?
Thanks for your help

[mccoys]

There are a few differences. The binning diag will count particles that are already there, while the screen will only count the passing ones. The screen diag might count several times particles passing and coming back.

And you should not have a "x" axis for screen, I think.

[vbrsi]

Thanks for the reply. Yes I did not mention that the Screen is cumulative but precisely because it is, if I have in the binning diag particles with E=30 MeV at t=t1, they should appear and also in the Screen because the must have come by crossing it. The Screen however should not count particles crossing from right to left (direction=forward) but also the binning should not contain those as they would be outside the domain I have specified. The difference should not be that then, unless I am missing something here.
Regarding the axes, I tested that these three are equivalent

DiagScreen(
#name = "my screen",
shape = "plane",
point = [Lsim[0]/2, Lsim[1]/2],
vector = [1., 0.],
direction = "forward",
deposited_quantity = "weight",
species = ["proton"],
axes = [ ["x", Lsim[0]/2, Lsim[0], 1],
["ekin", 0.0, 200., 1000]
],
every = my_every
)

DiagScreen(
#name = "my screen",
shape = "plane",
point = [Lsim[0]/2, Lsim[1]/2],
vector = [1., 0.],
direction = "forward",
deposited_quantity = "weight",
species = ["proton"],
axes = [ ["x", Lsim[0]/2, Lsim[0], 1],
["y", 0. * um , Lsim[1], 1],
["ekin", 0.0, 200., 1000]
],
every = my_every
)

DiagScreen(
#name = "my screen",
shape = "plane",
point = [Lsim[0]/2, Lsim[1]/2],
vector = [1., 0.],
direction = "forward",
deposited_quantity = "weight",
species = ["proton"],
axes = [ ["x", Lsim[0]/2, Lsim[0], 1],
["a", -Lsim[1]/2, +Lsim[1]/2, 1],
["ekin", 0.0, 200., 1000]
],
every = my_every
)

adjusting for the normalization.

[mccoys]

Is it just a problem of units? What if you scale it by some factor?

[vbrsi]

No it's a problem of units among the three Screen diagnostics, but the spectrum is intrinsically different from the Screen diag and the Binning diag: as you can see from the plot there are no particles with E> 25 MeV in the Screen (blue lines) and there are in the binning (yellow)

[mccoys]

Are there no particles beyond the limit already there from the start?

Are there periodic boundary conditions?

[vbrsi]

No there are not, as it should be because that is the rear side of a target were a laser is shot. BCs are 'silver-muller' for the fields and 'remove' for the particles