Merge pull request JuliaPhysics#443 from JuliaPhysics/efieldlines

Add keyword argument `Δt` to electric field line plotting

Author: fhagemann

docs/src/man/plotting.md

@@ -172,7 +172,8 @@ the plot recipe for electric field lines can additionally understand the followi
 * `sampling`: Specifies the steps at which the contacts are sampled to generate equally spaced charges at the surface. The default is `2u"mm"`, but the optimal value depends on the geometry of the detector and contacts. If no unit is given, `sampling` is interpreted in units of meter.
 * `offset`: The charges are created on the surface and have to be moved slightly inside the semiconductor to be able to drift (charges that are in the contacts will not drift). This keyword defines how much the charges will be moved inside along the normal vector of the surface. The default is `0.5u"mm"`, but the optimal value again depends on the detector geometry. Unitless quantities are interpreted in units of meter.
 * `skip_contact`: Detectors will usually have positively and negatively biased contacts. The charges should only be spawned on either of them, but not both. This keyword defines which contact should be skipped in the charge spawn algorithm. Default is `1`, i.e. that no charges will be spawned at the surface of contact with `id = 1`.
-* `max_nsteps`: After all, the generation of electric field lines is based on the charge drift code, which requires a maximum number of steps, which is set to `5000` by default here. If the drift ends before the charges reach a contact, the electric field line will not be fully displayed. In this case, it is recommended to increase `max_nsteps`.
+* `max_nsteps`: The generation of electric field lines is based on the charge drift code, which requires a maximum number of steps, which is set to `5000` by default here. If the drift ends before the charges reach a contact, the electric field line will not be fully displayed. In this case, it is recommended to increase `max_nsteps`.
+* `Δt`: The charges are drifted with a normalized mobility of $1\,\text{cm}^2\text{/Vs}$ (see [`ElectricFieldChargeDriftModel`](@ref)). If the individual steps are too large, it is recommended to reduce `Δt`.
 
 
 

src/PlotRecipes/ElectricField.jl

@@ -92,6 +92,7 @@ end
                     sampling = 2u"mm", # Specifies in what density the Contacts are sampled to generate equally spaced surface charges. Also see spacing.
                     offset = 0.5u"mm", # should be at least as big as sampling. In doubt sampling can be reduced and the spacing keyword can be used to thin out the lines.
                     spacing = 1, # If, due to fine sampling, too many lines would clutter the plot, the spacing keyword allows to skip some fieldlines. Spacing = 2 means plot every second line. Spacing = 3 every third.
+                    Δt = 1u"ns",
                     full_det = false,
                     skip_contact = 1) # Usually the "core" contact is skipped, and the other contacts are equally sampled for charges to drift
     sim = gdd.args[1]
@@ -165,7 +166,7 @@ end
 
     for el_field in (el_field_itp, el_field_itp_inv)
         paths::Array{CartesianPoint{T}, 2} = fill(zero(CartesianVector{T}), length(spawn_positions), max_nsteps)
-        last_step::Int = _drift_charge!(paths, Vector{T}(undef, max_nsteps), sim.detector, sim.point_types, sim.electric_potential.grid, CartesianPoint.(spawn_positions), ones(T, length(spawn_positions)), T(1e-9), el_field, Electron, verbose = false )
+        last_step::Int = _drift_charge!(paths, Vector{T}(undef, max_nsteps), sim.detector, sim.point_types, sim.electric_potential.grid, CartesianPoint.(spawn_positions), ones(T, length(spawn_positions)), T(to_internal_units(Δt)), el_field, Electron, verbose = false )
         for i in 1:size(paths, 1)
             path = @view paths[i, 1:last_step]
             @series begin