Example 5: Collisions

Collisional processes in EPOCH describe physics packages where incident particles interact with a background species, and significantly affect local background particles. When running a collisional process, both incident and background particles are re-sorted into cell-owned species-lists, in order to quickly identify these local particles.

The package test_process_collisions demonstrates this, and is of a similar form to a recombination package. The incident particle has a constant trigger cross section of 1e7 m², and the process destroys the incident particle when activating. Separately, the code estimates how many background particles are triggered each step, using the weight and trigger rate of the incident particle. Background macro-particles are triggered until the number of real background particles activated matches the expected number of real particles triggering in the incident macro-particle. In test_process_collisions, background macro-particles are sent to a secondary species upon trigger.

This example has a similar setup to Example 4, where a relativistic, low-density electron bunch passes through a static background electron plasma. The incident bunch consists of 1e5 macro-electrons describing a number density of 1e3/m³. The background electron plasma is a uniform 100/m³ density, with 100 macro-particles per cell.

Only one example is provided with this setup:

• example_5_0: Collisional process demonstration

Example 5.0

The incident electrons start in the first cell, and move across the domain at relativistic speed. A uniform density of background particles is present, and the cross-section of test_process_collisions is also constant - hence, the trigger rate is fixed for electrons in the incident species.

The physics package block takes the form:

begin:physics_package
   process = test_process_collisions
   incident = Electron_bunch
   background = Background_electrons
   secondary1 = Promoted_background
end:physics_package

The density of the incident bunch starts at 1e3/m³, and incident macro-particles are destroyed when the process is triggered. At the end of the simulation, the bunch density has dropped, and we see most incident electrons have been destroyed:

Since the background number density is initially uniform, and the reaction rate is constant, we expect an exponential decay pattern to occur for the spatial distribution of trigger locations. As background particles remain stationary both before and after promotion to the secondary species, we expect these secondary particles to show where reactions are taking place. Hence, we expect these to also follow an exponential distribution, which is what we see:

These secondary particles have been promoted from the background species, and we can see this by plotting the Background_electrons number density. Here, the exponential pattern from the Promoted_background species has been subtracted from the initially uniform 100/m³ background density: