Currently I need to generate secondary electron yield properties for Which I need Particle vs Time in CST PIC solver.
For hexahedral meshing, Eigen solver is used. A fixed frequency with 4 modes using JDM is implemented. Do I need a waveguide port with zero reflection or fields generated without wave port are acceptable? I have used periodic boundaries, but some referenced papers also suggest phase shifts in order to obtain travelling wave.
Further more, The PIC solver hampers ram (98%) when simulated with imported E and H-fields. Is only E-field enough. Various megawatt powers are applied through PIC waveguide excitation list.
Need help in these ambiguities to save simulation time and generate good SEY and PvsT.
Fields with no ports are acceptable unless you are interested in transient effects. For traveling wave periodic structures are needed, otherwise you need to create a correct superposition of the solutions with E- and H- walls in the resonant mockup with a correct number of cells. This can be a source of errors - periodic walls are much simpler.
Are you simulating multipactor discharge? Then H-fields are important too. You can consider TRACK solver, however.
Thankyou very much for your answer Sir.
Due to the unique geometry and its position on grid (waveguide), I have been concerned whether solution without waveguide port would generate correct directional travelling wave. The problem with waveguide port assignment is that Q-factor monitors are then not accepted.
I will use periodic boundaries as you have suggested for E and H field imports.
Yes I am simulating for multipactor discharge. For this particular setup, I need PIC solver and not Trk solver. Simulating for 4 modes in JDM with ports only progresses 13% in 14 hours (96% ram). Please let me know how to resolve this.
Kind regards.
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators