Noise field in a solid target

[HaoPengYY]

Dear developers,

Hi. I am not reporting a bug, but a question about the 2D simulation. We are trying to simulate the situation that a relativistic electron beam flys through the center of a round solid tube (in 2D, this might be a flat electron beam interacts with two flat target). But we find there are intense noise fields in the solid target, even though we use a very high temporal and spatial resolution and a large number of particles per cell. In the figure below, you can find the noise field labeled in the green rectangle.

And attached with the input script. Our question is that why there are such fields and how do we remove them? Thank you in advance.

[HaoPengYY]

REBu235c.py.zip

[mccoys]

Please do not use issues for questions

[HaoPengYY]

Sorry, I was not aware of that.

[mccoys]

Have you checked your resolution compared to the Debye length?

Note that the particles "feel" the fields averaged by a shape function. Use Probe if you want to get fields that particles see.

[HaoPengYY]

We have checked the resolution compared to the skin depth, but not to the Debye length. The plasma is initialized as cold plasma. I guessed the fields we saw are the numerical noises, which are quite large compared to the self field of the low-current electron beam.

You mean that the fields from the DiagFields and DiagProbe are different? This is interesting, because we did not see these noise fields in the 3D simulations, in which we use a DiaProbe to check the fields. While we usually do not put a 2D DiagProbe in a 2D simulation, we just use DiagFields. But why is this difference?

[mccoys]

The Field diagnostic is just the values on the grid. The probes are actually fake particles that use an interpolation technique to obtain the field at their position. Check the documentation to find out about interpolation

[HaoPengYY]

Thanks, we followed your suggestion and it works, but there is something wrong with the magnetic field, we are checking.