Update 1D-recycling example

[mikekryjak]

• CVODE version now consistent with beuler
• Reduced CVODE nout to 100 as while 400 may be necessary to get rid of all the tiny oscillations but may confuse new users
• Reduced d density floor to prevent unphysical temperatures
• Enabled diagnose=true in all components
• Changed sheath BC to sheath_boundary_simple as it's the most thoroughly tested one

To do:

• [ ] Change gamma setting to something more expected (4.5 for electrons, 2.5 for ions?)
• [ ] Test again
• [ ] Change collisions to Braginskii
• [ ] Improve comments

[mikekryjak]

Interestingly, sound_speed seems to slow down CVODE by about 10% - 20%. Probably because of the electrons being less damped now.

[mikekryjak]

Scratch that, adding sound_speed doubled the simulation time. @bendudson what do you think we should do with this...? On one hand, adding the component increases accuracy, on the other hand it will slow people's runs down and may lead to crashes during transients. In the end we are way less dissipative than SOLPS and other codes anyway...

Also, I found that the odd neutral temperature profiles were due to the neutrals hitting the floor. Lowering the floor to 1e-7 resolves the issue. Although you get the funny bump in Tn which Stefan spent a good amount of time tracking down - it's due to neutral friction and only appears when you correctly include the 5/2 factor in the neutral diffusion advection, which I did for 1D some time ago.