Plasmaline & Plasmaline-NG (microwave SWP and MVP)

Lukas Alberts, Nagoya University
Email:

* 2Daxisym-plasm... is a model of the so-called Plasmaline (E. Raeuchle, http://jp4.journaldephysique.org/articles/jp4/pdf/1998/07/jp4199808PR708.pdf)
it is running a 70 Pa Argon at 500W and 2,45GHz but only on a single side input.
* finger-upload is the single sided version.
* basic-mwp-lineNG is the a version of the plasmaline invented here in Nagoya (MVP here, or Plasmaline NextGeneration for me) which omits the quartz tube but therefore need a DC-sheath space. (it shows the expected behaviour but the chemistry has no metastables)
*MVP17 is my last model. It is converging within 1hr.40minutes with mesh2, you can find out steady-state by monitoring the Timeplot. It needs the right mobility values to get closer to reality and the sputtering effects.(may be not always converging.. sometimes the ion concentration makes problems if not some runaway of e-temperature)

The models do not reproduce reality. Depending on which rate constant description I rely on, the power loss structure is very different (xsec-tables or townsend look-up).
I need help on setting the right chemistry and mobilities..
What is also missing is the sputtering effects of the DC voltage.
The plasma color is definitively not the color of a pure Argon plasma but a lot of metal ions are present (10E15 to my extrapolation).

need to solve the BE for giving the models the right mobility coefficients..
looking for co-authors for a journal paper.

one big open question is still the resonance effects of the electric field at the cutoff plasma density . I included the doppler broadening parameter used for the ECR case (an erf function on Efield value), but I have no plan how to calibrate it.

what is a good value for secondary emission coefficient of Cu cathode material?

User Comments

Ruifeng Sun
Oct 30, 2013 at 8:52am UTC

more than 4 days???The CPU of your computer can bear it???when the calculation of the simulation is very solw,what i always do is to abort the solve.

Lukas Alberts
Entry submitter
Jan 20, 2014 at 5:08am UTC

yes, I agree, most of the time it is useless to continue calculating.
I would be happy to have a nitrogen plasma model?

Lukas Alberts
Entry submitter
Sep 17, 2014 at 6:42am UTC

By the time, it runs within 1 hour..

But still not happy.
With -250V DC bias I can have plasma densities above 1E17[1/m^3] which is much too high ? something is different between the DC and the MWP module.
Then using the townsend lookup-table or the collision rates of the xsec.txt file, the plasma behaves very differently.
Which is the real one is hard to tell.
Changing the mobility values from a microwave BE solution improves the picture.. but still not at the goal..
lukas

Lukas Alberts
Entry submitter
Sep 30, 2014 at 3:19am UTC

A case for the support: my microwave power is increasing before I ask him to do so? at the en of Study1?
is it some reflection effect of a 10E-20W power? something is wrong there!! from me, or from comsol..

lukas

Login to comment on this entry.