Other experiments on vtf

| Home | Introduction | The Machine | Research | Publications | People | Links |

Blob Experiments at VTF

In magnetically-confined plasmas, fluctuations in density/temperature/etc. near the edge are prevalent. The character of these fluctuations is often non-diffusive; in other words particles aren't moving around by random-walk, but rather they form coherent structures that convect outward towards the chamber walls. These propagating structures are often called filaments (since they tend to be aligned along the magnetic field) or blobs. Plasma transport by blobs across the magnetic field is generally unwelcome, because it can degrade the chamber walls, as well as plasma confinement.

Here is a movie from Stewart Zweben at the Princeton Plasma Physics Lab showing blobs at the edge of NSTX (using gas puff imaging).

Zweben (4 MB, mpeg). Taken from http://www.pppl.gov/~szweben/

And here is a movie of blobs we can create controllably in VTF:

VTF (200 KB, mpeg)

The advantage of VTF is that we can control the blobs and create them reproducibly, while in a tokamak one can only observe blobs when they are created by the main plasma. Furthermore, because the VTF plasma is short-lived and relatively low density, Langmuir probes can be used extensively without melting, to give the detailed internal structure of the blobs. The following diagram shows our experimental setup:

Our plasma is created by a short (60 µs) burst of microwaves at 2.45 GHz with a power of 15 kW. The resonance condition for electron-cyclotron resonant breakdown at this frequency is that the local magnetic field be 87 mT. The background toroidal magnetic field (into the figure) is only 40 mT, but it is enhanced locally by the solenoid's magnetic field (see the picture below) to give local breakdown. The result: a ring/filament of plasma near the inner wall that extends all the way around the machine. The blob then propagates outward at some fraction of the sound speed (c_s = T_e/m_i).

The solenoid used to enhance the toroidal magnetic field.

We have observed for the first time the nonlinear mushroom shape of the blobs. The following figure shows the blob at three times separated by 100 µs. The mushroom shape is consistent with that seen in many simulations, one of which due to O.E. Garcia is shown at the bottom right of the figure. At the upper right is the electrostatic structure, which shows charge buildup at the top and bottom of the blob. This charge build-up, which in the single-particle picture is due to grad-B and curvature drifts, gives a downward electric that which gives an outward ExB velocity.

We investigate the dependencies of the blob velocity, which is important because it determines plasma losses (transport) at the edge of tokamaks and other plasmas. The average blob velocity is found to scale inversely with the neutral density in the chamber (our ionization fraction is only about 1 %):

This result means that the blob loses momentum to the neutrals.

The electric field in the blob, and the resulting ExB drift velocity, can be quantitatively compared to the blob velocity determined by density measurements. The following plot shows the agreement. At low pressures, the measured electric field is too low due to systematic errors associated with low-density floating potential measurements. Note that the electric field is measured as the vertical gradient of the floating potential so we are assuming that the electron temperature is constant along the vertical direction.

The ExB graph shows three things:

1. The 1/Pn dependence
2. Agreement with velocity measured from ion saturation measurements (filled symbols)
3. That velocity decreases as blob propagates to larger R.

(The inset illustrates the layout of the Langmuir probe lines used for these measurements).