Real time boundary detection of toroidal plasmas from visible light images

Typically, the boundary of nuclear fusion plasmas is determined from magnetic signals, but alternatives are possible. For example, microwaves can be coupled into the plasma and reflected at the plasma boundary.

The plasma also emits visible light from the boundary region. The interpretation of the images not straightforward. In 2010, Hommen et al. ( ) showed that, with two assumptions, visible light images from tokamak plasmas can be interpreted as plasma boundaries. The assumptions are:

  1. The light is emitted from a toroidal surface, that is
  2. Tangent to the line of sight of the the camera.

Figure 1. shows how, for a given camera position, the lines of sight form a cone that is tangent to the toroidal surfaces at various toroidal and poloidal locations.Tangency cone. The lines of sight that are tangent at various poloidal and toroidal locations

Tangency cone. The lines of sight are tangent to the toridal surface at various poloidal and toroidal locations



The obtained image needs to be processed, in order to find the boundary. In the same paper, Hommen derived an expression to do exactly that! For the transformation, we refer to the RSI contribution. The expression was tested on data of the spherical tokamak MAST, in Culham, and compared to magnetic measurements of the plasma boundary. We found good agreement between the two methods.

Knowledge of the boundary combined with local spectroscopic measurements of the Motional Stark Effect is sufficient to non-iteratively assess the the lay-out of the plasma interior (see Hommen et al. Plasma Physics and Controlled Fusion Vol. 55 No 2 ).

MaST image and optical boundary reconstructionNow we (FOM / TU/e and TNO) are finalising a two camera, real-time optical boundary recognition system on the Swiss tokamak TCV. The optical relay system for the two cameras was designed by HAS and modified by TU/e.

To reach the specification of 1000 frames per second, we will use real-time cameras with Camera Link interface connected to frame-grabbers with on-board processing (FPGA/DSP). This system will provide sections of the plasma boundary and strike point locations. The choice of port views could be adapted per campaign, does not have to be hard coded into the system. For this campaign we have opted for the top and middle port.

Simulations show, for given camera locations, which parts of the boundaries can be resolved by the two cameras.

TCV vessel, two cameras (1000 fps) and the projected plasma boundary coverage

TCV vessel, two cameras (1000 fps) and the projected plasma boundary coverage

We are installing the system this week. Real-time images should be obtained at 1000 Hz as of next week, and RTC experiments of the plasma centroid position are foreseen before the shut down of TCV.

As soon as we have a 1000 Hz movie, I will post it here, plus the movie of the boundary.

Fingers crossed!


Over Marco de Baar
Dit bericht werd geplaatst in Hall of fame, Innovatie en getagged met , , , . Maak dit favoriet permalink.

Geef een reactie

Vul je gegevens in of klik op een icoon om in te loggen. logo

Je reageert onder je account. Log uit / Bijwerken )


Je reageert onder je Twitter account. Log uit / Bijwerken )

Facebook foto

Je reageert onder je Facebook account. Log uit / Bijwerken )

Google+ photo

Je reageert onder je Google+ account. Log uit / Bijwerken )

Verbinden met %s