 |
Workers Section |
Demonstration Script
Setting is that the SC03 Show Floor is a Tokamak Control Room with an
AG node. This specific tokamak is not running but
we want to be a remote collaborator to DIII-D that is running so we will use
this idle control room as our remote participation center. Schissel (DS)
will act as the remote scientist.
We assume DS' responsibilities for the day include the EFITs that are run
between shots, the TRANSP code that is run between shots, and data analysis
specific to the experiment.
We further assume that DS communicates directly to another scientist in the
DIII-D control room discussing data analysis. It is that local scientist in the
DIII-D control room that communicates to the on-site team for changes to be
made in time for the next pulse.
Setup before first pulse: 5 minutes total
- DS presents overview of demonstrations
- Uses PowerPoint Slides displayed on AG (PDF version)
- Key Points:...
- DS Configures EFIT to run between pulses - done ahead of time for first demo,
done now for next demo.
- He sets the following parameters: 1) Which EFIT input file to read
for the day, 2) How many time slices to run per pulse,3 ) what time interval
between time slices, 4) The time of the first analsis slice.
- The EFIT interface then lets him know how long this will take to run. If not
acceptable changes can be made or if it OK he exits with no need to revisit
this GUI.
- DS Configures TRANSP to run between pulses - done ahead of time for first demo, done
now for next demo.
- Input TRANSP file is set which includes what time interval the code
is run for.
- This time interval may need to change as the day evolves.
Information Displayed
- Via a web browser (Mozilla based) the electronic log book is displayed.
- Via a web browser the Real Time DIII-D Data Traces are Displayed
- Via a web browser the Shot Cycle Status and data acquisition status is displayed.
- A ReviewPlus window is up with graphics. Before the first shot of the day
a reference shot from a previous day is displayed.
- A EFITViewer window is up with graphics. Before the first shot of the day
a reference shot from a previous day is displayed.
- Elvis window is up and running showing reference TRANSP run.
- The AG node is fired up and communicates with DIII-D.
- A paper copy of the MiniProposal and Run Plan is on hand.
First pulse: 15 minute total
- Counting down to first pulse
- Chief operator (person at GA) audio is heard counting down via the AG (3, 2, 1, shot in progress).
- Pulse happens by script being run on GA laptop at SC.
- RT displays show plasma current and a few other
key parameters in Real Time (like an oscilliscope trace).
This is how I know the pulse is and has happened.
- DS describes the value and importance of the various displays.
- Data Acquisition Begins
- Data acquisition begins automatically.
- When magnetics data is acquired EFIT starts automatically
- Shot cycle status will change but not related to acquisition.
- While waiting for interesting data to appear people can talk depending on audience interest
- DS does this no matter what - describes EFIT and TRANSP and their importance.
- Keahey describes the computation on demand system
- The following topics are covered in PowerPoint (2 to 3 slides each)
- Wallace describes X sharing system
- Papka describes AG advances
- Sanderson describes SCIRun advances
- Thompson describes Akenti advances and authorization
- Feibush describes Elvis
- McCune describes Grid enabled TRANSP
- Schissel describes MDSplus and Globus
- Schissel describes FGM
- Keahey describes GT3
- First data of interest appears
- ReviewPlus, EFITViewer automatically listen to events that are declared by the
acquisition system that indicates when data has been acquired.
- ReviewPlus and EFITViewer begin to automatically update. EFITViewer displays
2D movie of plasma.
- As people begin analysis, they begin to write into the Electronic Logbook
and as they do these automatically appear.
- Kate's EFIT appears in about 4 minutes.
- If time budget allows run TRANSP after EFITs are done.
- Analysis phase
- DS uses ReviewPlus and EFITViewer to analyze the shot, a physics story begins to appear.
- As EFITs appear these are displayed. TRANSP data will appear later in the
analysis phase. TRANSP appears on Elvis window.
- Communicate with DIII-D control room, this is ongoing.
- Via AG a discussion begins on DS' analysis results. Interactive ReviewPlus sessions
are shared (both ways).
- A decision is reached on what to try next and this is communicated to the
session leader in the DIII-D control room.
- The DIII-D AG feed shows session leader, his local support team, and the
DIII-D Tiled Wall.
Second pulse: 8 minutes
Demonstration Wrap Up: 2 minutes
- We run TRANSP between pulses.
- Where were we 20 years ago? Compare to today.
- Quantify the difference we are making
- Future customers: FSP and ITER
