[alma-config]Re: Early science Configs

[Al Wootten]

Hi Mark

For the present, I agree that what we need is to know what is on our
present horizon.  We at least need a set of configurations for 6 antennas.
As you note, this becomes less critical the more antennas there are.
The reason I split the problem at 32 was that I assumed the innermost
model you mention will work from then on.

I think the 15th antenna will arrive with 16, ..., 23.  I agree that a tool
or rule on what to do with the next antenna or n antennas would be a good
solution.  I'm not sure that we need to consider more than four specific
numbers of antennas for specific configuration sets however.  We'll need
smaller sets for commissioning and verification--three, ... eventually also.

Al

Mark Holdaway writes:
 > 
 > Al,
 > 
 > I think it sounds reasonable to transition from "move all antennas when
 > reconfiguring" when we have 6, to "move 4 antennas every n days" when
 > we have more than m antennas.  However, we still have the problem of
 > "how do we configure 32 antennas on all these pads?"   One answer is 
 > just put them on the inner-most spiral pads, evacuate antennas from the
 > inside to the outside -- ie, the exact same reconfiguration strategy
 > used by the full 64 -- I'm sure that a somewhat better configuration
 > using the existing pads could be found, but at that point it might make 
 > only a small difference, not worth worrying about.  Reconfiguration
 > would certainly take longer, and the configurations would have a lower
 > long_baseline / shortest_baseline ratio.  A simple alternative would be to 
 > make "skeletal" configurations, ie, where every 2nd antenna pad in a
 > 64-pad configuration is not filled.  Some simulations would let us know
 > which style made more sense.  Of course, if you do the skeletal approach, 
 > you need a strategy for filling in the 33rd through 63rd antennas.
 > 
 > However, I think the approach of having special configurations for
 > 6, 14, 23, 32 antennas will be highly non-optimal, though it has the
 > advantage of being a finite problem which is solubale.
 > 
 > What exactly do we need this for?  For DSRP purposes, or to actually know
 > what to do when the 15th antenna arrives?  It seems that we want to have
 > a tool that can help us decide what to do when the next antenna 
 > comes rather than to have a full solution at this moment -- flexibility
 > may be required, as our presuppositions of what is required of these
 > configurations may not hold true as we are commissioning the antennas.
 > 
 > Take care,
 > 
 >    -Mark
 > 
 > > 
 > > Hi John and Mark
 > > 
 > > This sounds like a reasonable plan to me.  Early on, we'll probably have
 > > bursts of science--two weeks of science observations, then two or more
 > > weeks of commissioning and verification, followed by a period of science
 > > observations.  These science times might alternate between configuration
 > > sets which would be combined to obtain useful images.  Proposal calls will
 > > occur every six months or so in the paradigm we currently have, so the
 > > science array wouldn't get augmented on shorter timescales than that, at least
 > > for the first year or so.  
 > > 
 > > One major break point comes when 32 antennas are available.  At least then
 > > but perhaps before we might consider beginning an n day reconfiguration
 > > series, where n becomes 4 in the final stages in our current paradigm.
 > > The Earliest Science configuration plan would then cover the period until
 > > 32 antennas were available, and would consist then of perhaps four sets
 > > of configurations involving 6, 14, 23, 32 antennas.  Comments?
 > > 
 > > Al
 > > 
 > > John Conway writes:
 > >  > 
 > >  > Hi,
 > >  > 
 > >  >  Mark makes some good points, I agree a spatial dynamic range of 20
 > >  > is hard for 6 antennas. but for a fully 1D array of 6 antennas close to
 > >  > EW then  a baseline range of 20 could just about be achieved, but long
 > >  > tracks  are then of course required to get good azimuthal coverage.
 > >  > 
 > >  > Its more in line with the pad pattern and expandability of the array to
 > >  > full operation to have a 2D pattern of 6 antennas. For instance an outer
 > >  > triangle of three antennas and an inner triangle,will give azimuthally good
 > >  > uv covergae in 3-6hrs , but has a low spatial dynamic  range (like 3-5 maybe),
 > >  > however by combining two such arrays one  could get a  useful array for
 > >  > astronomical imaging,
 > >  > 
 > >  >       Jihn
 > >  > 
 > >  > 
 > >  > On Wed, 14 Jan 2004, Mark Holdaway wrote:
 > >  > 
 > >  > >
 > >  > > I think that these are reasonable boundary conditions (though a factor
 > >  > > of 20 between long and short baselines for 6 antennas seems to be
 > >  > > stretching it -- the VLA, with 27 antennas, has a ratio of 40), but
 > >  > > I think the defining concepts for this task may be:
 > >  > >
 > >  > > (1) that we usually want to use multiple configurations to get good (u,v)
 > >  > > coverage (this is ALMA, not OVRO).
 > >  > >
 > >  > > (2) that we want these configurations to be "expandable" in that
 > >  > > we will be adding antennas every month or so.  I know this was not
 > >  > > part of the current objective, but that is really the ultimate task.
 > >  > > Initially (ie, when there are 6 antennas), we may move all or almost
 > >  > > all between each configuration -- but at some point, more and more
 > >  > > antennas will NOT be moved as we change configurations.
 > >  > >
 > >  > >    -Mark
 > >  > >
 > >