From: Josh R Klein (jrk@mail.hep.utexas.edu)
Date: Sat Jul 02 2005 - 12:02:03 CDT
Hans,
Thanks for your comments.
>
> we are not taking appropriate credit for cross calibration at the end:
> Clearly we start with two detectors normalized to one another at the near
> site.
> Then we move one and track its parameters using all sorts of calibrations.
> It is important to bring those two detectors back together at a later time
> to verify that the tracking has been successful.
> This is not unlike surveying magnets in the Tevatron and then closing the
> loop to where one began.
> This improves the accuracy hugely and is really the only experimental way to
> put a number on the accuracy of tracking parameters.
> Suggestion:
> After the very first sentence add something like:
> "A second cross-calibration near the end of the experiments will both
> examine the success
> of the ccalibration system in tracking detector responses and
> reduce the uncertainties in the tracking process by having firm data at the
> beginning and at the end of the run."
I think in Mike's talk the plan was not to move the initial detector
back to the near location, but to swap the second set of detectors. We could
certainly change that plan, but I think the current plan probably gives the
same information, albeit perhaps less directly. If you remember, the initial
draft of this question had details about how we would use the information from
the different deployments to understand our overall relative efficiencies, so
I think we could have put this statement in that section.
>
> a. We are much too specific about deploying an articulated source
> throughout the volume with 2 cm accuracy.
> I don't think we want to commit and to "baseline" this device.
> There are good reasons not to:
I don't think there was anything specific in the final draft about
this, other than the 2 cm. The 2 cm applies to any kind of calibration system,
even one which can only deploy a source in the center of the detector. I
expect we will be able to do much better than 2 cm along the axis.
>
> First of all, if the positioning accuracy is 2 cm, then the volume
> uncertainty, even for perfect measuremsnts,
> we will only calibrrate at the level of (6cm/350 cm) which is at the 1.7 %
> . This is not at all useful.
We will not be determining our fiducial volume from the calibration
system, unless we have a disaster on our hands. The fiducial volume is
determined by the vessel itself. If we have to use reconstruction to tell us
the volume, we have a much bigger problem. The 2 cm is an estimate based on
the fact that we will need to know the solid angle subtended by the PMT's in
order to understand our detector response and extract all the relevant
parameters which go into it. In addition, the PMT response with angle and
wavelength can be sensitive enough at high angles that we don't want too much
slop in the position. I based the 2 cm on SNO's positioning along the central
axis (not off-axis) which is what we use for our best calibrations, as well as
KamLAND's stated goal for their new system (which I think is much better than 2
cm).
>
> Secondly, the presence of the mechanism and the source holder changes the
> local materials distribution,
> presumably also at least at the 1% level (needs detailed work)
I agree that a source deployment can, in principle, disturb the
distribution of materials. It can have other affects we haven't even thought
of yet. Any system we decide to go with should aim to minimize these potential
problems, while also ensuring that we will be able to calibrate the detector
accurately enought to make our final measurement.
> Thirdly, the requirement for a 3 ft opening at the top makes the detector
> highly anisotropic.
> This is a big concern, since the neutrino wind changes form 45 degrees at
> the near location to 6 degrees from the horizontal at the far location.
> I imagine one could get around this by tilting the detectors at 19.5 degrees
> in opposite directions at the two sites,
> but that has huge engineering and cost consequences.
> I cannot tell at what level the 3 ft diameter "special place, without
> PMT's" at the North pole creates near / far differences, but it is not a
> comforting thought.
The neutrino wind is an interesting problem, but I haven't yet seen the
demonstration that differences here are anywhere near something we have to
worry about. The neutrons do not travel that far in the Gd-doped scintillator,
though I suppose neutrons from the oil can. I personally think that we should
consider doping the oil with Gd, to reduce neutron leakage into the vessel, but
we would have to study that.
>
> Alternatively, if one can get by with a source/ LED device that scans only
> along the detector vertical axis,
> this can be done through a small (1/2 inch or so) opening at the top.
> I have a conceptual design for that.
> It can be done with superb accuracy (about 1 mm) at any point along the
> axis and it is simple and robust enough to run automatically and frequently.
> It requires no modifcations to the sphere at all.
> This device can also easily and accurately (1 mm) detect when it hits the
> bottom of the sphere and when it emerges from the oil at the top.
> Response as a function of distance from the acrylic can be mapped out in
> exquisite detail (along the axis only, though)
I still do not see how such a system can tell us about the effects on
the detector response from the supports along the equator, the variations in
PMT response from tube-to-tube (other than simple gain and QE), or
variations in reflectivities of the outer tank and PMT support.
>
> Bottom line: I suggest we delete all references to this articulated source,
> e.g. "by deploying sources near the edges of the active volume . ."
A source which is deployed only along the z-axis still goes to the
edges of the active volume, at the top and the bottom, so I don't see what the
statement would need to be changed even for your scenario.
>
> c. Under 9.2.1,
> the first sentence makes it sound like that changes in the PMT's and
> associated electronics are likely.
> I am confident that we will not see much change other than that of
> temperatur changes (if they should occur), which would be fully reversible.
> Suggestion:
> Change ":the most likely change . ." to:
> We do not expect any significant non-reversible change or any PMT failure
> during the moves. If any changes should occur, or PMT's be damaged, then
> these can be easily determined . . ."
The statement is "most likely". I would expect that most people who
have dealt with PMTs would expect that moving 1000 of them is likely to lead to
at least a few problems. I think those few problems may be the worst we have
to deal with, which is why it says "most likely". It would be great if
everything turns out to come back on OK, I'm just not sure it is prudent to
make statements which will counter the experiences that others have had,
without anything to back them up.
>
> d. Under 9.2.1:
> I would suggest to omit the sentence "Only catastrophically large changes .
> .."
> We have no expectation or reason to believe that the occurrence of such
> changes is credible.
Right. But NuSAG has asked us to talk about them anyway.
>
> e. Under 9.2.2
> What exactly is " a diffuse optical source" ?
> Do you mean many sources thorughout the volume (which I fully expect to
> have, in several colors) or do you mean a moving source with an isotropic
> optical emission?
"Diffuse" means "integrating". LEDs and lasers are not by themselves
isotropic sources. The optical calibration will be much easier if we have a
source as isotropic as it can possibly be, and that is what a diffuse source
can give us. A collimated source would also be nice, and we should be thinking
about this as well.
>
> f. Under 9.2.3 you say:
> " . .therefore small changes can have a big effect .".
> This is in contrast to our statement under 9.2.1 :
> "We are very insensitive to changes in these efficiencies . . ."
> Which is it ?
As you see in the document, Section 9.2.3 refers to a change in the
scintillator response, and the small changes in this have a big effect on the
energy response, not necessarily on the overall efficiency. I believe that is
what the statement says. Section 9.2.1 is discussing the efficiencies of the
PMTs, and is saying that these are a small effect on the overall efficiency. I
agree, however, that this could be made clearer.
>
> g. Under 9.2.4
> I would suggest to omit the SNO anecdotal evidence, even though it is cute
> and intriguing; it just doen't apply to Braidwood.
>
The point here is to show that we are aware of problems both gross and
subtle and have been thinking about them. Ignoring previous experiments'
experiences, without good reason, just makes us look ignorant. Far better to
say we are aware of something and that it doesn't apply.
>
>
Thanks,
Josh
This archive was generated by hypermail 2.1.6 : Mon Jul 04 2005 - 03:10:17 CDT