January 21, 1998
Reinhard Schumacher
CLAS-Note 98-001 (v.1)
_____________________________________________________________________________
The Jefferson Lab survey group performed measurements of the Region
One Drift Chamber location inside the CLAS spectrometer on October 3,
1997, prior to the December 1997 e1 run. The raw numbers provided by
this group are analyzed in this report to provide positional offsets
needed in the drift chamber tracking programs. The raw data are given
in Appendix I. The handwritten notes and calculations leading to the
results given below are in Appendix II.
The overall accuracy of the survey group numbers seems to be very
good. They claim +/- 0.4 mm in Z and +/- 0.15 mm in X and Y. Various
cross checks between their results and design numbers unknown to them
but known to us indicate that these estimates are on the conservative
side.
Region One is supported by two dowel pins at its downstream end which
slide into holes in a mating piece attached to the cryostat. At the
upstream end the detector has three legs which attach to "landing
pads" attached to the cryo-ring at 10, 2, and 6 o'clock. The accuracy
of the downstream positioning was determined by how well the mating
hole bracket was aligned and attached to the cryostat. Doug Tilles
aligned this mating piece with a digital level which had an estimated
precision of 1/2 degree. The accuracy of the upstream positioning was
determined by how well the cryo-ring was manufactured and aligned to
the torus. During installation of Region One it was noted that the
cryo-ring was machined to fairly low tolerance, since we could not get
the dowel pins in the three legs of Region One to slide simultaneously
into their holes. The legs of Region One were made such that the
position tolerances of the dowel pins was less than 10 mils, while the
mismatches to the cryo-ring were of the order of 1/4". Only one of
three dowel pins (the one at 10 o'clock) was inserted, and can be used
in the future to ensure excellent reproducibility of the Z chamber
position if and when Region One is removed and reinstalled. The
upstream end of the detector was not surveyed during the installation
process. The upstream positioning was done to eyeball accuracy by
equalizing the small (roughly 1/2") gaps between the six sectors and
the cryostat. Hence the present survey is the only accurate means of
determining the X and Y positioning of the chamber package.
There are 10 fiducial points on the Region One detector package. The
upstream end of Region One has four 1/2" tooling balls, mounted at 12,
3, 6, and 9 o'clock on the inner surface and flush with the downstream
face of the Boss Ring. Each of the six sectors has a scribed cross on
the downstream sector plates which is visible from the upstream end of
the detector.
The sighted (X,Y) coordinates for each of the six downstream crosses
were fitted to a circle to determine the mean rotation and offset of
the downstream end of the detector. The residuals for all sectors
were very small, on the order of tens of microns, and the radius of
the fitted circle was in excellent agreement with the design value.
Thus we have confidence that the downstream positioning is well
understood. We found
X offset 0.72 +/- 0.03 mm
Y offset -0.29 +/- 0.03 mm
Rotation -7.1 +/- 0.8 millirad (counterclockwise about Z)
where the X and Y offsets refer to the difference between the center
of the fitted circle and the CLAS Z-axis which point downstream. Y
points opposite to gravity and X points to beam left to form a
right-handed coordinate system. The rotation, about 4/10 of a degree,
is consistent with the estimated positioning accuracy of the
downstream mounting bracket. (The negative value for the rotation is
correct for a counterclockwise rotation, but may differ from the
definition of rotation angle in the off-line analysis program.)
The upstream tooling balls were used to determine the offset and
rotation of the upstream end of the detector in a manner similar to
the downstream measurements. The offsets determined from both the
vertical and horizontal pairs of tooling balls agreed within errors.
We found
X offset +0.58 +/- 0.1 mm
Y offset -1.88 +/- 0.1 mm *
Rotation -1.277 +/- .006 millirad (c.c.w. about z)
The offsets and rotation are again consistent with our estimated
positioning accuracy known from the installation period. The offset
values basically agree with the memo sent out by Dan Carman on
11-10-97 (see Appendix II), but the number given with the (*) differs
in sign. The rotations are given above for the first time.
The measured distance between pairs of tooling balls both horizontally
and vertically was consistently 2.2 mm larger than the distance
calculated from the Boss ring and tooling ball geometry. We do not
understand the reason for this difference, but we do not have complete
information about how this part of the survey was done (which part of
the tooling ball was targeted, which tooling ball type was actually
used, etc). This lack of knowledge was not important in extracting
good results below.
Using all the survey points it was possible to estimate the offset of
the entire chamber in X, Y, and Z. The design distance in Z from the
origin to the downstream scribe marks was 31.000", while the design
distance to the upstream tooling balls was 45.265". Using the
centerline offsets given above together with these lengths led to the
following offset of Region One from the nominal origin. We found
************************************
X offset +0.66 +/- 0.08 mm
Y offset -0.94 +/- 0.08 mm
Z offset -1.40 +/- 0.50 mm
Rotation about X -0.82 +/- 0.05 mrad
Rotation about Y +0.07 +/- 0.05 mrad
Rotation about Z -4.2 +/- 4.1 mrad
************************************
The rotation about Z is given as the average of the upstream and
downstream rotations. Taken on their face, the Z numbers imply a
small helical twist to the detector induced by the mounting
arrangement, albeit one consistent with our mounting accuracy. The
rotations about X and Y are computed from the upstream and downstream
centerline offset; alternate values derived using the tooling ball
measurements were consistent with these, though with larger estimated
uncertainties. The signs of the rotations are defined as positive for
clockwise and negative for counterclockwise about their respective
axes.
The X and Y offsets given above can be rotated into the sector
coordinate system for each sector for the purpose of single-sector
tracking. These numbers are as follows (rotation numbers from Rob
Feuerbach):
Sector "X" offset "Y" offset "X" rot "Y" rot "Z" rot
1 0.66 mm -0.94 mm -.82 mr +.07 -4.2
2 -0.48 -1.04 -.47 -.68 -4.2
3 -1.14 -0.10 +.35 -.75 -4.2
4 -0.66 0.94 +.82 -.07 -4.2
5 0.48 1.04 +.47 +.68 -4.2
6 1.14 0.10 -.35 +.75 -4.2
The raw survey numbers from Appendix I can also be compared with the
nominal design values for those numbers. They are (in millimeters):
Z X Y Z X Y
Survey Survey Survey Nominal Nominal Nominal
Target 0.00 0.00 0.00
Downstream
Sector
1 786.21 80.20 -0.92 787.4 79.375 0.000
2 786.33 40.92 68.27 787.4 39.688 68.741
3 786.07 -38.54 68.70 787.4 -39.688 68.741
4 786.23 -78.67 0.30 787.4 -79.375 0.000
5 786.35 -39.50 -68.80 787.4 -39.688 -68.471
6 786.46 39.92 -69.30 787.4 39.688 -68.471
Upstream
12o'clock -1152.11 1.30 576.27 -1149.73 0.00 576.00
3 o'clock -1151.69 -510.13 -1.22 -1149.73 -508.48 0.00
6 o'clock -1150.69 -0.17 -580.06 -1149.73 0.00 -576.00
9 o'clock -1151.08 511.31 -2.53 -1149.73 508.48 0.00
In conclusion, the survey data for Region One are mostly consistent
with known features of the detector, and internally consistent with
themselves. Numbers given here should be suitable for inclusion in the
RECSIS or SDA analysis packages. Each time Region One is removed and
re-installed in the future, it must be re-surveyed in order to ensure
that the true detector location is known.
_____________________________________________________________________________
APPENDIX I
Raw survey data as transmitted from Kelly Tremblay to Robert Feuerbach:
"The following data reports the location of the Hall B CLAS Region 1
chamber tooling balls and scribe locations relative to the the target on
October 31st, 1997. The target is defined by Jefferson Lab drawing
66210-E-01738 and our groups original cryostat survey (cira 1995). The
data is in millimeters with the origin at the target center, Z axis
follows the beam positive towards the dump, X is in the horizontal
plane, positive beam left, and Y is positive opposite to gravity.
Please note that there is a slight rotation of the entire torus magnet
of 4 millirads, clockwise about the Z axis.
Target Z X Y
Beam Target 0.00 0.00 0.00
DownStream Scribe 1 o'clock 786.07 -38.54 68.70
DownStream Scribe 3 o'clock 786.23 -78.67 0.30
DownStream Scribe 5 o'clock 786.35 -39.50 -68.80
DownStream Scribe 7 o'clock 786.46 39.92 -69.30
DownStream Scribe 9 o'clock 786.21 80.20 -0.92
DownStream Scribe 11 o'clock 786.33 40.92 68.27
UpStream Tooling Ball 12o'clock -1152.11 1.30 576.27
UpStream Tooling Ball 3 o'clock -1151.69 -510.13 -1.22
UpStream Tooling Ball 6 o'clock -1150.69 -0.17 -580.06
UpStream Tooling Ball 9 o'clock -1151.08 511.31 -2.53
The positional accuracy of the points is approximately +/- 0.4 mm in Z, and
+/- 0.15 mm in X and Y. Please contact me if you have any questions.
Kelly Tremblay"
APPENDIX II
Handwritten notes by R.A. Schumacher with the detailed calculations.
.
.
.
Source file: uquark.phys.cmu.edu::[schumacher.region1]survey.txt