Forward Simulations

Forward Simulations Forward Simulations

Jan 24, 2005
w.j. llope
[ PDF Version ] [ HTML Version ]

1 Introduction
2 Geometry
    2.1 Input
    2.2 Configurations
    2.3 Modifications
3 Results
    3.1 Y2004X vs. Y2004Y
    3.2 Y2004/5 vs. Y2006x
Index

1 Introduction

A simulations study concentrating on the very forward areas of STAR near the beam pipe was performed. The main goals of this effort were the following:

a strict comparison of the forward geometry in star (beam pipe, and pVPD [1]) as defined via "*.g" files in the STAR simulations framework to CADD drawings obtained from STAR STSG [2,3], and any revisions to the existing star geometry needed to make the two geometry definitions consistent.
the definition in the simulations framework of a number of inactive metal pieces near the beam line that were previously missing from the simulations geometry definitions. The proper inclusion of these pieces could increase the backgrounds seen in any pVPD, PMD, or FPD simulation.
the definition of new simulations geometry configurations which replace the pVPD of Runs 2-5 with a new, larger-area detector (here called the upVPD), which is needed to improve the start time efficiency in p+p and light-ion reactions for the large-area STAR TOF system [4].
the geant simulation of minimum bias p+p and Au+Au events through these updated or new geometries to compare the performance of the different designs for the upVPD.

Section 2 describes the changes made to the present STAR geometry to match the STSG CADD files, and a few possible designs for the upgraded pVPD. Section 3 describes the effects of the matching to the CADD geometry and the addition of all the missing metal. Also described in this section is the performance comparison pVPD vs. upVPD.

2 Geometry

2.1 Input

The present STAR simulations framework takes the geometry information from xxxxgeo.g files, and parses it using Pavel Nevski's agi into geant-standard volume definitions in ForTran. It is not clear if there has ever been a strict comparison of the size and positioning of the volumes so defined to the actual hardware as it exists now in the WAH. Such a comparison was performed here for the very forward regions of STAR outside the pole-tips and near the beam pipe, i.e. at (|Z|,r) positions in the range (400-600cm, 5-50cm).

Figure 1: A view of the CADD file that was used to define the hardware that supports the I-beam from the balcony, and that supports the pipe from the I-beam. [EPSF] [JPEG] [PDF]

Figure 2: A views of the updated CADD file that was used to position the pieces seen in figure 1 with respect to the STAR coordinate system. [EPSF] [JPEG] [PDF]

Two sources of information were combed for the details on the actual STAR geometry. CADD files from STSG [2] were the primary source of information. These were opened in CADD programs allowing one to pick off any dimension or distance. Some discrepancies in these files were found, which were solved by direct measurement [3] followed by an update to the CADD files [2]. Following these discussions, the updated CADD files were then used to make 1:1 comparisons to the simulations geometries.

The two most important views from the CADD files are shown in Figures 1 and 2. The upper one, figure 1, shows multiple views of the beam pipe and support structure. With the exception of the 4" I-beam itself, none of this hardware exists in the simulations geometries. Of the missing hardware, perhaps the most important are the two sets of brackets near |Z| ~ 4 m, which clearly can affect the particle production into the pVPD, PMD and FPD. Most of this missing hardware is Aluminum, although also missing are a large number of bolts, which are assumed to be stainless steel. Some of the aluminum pieces, for example the vertical sides of the pipe support brackets, are several inches thick with respect to the typical particle trajectories in this region. This upper view was thus used as the primary source of information on the dimensions and positioning of the various pieces of the beam pipe support structure and brackets. The lower image, figure 2, depicts a side view of STAR as viewed looking North, following an update based on the recent measurements [3]. This lower view was used primarily for the positioning of the pVPD and the support structure relative to the magnet steel, hence relative to the STAR coordinate system. The magnet is assumed to be centered at Z=0, and to extend in |Z| to 11' 10 3/4", or 362.6 cm.

Figure 3: Photographs of the actual beam pipe and pipe support structure that were used to check the accuracy of the new simulations geometries.

The most important distances and dimensions picked off the CADD files are listed in table 1. With the exception of the last line (length of the I-beam), the equivalent volumes in the STAR-standard simulations geometry have different dimensions or positions. For example, the I-beam in the simulations is too close to the magnet by ~ 4". Numerous small changes to the beam-pipe simulations geometry are also necessary apparently.

Table 1: The more important measurements picked off the CADD files [2]. The |Z|-position of the magnet steel is 11' 10 3/4" [362.6 cm].

Quantity	West	East

DZ, magnet steel to start of I-beam ^f	12 1/8"	12 7/8"
DZ, magnet steel to pVPD front face	78"	77 7/8"
DZ, magnet steel to cross-piece	88.125"	88.26"
Z-extent, 3" pipe to 5" pipe neck-up	10.69"	10.54"
Z, pVPD front face ^f	220.72"	220.59"
DZ, start of I-beam to Z-center Bracket #1	1.85"
DZ, start of I-beam to Z-center Flange #1	4.51"
DZ, start of I-beam to Z-center Flange #2	12.37"
DZ, Z-center Flange #2 to end of neck-up	11.54"
DZ, start of I-beam to Z-center Bracket #2	19.59"
Z-extent, I-beam	116.1"

^f Recently re-measured [3]. ^f Values shown for this quantity are specific to Run-5.

The second source of information used here, primarily to confirm the geometry described by the CADD files, was photographs of the actual hardware. A few of the more useful of these are shown in Figure 3. The upper photo shows the East side of STAR during Run-1, which gives an excellent view of the pipe support I-beam and the I-beam support structure as viewed from a balcony. The two photos in the middle row show the East and West sides, respectively, of STAR during Run-3. The two photos in the lower row show the East side of STAR during Run-4 from two different angles (behind and below, respectively). The middle left and lower left photos indicate certain objects that are not only not found in the simulations geometry but are also not seen in the STSG CADD files.

2.2 Configurations

Given the input described in section 2.1 above, local versions of the STAR simulations geometry files were produced. The changes made to the existing files to make them consistent with the CADD files and the photographs, as well as the extensions to these files to define new geometries for the upgraded pVPD, are described in this section.

The various detector geometry options are controlled via the geometry.g file after the user-specification of the configuration name. The configuration names used here are the following.

Y2004X - This is the current library version of the STAR "standard" geometry configuration. This is unchanged in the local code with the following exceptions. First, a bug was found in the orientation of the "boats" for two of the three pVPD PMT assemblies one each side, which was fixed. Second, the pVPD FEE box is mounted onto a side-strut of the pVPD in Runs 4 and 5, whereas the Y2004X pVPD geometry places it under the pVPD base plate (as it actually was in Runs 2 and 3). This was fixed. Third, the thickness of this (Aluminum) electronics box was incorrect - it was 0.25" but should be 50 mils. Other than these trivial modifications, Y2004X should be considered as a close reflection of the forward geometry presently used throughout STAR. Note also that in this "standard" geometry, the only component to the pipe support structure that is defined in the simulations is the 4" I-beam itself.
Y2004Y - This version makes additional changes to the Y2004X configuration in two major areas. The first of these is the new definition of the components of the pipe support structure. These components include the "cross-piece" underneath the I-beam, the diagonal struts of 3" aluminum angle that support the cross-piece from the balcony, the vertical pieces of aluminum angle at the highest-|Z| extent of the I-beam, as well as the brackets and bolts that hold the pipe in place with respect to the I-beam that are near the lowest-|Z| extent of the I-beam. Also defined is the long threaded rod that holds the pipe support structure in place laterally with respect to the pole tip. This rod does not exist in the latest STSG CADD so a position and diameter (3/8") for this rod was invented based on the middle left image of figure 3. The second group of changes involved modifications to the positioning of all of the forward hardware with respect to Z-position of the magnet steel to match the (updated) CADD files. To make the simulated hardware match up with the CADD and the photographs, it was necessary to change geometry parameters for the I-beam, the pVPD (both in vpddgeo.g), and the beam pipe (in pipegeo.g). Also implemented for the first time in this configuration (and the later ones below) is the 3/4" offset between the East and West balconies with respect to STAR that I discovered about a year ago. This offset was recently confirmed [3] and subsequently included into the STSG's CADD files (figure 2 shows the updated version).
Y2005 - This version is the same as Y2004Y, except for a single change. The pVPD Z-positions were changed to match STAR as it exists right now during Run-5. In the previous runs, the pVPD was positioned symmetrically with respect to the I-beam support structure, specifically the "cross-piece."¹ For Run-5, STSG positioned the pVPD equidistantly East and West from a piece on the BBC [5], and the pVPD Z-positions used in this configuration are those that were directly measured with respect to the BBC following the pVPD installation for Run-5.
Y2006A - This version is the same as Y2004Y or Y2005, except that the pVPD, with three 2" PMTs inside 3" O.D. steel shields, is replaced by detector with more channels in the same integration volume. Assumed for this detector are 1.5" PMTs (e.g. the Hamamatsu R5946 mesh-dynode PMTs used in TOFp and the CTB) inside 2" O.D. thin-walled Aluminum cylinders. In the Y2006A configuration, nineteen (19) such detector assemblies are placed in a non-cylindrically symmetric arrangement due to the presence of the pipe support I-Beam just under the pipe.
Y2006B - This version assumes the same 2" OD detector assemblies but places twenty-two (22) of them cylindrically-symmetrically around the pipe in two rings. This was achieved by lowering the I-beam with respect to the pipe by 2.25". The present pipe support brackets (see the middle left and lower left images in figure 3, or figure 5) seem capable of allowing such a shift without modification.²
Y2006C - This version assumes the same 2" OD detector assemblies but places thirty-three (33) of them cylindrically-symmetrically around the pipe in three rings. This was achieved by lowering the I-beam with respect to the pipe by 3.25". The present pipe support brackets seem capable of allowing such a shift with the only modification being the replacement of the present vertical bolts with longer ones. Note that the detector assemblies in all designs above have long axes parallel with the beam axis, while the third ring in the Y2006C geometry is at a radius R ~ 17cm. A Y2006D geometry was built and tested for which the upVPD Pb layers were placed in the same (R,Z)-positions but the positioning of these detector assemblies included a phi-dependent angle to "point" the detectors at Z=0. The Y2006D results were indistinguishable from the Y2006C results.³

Figure 4: Geant renderings of various views of the new simulations geometries; from upper left to lower right: Y2004X, Y2004Y, Y2005, Y2006A, Y2006B, Y2006C.

Multiple views of these six modified or new configurations are shown in Figure 4 - the views of Y2004X are in the upper left frame, while the views from Y2006C are seen in the lower right frame. In each frame, a side view of the entire beam pipe and support structure is shown at the top. Just below this is a close-up of this hardware on one side (the East). Just below this and to the right is a side view of a detector assembly. At the bottom left is an isometric hidden-line view, and on the bottom right is the view along the pipe.

The comparison of the upper two frames of Figure 4 indicates that both the beam pipe's "transition region" and the pipe support I-beam have been moved outwards in |Z| in the present versions of the simulations geometry in order to match the CADD files. The new materials added to the forward geometry (I-beam support pieces, pipe support brackets, etc.) can be seen by comparing Y2004X (upper left) with any of the other views.

Figure 5: Geant renderings of the beam pipe transition area and support structure in the new simulations geometries; from upper left to lower right: Y2004X, Y2004Y, Y2005, Y2006A, Y2006B, Y2006C.

Close-up views of the beam pipe transition region for these six modified or new configurations are shown in Figure 5 - the view from Y2004X is seen in the upper left frame and that from Y2006C is seen in the lower right frame. The comparison of the upper two frames of this figure again reflects the local modifications to the beam pipe and the I-beam positioning. Also apparent in these first two frames are the changes made to the (steel) bellows in between the two pipe flanges. Beginning with Y2004Y, one also notices the appearance of the pipe support brackets and related bolts. While the brackets are aluminum, in some areas they are are several inches thick as seen by forward-going particles. The bolts, while not "thick," are however not aluminum but stainless steel, which is approximated in the geometry files by iron.

2.3 Modifications

This section tabulates the specific changes made to the simulations geometry files to try to better match the CADD files and the photographs. Changes were made to the beam pipe (in pipegeo.g), to the pVPD and pipe-support structure (both in vpddgeo.g), and for local configuration control (in geometry.g). The PMD or FPD |Z|-positioning was not modified in the local geometry, but this should be checked. Especially in the case of the pipe, the changes made may not be unique - i.e. I needed to move the flanges out, and did so by arbitrarily picking the subsection of the pipe to extend. So, this document, and especially this sub-section, is basically a first stab at something that will be fine-tuned over the coming months with wider discussion and a few more direct measurements.

Table 2: Changes made to pipegeo.g to better match the simulations geometry to the CADD and photographs.

Bank	Variable	Y2004X	Y2004Y	\|D\|

PIPG	S1Leng (cm)	153.4	164.2	10.8
	S2Leng (cm)	18.0	15.7	2.3
	S3Leng (cm)	1.0	0.5	0.5
	ConeLen (cm)	12.5	13.7	1.2
	RibNum	8	20	12
	RibCent (cm)	N/A	5.0	N/A

Table 3: Changes made to vpddgeo.g to better match the simulations geometry to the CADD and photographs.

Bank	Variable	Y2004X	Y2004Y	\|D\|

VPDG	pVPD West Z	573.57	568.1
	pVPD East Z	-574.57	570.0
	I-Beam West Z	530.48	540.45
	I-Beam East Z	530.48	-542.35
	BXthick	0.635	0.127

Table 4: The pVPD (mother) Z position versus the RHIC Run number. The left two columns indicate the West and East, respectively, positioning as it exists in the present STAR CVS library, while the two columns on the right indicate the values that result from the matching to the CADD files.

Run	In CVS:		Should be:
	Z_west (cm)	Z_east (cm)	Z_west (cm)	Z_east (cm)

2	563.1688	561.2638	556.7 (est.)	556.6 (est.)
3	563.4069	564.4388	557.0 (est.)	559.9 (est.)
4	573.5669	574.5688	568.1	570.0
5	N/A	N/A	583.5	583.8

3 Results

The results from the simulation of collision events through the modified geometries are described in this section. The comparison of the Y2004X and Y2004Y geometries is done in subsection 3.1. This shows the results of the changes made to the simulations geometry following the comparison to the CADD files, and of the addition of the previously missing inactive metal of the pipe support structure. The comparison of the performance, specifically the detector efficiency per event, of the upgraded pVPD designs to that of the present pVPD using the full simulation of collision events is then described in subsection 3.2.

The simulations were performed starting with the latest version of the STAR geometry, i.e. "*.g, files, and the local modifications described in the previous section were done. The analysis shell used to compile and "make" the geometry, read in the evgen information, and perform the full gstar simulation through the STAR geometry was STaF. This was mainly because STaF allows a user-routine to be called inside the geant event loop. Given the tough competition for the scarce RCF resources, writing large RZ files containing mostly unused information and then reading these into root to perform the simple user analysis seemed excessive and impolite. So instead a local "pirate" version of STaF was built for the latest RedHat and GNU compiler versions now supported and in use at the RCF.⁴ With this shell, only the few (tens) of quantities needed from each simulated event were either simply looked up or were calculated in the user-routine and then saved directly as part of the event loop. This reduced my huge factors the disk space and operator-hours needed to complete these simulations. An additional advantage of using STaF for these simulations was that the user codes developed for earlier but very similar simulations [6] could be implemented immediately, saving quite some time in the code development.

Like the previous simulations, the present ones pay considerable attention to the definition of a "hit". The details are described in Ref. [6], but the basic idea involves "full-stacking" to allow the complete parentage tree of any particle resulting in an energy deposit in the (u)pVPD scintillator to be available to the user-routines and traversable by them. This parentage information is used extensively in section 3.1 below.

The STAR-standard Pythia was run minimum bias (MSEL 2) with no P_T limits nor thresholds and pseudo-rapidity limits of �8. The minimum bias Au+Au events from Hijing are the STAR standard set.⁵ Approximately 100k p+p, and 3k Au+Au, events were run through each of the six geometries described in section 2.2. The primary vertex was smeared with Gaussian standard deviations of 40cm(15cm) in the p+p(Au+Au) events.

3.1 Y2004X vs. Y2004Y

The previously missing pipe support hardware must increase the production of (simulated) secondaries into the STAR forward detectors. The cleanest "before & after" comparison of the effects of all the new metal is to simply compare the results from Y2004X to Y2004Y. Such comparisons, which also indicate the changes to the geometry files needed to make them consistent with the STSG CADD, are discussed in this subsection.

Figure 6: The number of pVPD hits per minimum bias p+p event using the Y2004X geometry (blue) and the Y2004Y geometry (magenta), plotted as a function of the Z-value of the geant parent vertex for the particle resulting in a pVPD hit. The inset shows a close-up of the region 380cm < |Z| < 470cm (West). Each bin in either frame is 2cm wide.

Shown in Figure 6 is the number of pVPD hits per Au+Au event using the Y2004X geometry (blue) and the Y2004Y geometry (magenta), plotted as a function of the Z-value of the geant parent vertex for the particle resulting in a pVPD hit from Z=-600cm (East) to +600cm (West). The peak at Z=0 corresponds primarily to charged primaries which proceed mostly-unaffected through the geometry and then leave energy in the scintillator layer of the pVPD detectors. For these hits, the parent vertex is the primary vertex. This is supported by the observation that the fitted width of this peak for the Y2004X geometry is 14.7�5cm which agrees with the gaussian width specified for the primary vertex smearing in the simulations of 15cm.

The two wider bumps near |Z| ~ 250cm correspond primarily to hits in the pVPD detectors from particles produced in the beam pipe. The region 380cm < |Z| < 470cm is the pipe "transition" region (see figure 5) which includes the two sets of pipe flanges, the steel bellows, the 3" to 5" "neck-up", the beginning of the pipe support I-beam, and for Y2004Y and later, the pipe support brackets and bolts, and the I-beam support pieces including the horizontal support rod. This region is also shown for the West side in the inset to 6. The very large peaks at |Z| ~ 550cm correspond to hits produced in the Pb layer immediately in front of the pVPD scintillator layers primarily by the conversion of energetic primary photons.

One notices significant differences between the Y2004X and Y2004Y geometries in two regions. Near the pVPD itself (|Z| ~ 550cm) one can see the forward shift of the pVPD for Run-4 that resulted from the comparison to the CADD files. The transition region is clearly quite different. The two beam pipe flanges have been move outwards in |Z| going from Y2004X to Y2004Y. Also, there are in general more pVPD hits resulting from this region as a result of the addition of the pipe support hardware.

Figure 7: The (x,y) positions of the (secondary) production vertices for particles resulting in (u)VPD hits. The upper row is for the Y2004X geometry (star default, pipe support structure consists only of the I-beam), and the lower row is for the Y2004Y geometry (which includes the newly defined pipe support hardware). The left frames include all secondary vertices leading to (u)VPD hits, while the middle frames require these secondary vertices to be in the range 3.8m < |Z_SV| < 4.7m, and the right frames require 4.7m < |Z_SV| < 6m.

Figure 8: The ratio of the number of pVPD hits in the Y2004Y geometry to the number of pVPD hits in the Y2004X geometry as a function of the |Z| position of the (secondary) vertex at which the particles leading to pVPD hits were produced. The blue(green) points correspond to the West(East) side of STAR. Each bin is 20cm wide.

This is seen in Figure 7, which depicts the (X,Y) positions of the parent vertices for particles resulting in pVPD hits. The upper row is for the Y2004X geometry, and the lower row is for the Y2004Y geometry. The left frames include all secondary vertices leading to (u)VPD hits, while the middle frames require these secondary vertices to be in the range 3.8m < |Z_SV| < 4.7m, and the right frames require 4.7m < |Z_SV| < 6m. The new pipe support hardware in the transition region is clearly "imaged" in the middle lower frame.

Shown in Figure 8 is the ratio of the number of pVPD hits in the Y2004Y geometry to the number of pVPD hits in the Y2004X geometry, plotted as a function of the |Z| position of the geant vertex at which the particles leading to pVPD hits were produced. The blue(green) points correspond to the West(East) side of STAR. This figure indicates that the number of pVPD hits resulting from primaries directly or from secondaries produced in the beam pipe, (i.e. |Z|[ < || ( ~ )]300cm in this plot) are very similar. The major dip near |Z| ~ 400cm occurs because the two pipe flanges on each side were moved outwards in |Z| from Y2004X to Y20004Y. The major peak at ~ 430cm includes these flanges in the new positions consistent with the CADD files, as well as the newly-defined pipe support brackets and bolts.

In 3k minimum bias Au+Au events, the simulations resulted in 402,668(462,332) pVPD hits in the Y2004X(Y2004Y) geometries. The newly added I-beam and beam pipe support structures thus result in an increase in the number of pVPD hits per event of 15%, which is not an insignificant increase.

3.2 Y2004/5 vs. Y2006x

The three "2006" geometries replace the pVPD with another detector (upVPD) which has more, but lower diameter and much lower mass, detector channels. Such an increase in the number of channels in the start detector is needed to increase the start detector's efficiency for providing start times to the full TOF system in low-multiplicity events such as those from p+p, asymmetric ion, or peripheral heavy-ion, collisions. The performance comparison Y2006x upVPD vs. Y2004/5 pVPD is described in this subsection.

Figure 9: The efficiency per event of the various designs of the (u)pVPD versus the impact parameter in femtometers from the full simulation of minimum bias Au+Au events from hijing. The vertical line at 14fm is twice the hard-sphere radius for Au.

Figure 10: The efficiency per event of the various designs of the (u)pVPD versus the Z-value of the primary vertex from the full simulation of p+p events from pythia. The efficiency for at least one hit (east or west) is shown in the left frame, while the efficiency for at least one hit on both the east and west is shown in the right frame.

Figure 11: The same as figure 10 but including the requirement that there was at least one hit in both the East and West BBC detectors in the same event. The yellow band depicts the values observed experimentally during the p+p phases of RHIC Runs 2 and 3.

Shown in Figure 9 the start detector efficiencies per event versus the impact parameter, b, in minimum bias Au+Au collisions for the different (u)pVPD geometries as labelled across the top of the figure. In the left frame is the start detector efficiency per event using the condition that there was at least one hit on either the East or West sides, and on the right is the efficiency per event using the condition that there was at least one hit on both the East or West sides. As has been seen experimentally in Runs 2 and 4, the pVPD efficiency (bluish points) is excellent in Au+Au collisions, and only tails off for extremely peripheral impact parameters. The various upVPD designs (reddish points) are of course also perfectly efficient for b[ < || ( ~ )]11fm, and have a slightly improved efficiency for more peripheral collisions.

Shown in Figure 10 is the efficiency per event of the various designs of the (u)pVPD versus the Z-value of the primary vertex from the simulated p+p events. The two frames correspond to the same two "local trigger" conditions as used in figure 9. No significant Z_vtx dependence in these efficiencies is observed. The efficiencies per p+p event increase significantly when comparing the pVPD geometries (bluish points) to the upVPD geometries (reddish points), due to the larger number of detector channels in the upVPD geometries. The Y2006C upVPD geometry with 33 detectors in a cylindrically-symmetric arrangement has a 10-15% better efficiency per event than the Y2006A upVPD geometry of 19 detectors in a non-cylindrically-symmetric arrangement.

Shown in Figure 11 are the same efficiencies as shown in figure 10. In this figure, however, only p+p events that result in at least one hit in each of the East and West halves of the BBC are included. This sample of events is expected to more closely match those that STAR collects during RHIC p+p running, as such a condition on the BBC is typically required at Level-0 during data-taking. For these pseudo-"STAR-triggered" events, the (u)pVPD efficiencies are significantly higher than those for the truly minimum bias p+p collisions used to make figure 10. The upVPD efficiencies for the " � 1.or. � 1" local trigger (left frame) exceed 80% for all three upVPD designs. It's ~ 90% for the Y2006C design. The efficiencies for the " � 1.and. � 1" local trigger (right frame) exceed ~ 35% for all three upVPD designs. It's ~ 45% for the Y2006C design.

The yellow band indicates the local trigger efficiencies observed from the actual pVPD in the p+p phases of Runs 2 and 3. The experimental values are in the range of 8-12% per STAR event, which presumably would have been triggered via the CTB in Run-2 and both the CTB & BBC in Run-3. The present simulations for the pVPD configurations (bluish points in the right frame) are consistent with this observation once the BBC requirement is included in the simulations to simulate the Level-0 trigger.

References

[1]: "The TOFp/pVPD time-of-flight system for STAR," W.J. Llope et al., Nucl. Inst. and Methods, Section A, 522, 252 (2004).
[2]: J. Scheblein, private communication.
[3]: T. Krupien and B. Soja, private communication.
[4]: "Proposal for a Large-Area Time-Of-Flight System for STAR," (The STAR TOF Group), http://wjllope.rice.edu/^~TOF/TOF/Documents/TOF_20040524.pdf
[5]: R.L. Brown, private communication.
[6]: W.J. Llope, John Mitchell, and F. Geurts, "pVPD Simulations", Star Note 0416, March 6, 2000.

Footnotes:

¹Hence the 3/4" East/West balcony offset is visible in the pVPD Z-positions defined in vpddgeo.g based on the configuration name.

²Such a shift could cause an interference with the PMD. If this turns out to be the case, one could imagine simple revisions to the I-beam itself to clear such a conflict.

³The outer ring of detectors in Y2006C only needs to be rotated radially by 1.7 degrees to point to Z=0, hence all start detector assemblies can simply be parallel to the Z-axis for mechanical simplicity without significant loss of geometrical acceptance.

⁴Thanks to Maxim Potekhin for the Makefiles even though STaF has been unsupported in STAR for about a year now.

⁵On the RCF, they're in /star/simu/evgen/auau200/hijing_382/b0_20/done/.

File translated from T_EX by T_TH, version 3.38.
On 24 Jan 2005, 14:12.

Contents