For many of the muon tracker alignments, the structural support of the system will be used to hold the stations in place to within the alignment tolerances (relative to the interaction point, magnet, etc.) and the exact placements of the tracking stations will be determined by a survey at installation time. For the internal station-to-station alignment in the x-y plane, however, we will not only survey the positions at installation time, but also plan to actively monitor the positions with line-of- sight monitors after installation because of the tight alignment tolerances.
The station-to-station alignment requirements in z of 0.4 mm should be determined by a survey with respect to the nominal interaction point, as specified by a RHIC or PHENIX survey mark which is used by the whole PHENIX detector. If these alignments can be obtained, then the alignment of the tracking stations to the vertex (or MVD), the muon magnet, and between the two tracking arms should be automatically provided at the same time.
The placement of the tracking chambers in the x-y plane must
meet the global alignment requirements, and must be good enough
for the active alignment monitors to fall into a line-of-sight
so that a final alignment of 25 
m can be achieved. Since
most active alignment monitor systems that we have looked at
are capable of covering a displacement range of 1 to several
millimeters, and the global alignment requirements are also on the order
of 1 to a few millimeters, we will require the chamber positions
to be aligned in the x-y plane relative to a global survey mark
to within at least
1 mm. This must be a common survey point which is used by both
muon arms, the muon magnet, and the MVD. To maintain the
alignment of the outside fiducials and/or alignment monitors
to the internal cathode strips, tight tolerances will be
required during assembly of the chambers. The tolerances
that will be maintained have been covered in the muon
note PHENIX-MUON-95-9, by Dave Lee.
The line-of-sight (LOS) monitoring system which we are most actively pursuing uses several LED-lens-detector systems, with the LEDs mounted at one station, the lenses mounted at a second station, and the detectors mounted at the third station. Previous work by the GEM muon system (GEM TN-92-202) has shown that six separate LOS monitors on a chamber are adequate to measure all displacements and rotations which would contribute to an error in the phi measurement. If eight monitors are used, then distortions caused by gravitational sag can also be corrected for. The six monitors would be placed at the four corners of a chamber with two additional monitors on the middle of the inner and outer radius of the chamber, and the eight monitor system would add two more monitors to the sides of chamber. This system would require that the muon tracking volume have several lines-of-sight which extend from station 1 to station 3. For monitors near the piston, this can only be achieved by providing holes in the station 2 chamber frames where lenses could be accurately placed. These lines of sight would also need to be extended through the support structure that is used to hold station 2 in place. Lines of sight at the outer radial edge of station 2 can be provided again by holes in the chamber frames or by making sure that an adequate gap is provided between the outside edge of the chamber frames and the magnet lampshade.
If the lines-of-sight can not be adequately provided, an alternative
alignment system that could be used would be to separately align
stations 1 and 2 and
stations 2 and 3 using similar light source, LED, and detector systems
or using a light source, reflector, detector system. All of the light
source-detector systems that we have investigated have been able to
rather easily achieve displacement measurements of 10 
m or better
over several meters of distance.