Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stecf.org/conferences/adass/adassVII/reprints/heh.ps.gz Äàòà èçìåíåíèÿ: Mon Jun 12 18:51:46 2006 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 02:23:59 2012 Êîäèðîâêà: IBM-866 Ïîèñêîâûå ñëîâà: ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ð ï ï ï ï ï ï ï ï ï ï ï ï ï ï ï ï

Astronomical Data Analysis Software and Systems VII
ASP Conference Series, Vol. 145, 1998
R. Albrecht, R. N. Hook and H. A. Bushouse, e
ó Copyright 1998 Astronomical Society of the Pacific. All rights reserved.
ds.
ASC Coordinate Transformation --- The Pixlib Library, II
H. He, J. McDowell and M. Conroy
HarvardíSmithsonian Center for Astrophysics
60 Garden Street, MS 81
Cambridge, MA 02138, Email:hhe@cfa.harvard.edu
Abstract. Pixlib, an AXAF Science Center (ASC) coordinate library,
has been developed as the continuing e#ort of (He 1997). Its expansion iní
cludes, handling of the High Resolution Mirror Assembly (HRMA) Xíray
Detection System (HXDS) stage dither and the fiveíaxis mount (FAM)
attachment point movements, correction of misalignments of the mirror
mount relative to Xíray calibration facility (XRCF) and to the default
FAM axes, as well as solution of sky aspect o#sets of flight, etc. In this
paper, we will discuss the design and the configuration of the pixlib sysí
tem, and show, as an example, how to integrate the library into ASC
data analysis at XRCF.
1. Introduction
The work of He (1997) established a preliminary framework for the pixlib sysí
tem, including the parameteríinterface data I/O structure, matrix calculation
algorithm, and coordinate transformation threading baselines. Since then, the lií
brary has undergone thorough reíorganization and expansion to meet the AXAF
onígoing requirements of both ground calibration and flight observation. At the
time of writing, the library is about 95% completed with approximate 6000
source lines of codes. It was successfully integrated and built during the XRCF
calibration phase.
In this paper, we will highlight the system design and architecture of the
library, complementary to the early work, and describe the system configuration
in terms of user application. The complexities of coordinate transformation at
XRCF and the resolutions will be discussed.
2. Architecture of the ASC Coordinate Library
The building blocks of the Pixlib library are three subísystems, core, auxiliary,
and application interface (API), and the foundation of the library is built with
the parameteríinterface structure. Figure 1 sketchs the architecture of the lií
brary.
As discussed in He (1997), the design of pixlib is modular to allow system
expandibility, easy maintenance and simple ways to incorporate new scientific
knowledge. The core subísystem, which includes 8 modules (see Figure 2 for
208

ASC Coordinate Transformation --- The Pixlib Library, II 209
Application Interface
Core Modules
Parameter Files
Modules
Auxiliary
Figure 1. Pixlib library architecture, constructed on three subí
systems which are layered on the parameterífile structure.
details), builds the ASC coordinate frameworks of grating, FAM motion, sky
aspect o#sets, telemetry (raw) reads, detectoríchip tiling, and coordinate transí
formation among chip pixels and celestial angles. Because of the common needs
of generic data sources, handy utilities, moduleítoímodule communication, etc.,
the library is supported with a 4ímodule auxiliary subísystem, as shown below.
pix_errstatus.c íí error handling
pix_utils.c íí utility functions
pix_common.c íí common data sources to all modules
pixlib_hiden.c íí internal configuration, bookkeeping
The upperílevel interface of the library is implemented in the module pixlib.c,
which distributes functions between the lowerílevel modules. pixlib.c, in large
part, provides setup functions for system configuration, and other API functions
are implemented locally without the need for crossímodule function calls. All
the API functions are identified by the ``pix_'' prefix.
The data inístream of the parameteríinterface approach simplifies system
configuration and data readability. The number and organization of those data
files have remained almost same as described in He (1997) with few updates.
pix_pixel_plane.par, substituting the original pix_size_cnter.par, groups
2íD pixel system parameters of focal plane, tiled detector, grating dispersion
together; pix_grating.par is added to define dispersion period and angle of
grating arms.
3. System Configuration
Prior to application program execution, the library needs to be configured propí
erly. The system configuration is optionally either static or dynamic, as illusí
trated in Figure 2. A setíparameterívalue to a parameter file, pix_coords.par,
handles the static configuration and the user can set values for the following
parameters.

210 He, McDowell and Conroy
tdet cpc2stf stf2tpc grating
tpc2src aspect fam
Configuration
pix_init_pixlib()
API
Parameter Files
raw
call pix_set_*
(override pset)
pset pix_coords.par
(default)
static
setups
dynamic
setups
Figure 2. Pixlib data flow and system configuration. Each smaller
box above represents a module. For example, the ``raw'' and ``cpc2stf ''
boxes denote the modules of pix raw.c and pix cp2stf.c, respectively.
flength = Telescope focal length in mm
aimpnt = Name of aim point of detector
fpsys = Focal Plane pixel system
tdetsys = Tile Detector pixel system
gdpsys = Grating Dispersion pixel system
grating = Grating arm
align = FAM misalignment angle in degrees(pitch, yaw, roll)
mirror = mirror misalignment angle in degrees(pitch, yaw, roll)
In the course of the system initiation, executed through pix_init_pixlib(), iní
ternal functions lookup the parameter table to parse the information down to
relevant modules, which are then configured accordingly.
An alternative way to configure the system is to make function, ``pix_set_*'',
calls in application program following the initiation. pix_set_flength(int),
for instance, is equivalent to the pset ``flength'' for pix_coords.par, and
pix_set_fpsys(int) to the pset ``fpsys'', to name a few. The consequence of
those calls is to override the static configuration which is the system defaults.
4. Coordinate Transformation at XRCF
Coordinate transformations at XRCF need to be carefully handled when the
FAM feet move and the HXDS stage dithers. In the default, boresight configuí
ration the FAM axes are parallel to the XRCF (and Local Science Instrument,
LSI) axes, but they may undergo some movements in addition to the HXDS

ASC Coordinate Transformation --- The Pixlib Library, II 211
stage dithering and possible mirror mount movement. Therefore, those e#ects,
as listed below, must be accounted for before coordinate transformations beí
tween focal plane and LSI system are made:
. misalignments of the default FAM axes from the mirror due to FAM atí
tachment point motion,
. motion of the HXDS stage from the default FAM reference point,
. possible misalignments of the mirror mount relative to the XRCF preí
scribed by pitchíyaw angles,
. misalignments of the FAM axes from the default due to FAM feet motion.
The following two functions, in addition to other generic configurations,
e#ectively supply the system configuration for coordinate transformation at
XRCF. The routine
pix_set_mirror (double hpy[2], /* in degrees */
double stage[3], /* in mm */
double stage_ang[3]) /* in degrees */
corrects misalignment from the mirror axis by measuring its displacement from
the boresight configuration of the default FAM frame (stage_ang) for a given
mirror position (hpy) in mirror nodal coordinate system. The hpy is measured
in HRMA pitch and yaw axes, and the HXDS stage position (stage) monitored
relative to the default FAM reference point. The routine
pix_set_align(
double mir_align[3], /* (yaw, pitch, roll), in degrees */
double stg_align[3]) /* (yaw, pitch, roll), in degrees */
serves to assess
. misalignments of mirror mount (mir_align) relative to XRCF axes are
measured in the given yawípitchíroll Euler angles in the mirror nodal coí
ordinate, and
. misalignments of the default FAM (stg_align) relative to XRCF axes are
corrected in terms of yawípitchíroll Euler angles in the default FAM frame.
The system configuration above was successfully applied to and integrated
into ASC data analysis during the Xíray calibration.
Acknowledgments. We gratefully acknowledge many fruitful discussions
with ASC members. This project is supported from the AXAF Science Center
(NAS8í39073).
References
McDowell, J., ASC Coordinates, Revision 4.1, 1997, SAO/ASCDS.
He, H., McDowell, J., & Conroy, M., 1997, in ASP Conf. Ser., Vol. 125,
Astronomical Data Analysis Software and Systems VI, ed. Gareth Hunt
& H. E. Payne (San Francisco: ASP), 473.