Integration
of the OMC and WFS at the ATC
wht-naomi-72
Document number
AOW/GEN/RAH/9.0/02/98/OMC/WFS Integration
Version date: 24 February
1998
This document is intended to
provide guidance in selecting the approach to the integration of the OMC and
the WFS at the ATC. Many of the points presented may be obvious and possibly
trivial but the author hopes that their documentation will help in arriving at
a decision.
As a starting point the two
extreme approaches to the integration will be addressed. At one extreme the
minimum possible set of tests would be performed. These tests would verify that
the OMC and WFS perform all basic functions, transmit images through the entire optical train and produce correct
output data These data would include
signals indicating positions and states of motorised components together with
WFS camera data and signals from the pre-correction camera. WFS camera data
would be evaluated for the correct format in its range of operating modes but
there would be no real-time image processing . Tests assessing measurement
accuracy, reliability, sensitivity to thermal changes and vibration would not
be performed at this stage. Only problems that
would absolutely prohibit one from progressing to the next stage of
testing would be corrected during this phase. The philosophy here is to proceed
to complete system testing at Durham as rapidly as possible. Extensive testing would be performed at the system
level and correction of problems would be performed in parallel.
At the other extreme one
would perform thorough testing to uncover as many faults as possible; this
approach might involve shipping part of the RTCS to the ATC. It would certainly
involve the use of some auxiliary equipment and software to process output data
from the WFS. All faults would be corrected before shipment to Durham.
Proponents of one of these extreme
approaches may well regard the other approach as ridiculous but , depending on
the circumstances, either approach may be used. There is, of course, the option
of choosing middle ground.
The choice of the integration
approach depends on several factors. These include at least the following:
1. The nature of the problem,
i.e. software, mechanical, electrical, environmental, system control, etc.
2. The level at which the
problem can be first identified, i.e. subsystem testing, OMC/WFS integration at
the ATC or system integration at
Durham.
2. The time and effort
required to correct the problem.
3. The availability of the
required resources.
4. The probability of
schedule delays elsewhere in the project (e.g. in software development) and
their effect on the adopted plan.
Risk assessment can help
in evaluating these factors but in
reality one seldom has the time or money to perform a thorough risk assessment.
Given NAOMI’s present circumstances,
i.e. time limitations and limited risk
assessment, one can do little more than
perform a cursory evaluation of the
above factors. The consequence may be that
one will encounter problems that were completely unanticipated.
The first approach may be quite acceptable if one is confident
that most problems will occur during system integration, e.g. software problems
associated with system control, and that these problems could not be readily
identified at an earlier stage. The assumption here is, of course, that one is
in a position to proceed rapidly to system integration. There is the risk that
during system integration one may have to ship a component or even the WFS back
to the ATC to correct a problem or send ATC personnel to Durham. However with
this approach one should allow for such contingencies. An advantage is that one can correct problems in
parallel. Any problem that requires substantial time to correct will have
serious consequences for either approach.
The other extreme obviously
requires greater resources at the ATC, including additional test equipment. At
first sight correcting problems sequentially would appear undesirable from the
schedule viewpoint. However if there are schedule slippages elsewhere and the faults
uncovered during the OMC/WFS
integration require little time to correct, this approach may well be
acceptable. It should minimise the number of problems uncovered later during
system integration and testing but it provides no guarantee that the OMC and
WFS will be completely free of faults. Thus even for this extreme approach
there is a still a risk, albeit very small, that one may have to ship the WFS
or an OMC component back to the ATC or send ATC personnel to Durham.
Table 1 has been divided into
four levels of integration tests. The division of tests between the two
intermediate levels is somewhat arbitrary and one could add levels if desired.
The maximum level stops at open-loop tests. Closed-loop tests would require
much of the RTCS at the ATC and as such it would almost amount to performing
the system integration at the ATC. At the time of writing the author finds it
difficult to make a strong recommendation based on a cursory assessment. The
uncertainty in the delays associated with software development and the transfer
of the WFS to the ATC contribute to the
difficulty. Selection of the either the second intermediate level or the
maximum is suggested.
Table 1. Suggested levels of integration testing for the OMC, NCU
and WFS.
|
Category |
Summary of Integration
Tests |
Comments |
|
Minimum level of
integration tests. |
Determine that all
mechanical interfaces are satisfactory and that there are no mechanical
interferences. Verify all basic OMC & WFS functions using engineering
level software and WFS independent control module. Verify that images are
received by pre-correction camera, IR/optical science ports and WFS but
exclude a quantitative evaluation of image quality. WFS tests limited to only
those possible with pixel stream from the WFS camera, i.e. no real-time image
processing capability. Optical flat used in place
of DM but mechanical interface with DM would be checked. Note: Tests will not assess
sensitivity to temperature changes and vibration, repeatability, reliability,
noise and failure modes. |
Advantages Shortest route to
evaluation of entire system. No additional equipment
and/or software required. Disadvantages Highest risk. Major concern is that
failure to uncover any vibration or stability problems could lead to
significant delay later. |
|
First intermediate level. |
All tests in first category
plus the following: Limited open-loop tests of
FSM (probably using position-sensing detector at f/16.8 focus). Assessment of
vibration effects with accelerometers if sensitivity is sufficient. Simple static wavefronts
generated by the NCU and received by the WFS will be measured and displayed.
Performance variation with light level will be investigated. Limited open-loop test of
WFS dynamic response using tip/tilt injection capability of the NCU. Distortion mapping over the
WFS field using the NCU’s array of point sources. OMC/NCU/WFS sensitivity and
repeatability evaluated when subjected to temperature cycling (within limits
set by laboratory environment) will be assessed. Note: DM would be replaced
by optical flat as for minimum level. |
Advantages Moderate risk. Limited evaluation of
vibration effects. Provides initial but
incomplete assessment of temperature effects, stability and repeatability. Disadvantages Requires additional
equipment and software for display and processing of WFS spots. No information on DM
interaction with subsystems. |
|
Second intermediate level. |
All tests in above
categories plus limited use of the DM as follows: Verification of method of
aligning DM to WFS. Generation of simple
wavefronts with DM and NCU point source. Measurement of these wavefronts with
the WFS. |
Advantage Further reduction of risk. Disadvantage Additional software and
test effort. |
|
Maximum level of
integration testing without use of
RTCS. |
All tests in above
categories plus the following: Determination of
non-common-path aberrations using IR camera at science port. Full completion of
open-loop characterisation tests not already covered above. Assessment of vibration
effects with simultaneous open-loop operation of both the FSM and DM. ( Note:
Approach suggested by RMM requires further study.) Rigorous evaluation of
image quality at all ports. Measurements of noise over
full range of light levels. |
Advantages Lowest risk. Well characterised
subsystems provided for system integration. Provides earliest detection
of most problems associated with the OMC/NCU/WFS as an integrated assembly. Unlikely to require return
of components to ATC for correction of faults. Disadvantages Longest route to evaluation of entire system. Substantial effort required
by ATC with some support from Durham. Sequential correction of
problems may cause significant delays. |