Simon Morris (Durham University, Chair)
Klaus Hodapp (U. of Hawaii)
John Rayner (U. of Hawaii)
Robert Nolan (Gemini)
Mark Hunten (Gemini)
Don Hall (U. of Hawaii)
Corinne Boyer (Gemini) - telecon part time
Doug Simons (Gemini)
The intent of this Critical Design Review is to identify whether the NIFS design as presented in the 3 volumes of CDR documentation, along with viewgraphs presented during the meeting, can be built as described and will meet all the performance requirements, producing interesting scientific results. The committee is also meant to identify any areas where the information presented is not enough to assess whether the design will work. It is not the job of this committee to suggest alternatives or minor design improvements, but to assess the instrument as presented for major flaws or risks. An alternative way of stating the above is that the committee is meant to decide whether it agrees with the risk analysis table the NIFS team presented in the systems analysis section, and whether we agree that the mitigation plans the NIFS team presented are sufficient to pass the CDR milestone.
The committee thinks that the NIFS team has once again presented an extremely well documented and well-prepared review. In all areas, we think the plans and documentation are at, or well above, CDR level, and so recommend proceed to completion as rapidly as possible. The committee was very impressed by the high quality of the work done and the planning presented.
If there had been issues that the committee had felt were enough to cause NIFS to fail the CDR, or require some sort of additional formal review before proceeding to full construction, we would have listed them here.
Listed here are areas where the committee thinks there are still uncertainties or risks that could significantly impact either NIFS performance or cost and schedule. The NIFS team in the CDR documentation already identified most of these. In all cases we think that the NIFS team's proposals for how to handle these uncertainties or risks (with a few minor modification discussed below) are sufficient to allow them to proceed to construction.
Performance modeling. The committee agrees that the comparisons presented in the CDR documentation are enough to give one reasonable confidence in the predicted NIFS performance. Residual mild areas of concern are the loss of chip area due to MUX glow, and possibly other effects, leading to a smaller effective slit length and wavelength coverage. The unknown effect of persistence in the detector could also end up being a driver on performance unless heroic operational efforts are able to mitigate this possible problem. We encourage the NIFS team to continue to explore operational strategies to minimize this problem. The upgrade path towards an MBE detector may ultimately be the best solution for this. We encourage the instrument team to continue their efforts in preserving this upgrade path.
Guide Star Availability. The comprehensive analysis shown in the CDR documentation demonstrated clearly to the committee that (a) one will have to plan experiments with NIFS carefully, but also (b) that the definitely is enough interesting science to be done with NIFS to justify its construction. As with all instruments though the faster it is delivered the more exciting science it will be able to cream off before its competitors.
We also encourage the NIFS team to continue to track the NIRI OIWFS performance tests and consider the effects of the AO corrected PSF feeding the OIWFS, and the possibilities of wider filters such as J+H. After the objects with suitable OIWFS guide stars are all observed, we note that there is a further option of then removing the OIWFS prism to go a little fainter at the cost of reduced AO performance.
NIFS RESPONSE – We will continue to
pursue an aggressive construction schedule leading to an early commissioning
date. A J+H filter will be provided in the OIWFS filter wheel.
Ghost
images
We encourage the NIFS team to investigate post-processing algorithms for any (possible) residual structure from OH spectra ghosts.
NIFS RESPONSE – Such schemes are likely to be cumbersome. We will explore this approach if it is demonstrated with real data that it is necessary.
Image
slicer, pupil and field array manufacture.
The committee was impressed with the figure achieved on the test pieces, but share the NIFS team concerns about the surface finish. We encourage the NIFS team to contact Martin Wells at the ATC and Dave Robertson at Durham for information about diamond machinable materials.
NIFS RESPONSE – They will be contacted and any suggestions passed on to the University of Bremen. Further trials will be conducted using the alloy preferred by the University of Bremen and using nickel-plated blanks. A last resort will be to ask POE to diamond turn the arrays in a manually advanced jig.
Lens
Mounting
The committee agrees with the careful and conservative analysis for the lens mountings, and thinks that the proposed approach of machining a larger hole will work. Given the very high cost of a failure though, we also strongly encourage the team to proceed with a test using a blank.
NIFS RESPONSE – A test will be performed
with a test lens.
OIWFS
gimbal flexure
We recommend that the NIFS team continues to monitor the NIRI OIWFS tests, but should proceed assuming they will be successful. In the event of failure, then solutions will have to be investigated in collaboration with the NIRI team and Gemini.
NIFS RESPONSE – Gimbal mirror flexure
will be ignored unless data to the contrary are supplied by Gemini.
Grating
turret flexure
We endorse the NIFS team proposal to put some effort into improving the flexure performance of this item, but do not want this to impact the schedule.
NIFS RESPONSE – Ways of reducing the flexure component of the grating mount will be explored during the design detailing.
Detector
stability
The committee agrees that the mK control and the reference signal in combination mean that detector stability is no longer a serious concern. The committee suspects that the temperature sensor on the detector may not be a good idea, because it will not contribute to the instrument performance and introduces some risk.
NIFS RESPONSE – Gemini will be asked to contact Rockwell and not proceed with mounting a temperature sensor on the chip carrier. In no event should Rockwell attempt to mount a temperature sensor on the detector itself.
Array
Controllers
The committee believes it is crucial for the current schedule that the NIFS team obtains 2 array controllers to allow them to work with the test dewar and the NIFS dewar simultaneously.
NIFS RESPONSE – Gemini will purchase the necessary spares for the controller and have agreed to make these available to RSAA to set up a second controller based on an existing SDSU-2 development system.
Software
NIFS detector control - the committee (Boyer) think that sharing the CPU between the data task and the readout task is very risky. We feel this should be investigated further. It is possible that a second CPU (MVM2700 board) may be needed to handle the data task and the DHS transfer. This will allow two network connections (one with control LAN for the epics CPU, the second one with the data LAN for the DHS CPU). Also this will simplify the design for the DC and the implementation of this design.
The NIRI implementation of the apply record within the instrument sequencer may be flawed, and solutions to this should be investigated in collaboration with Gemini and the NIRI team. The contact person for this is Mathieu Bec.
The NIFS team needs to provide an ICD for the OIWFS or agree that the NIRI ICD is adequate.
We recommend that the NIFS team continue their dialog with Gemini regarding the data reduction and Quick Look software to ensure that it is compatible with the data pipeline.
NIFS RESPONSE – Tests are being
performed to determine the extent of any problem sharing the CPU between the
data task and the readout task. The impact will be assessed when specific
information is available.
Contact has been made with Mathieu Bec.
A problem with sequencing the NIRI DC from the IS has been identified. This
will be resolved in the NIFS implementation of different DC code.
Our dialog with Gemini will continue after the recently-released Observing Tool has been installed at RSAA.
Flexure
Testing
The committee would like to see a fully developed flexure test plan for 2 axes. We agree that it would be sufficient to simply flexure test the cryostat.
NIFS RESPONSE – A full flexure test plan
for 2 axes will be developed and submitted to Gemini.
Management
plan, cost and schedule
The committee thinks that the NIFS management plan, cost estimate and schedule are extremely well thought out, and stand one of the highest chances of being achieved of the Gemini instruments. As identified by the NIFS team, personnel and detector delivery remain the high posts among the risks.
NIFS RESPONSE – Agreed. Efforts continue to fill vacant staff positions.
Other
We encourage the NIFS team to work with Gemini to reach compliance with the ICD on space envelope and strip the paint on the surface contacting the ISS.
NIFS RESPONSE – Agreed. This work is in progress.
There were a number of comments made, or discussions had, during the CDR that do not qualify as concerns but which might be of value. I have a fairly complete set of notes from the meeting that will be distributed to both the NIFS team (via Peter McGregor) and the IGPO (via Doug Simons) that contain these comments or discussions.