United States Gemini Program

Quarterly Review

of

GNIRS

Held October 13, 1999

at Tucson, Arizona

1. Meeting Background

A USGP quarterly review of GNIRS was held on October 13, 1999. The meeting was attended by Todd Boroson and Mark Trueblood (WPM) from the USGP; David Montgomery of IGPO; and Larry Daggert, Neil Gaughan, Jay Elias, and members of the GNIRS team from NOAO/ETS.

The goal of the quarterly reviews is to evaluate each instrument project's progress in a number of different areas, with emphasis on management and high level concerns. Specifically, the USGP wishes to have a formal mechanism to determine whether a project is on track with respect to budget and schedule, and to identify potential problems before they significantly impact progress.

2. Major Findings

The project appears to be somewhat behind schedule for holding the Pre-Fabrication Review in February 2000 but in the opinion of the WPM and the Project Manager, it is possible to make up that schedule deficit, which is primarily in mechanical analysis of the instrument's flexure and thermal properties.

A major deficiency identified at the Restart Review that had not been corrected at the time of the review is software. No software design or plan was presented at the Quarterly Review, and no software plan appeared to be in place at that time. Since then, NOAO has submitted a very preliminary software plan indicating a more detailed plan will follow next April.

3. Project Summary

3.1 Project Overview and Key Accomplishments

The key accomplishments since the Restart Review have been:

·Concluded negotiations with IGPO concerning the terms and conditions for restarting the project

·Developed a detailed schedule of all tasks needed to complete the design and to fabricate and test the instrument

·Put into place a management information system to status the schedule, obtain financial data in accordance with the WBS, and to produce earned value and other management reports

·Advanced the mechanical design of the instrument in several areas

·Began structural and thermal modeling

·Executed a comprehensive cold motor testing Project to select cold motors and to characterize them for use in the instrument

·Built a large test dewar capable of handling the large mechanisms to be used in the instrument, to test subassemblies before they are installed in the instrument

·Explored various alternatives to obtaining the needed software engineering support

3.2 Project Status and Plans

The GNIRS project is behind schedule in the area of engineering design of the main optical bench, and the structural and thermal analysis of this design. The Project Manager believes the schedule is tight, but that the Pre-Fab review can still be held as scheduled in February. As the Restart Review committee discovered, a detailed software design and management plan are not yet in place, due to lack of NOAO resources. The current schedule is for these to be put into place in April 2000. All other areas appear to be on schedule. All effort is focussed on preparing for the Pre-Fabrication Review scheduled for February 2000.

3.3 Project Problems and Concerns

As noted above, mechanical design and analysis are somewhat behind schedule. It is not clear what the software status is, since no schedule yet exists for software.

3.4 Project Schedule

The summary-level project schedule is attached to this report. An analysis of the schedule performance of each major engineering discipline appears below.

3.5 Project Milestones, Cost, and Manpower Charts

Major milestones for the next few months include the Pre-Fabrication Review and developing the detailed software plan. The cost of the project since the beginning of the instrument redesign in January 1999 is estimated to be $4.2M, not including the $2.4M spent between the project start in October, 1995 and the discovery of serious design flaws in late 1998. Manpower charts are not available at this time, as manpower leveling is still being performed as the schedule is being fine-tuned by the Project Manager. Beginning in the November 1999 report, the USGP expects to see Earned Value reported in the monthly report.

3.6 Project Budget and Expenditures to Date

The Project Manager reports budget and expenditures monthly to the USGP (see attached report for October). This report shows budgeted and actual expenditures for the month and cumulative to date. As of this review, expenses are on track for staying within the budget at completion.

3.7 Organization

The project appears to be staffed with a sufficient number of competent people, except in software. A software manager is due to begin full time in January 2000.

4. Optics Design

The optics design is complete. All further optical work is in the area of procurement, installation, alignment, and integration.

5. Optics Fabrication

All powered elements and the filters are on order, with the other items not yet scheduled to be procured.

6. Mechanical Design

6.1 Mechanical Design Overview and Key Accomplishments

The mechanical design strategy is to use the two mechanical engineers and AutoDesk Mechanical Desktop to make a 3-D model of the instrument, then to use the drafting staff to extract 2-D drawings from the 3-D model, to dimension and tolerance the drawings, and to prepare them for use in fabrication.

As of the Restart Review, the 3-D engineering design was almost at a PDR level, but without any analysis having been done. Between the Restart Review and the Quarterly Review, work concentrated on fleshing out the design and getting it ready to perform the structural and thermal analysis. A major effort was also devoted to designing prototype mechanisms that will be fabricated and cold tested to ensure that repeatability and other requirements can be met.

6.2 Mechanical Design Status and Plans

The prototyping effort is on schedule. The completion of the design, especially the optical bench, and other tasks that need to be completed before structural and thermal analysis can begin are behind schedule, and are now on the Critical Path.

6.3 Mechanical Design Problems and Concerns

WBS elements 5.3.2.1 and 5.3.2.2 are behind schedule. It appears that this is due in large part to a miscommunication between the Project Manager and the engineering staff concerning the ordering and timing of tasks leading up to the structural and thermal analysis of the instrument. The GNIRS must meet some very demanding requirements for both flexure and thermal gradients, so the mechanical engineer performing these analyses is interested in using a model that accurately represents the instrument design. The intended process is that the final stages of design and analysis are parallel efforts each informing the other; when the analysis indicates a stiffer structure is needed in a particular place, the engineer adds the required material to the design, regenerates the model, and repeats the analysis. This cycle is repeated until the design converges on a solution that analysis indicates will meet the requirement, or the engineer determines the requirement cannot be met.

The original schedule had the design of the science and OIWFS optical benches being completed after the Pre-Fabrication Review in February. This activity was moved forward to provide the engineer with an adequately realistic starting point for the analysis. If this work can be completed in time to hold the Pre-Fabrication Review in February, then the mechanical design will actually be slightly ahead of schedule, since a task scheduled for after the review (completing the 3-D design of the two optical benches) will have been completed before the review.

7. Mechanical Fabrication

No mechanical fabrication was scheduled to have begun before the Quarterly Review.

8. Electronics Design

8.1 Electronics Design Overview and Key Accomplishments

The basic electronics design was presented at the Restart Review, and is at a level somewhere between a PDR and a CDR. All assemblies are identified and designed, with remaining work being in the area of some detailed design, such as connector pinouts and wiring lists. The electrical engineer has spent most of his time since the Restart Review working on cold motor qualification and characterization tests, which have been completed.

8.2 Electronics Design Status and Plans

Electronics design is on schedule.

8.3 Electronics Design Problems and Concerns

None

8.4 Electronics Design Schedule

Electronics design is on schedule.

8.5 Electronics Design Milestones, Cost, and Manpower Charts

The next major milestone is completion of mechanism prototype testing. Cost is within budget, and the tasks are adequately staffed with available personnel. After the QR in early November, Andy Rudeen, the GNIRS electronics engineer, gave notice of terminating his employment. The Project Manager does not believe this will have an adverse effect on the project, as electronics is not on the Critical Path, and he believes he will be able to hire a replacement in time to avoid schedule slips.

9. Electronics Fabrication

No deliverable fabrication tasks are scheduled to begin until February 2000. The only fabrication that has been done to date is for motor qualification and characterization, and for mechanism prototype testing. With regard to the latter, a test rig has been assembled and demonstrated.

10. Software Design

10.1 Software Design Overview and Key Accomplishments

Due to major changes in the mechanism control philosophy between the original instrument design and the current design, most of the prior work on software, at least at the detailed level, must be discarded. Some top-level design work concerning the OCS interface may be transferable.

Richard Wolff has agreed to serve as the GNIRS software group manager beginning January 2000. He will be responsible for all GNIRS software efforts. The software requirements are now being documented by the systems engineering staff, primarily Jay Elias. These requirements will be refined by Richard, then he will produce a top-level design, a detailed design, and a management plan by April, 2000. The master project plan will be updated at that time to reflect his estimates for completing the software.

10.2 Software Design Status and Plans

Software design will begin in early 2000.

10.3 Software Design Problems and Concerns

In the experience of the GNIRS Work Package Manager, software usually takes far longer than anticipated. For example, the NIR InSb array controllers NOAO built for Gemini were over a year late, due primarily to delays in software caused by the difficult learning curve for EPICS and underestimating the effort required. With the instrument now scheduled to be delivered about 2.5 years late, NOAO cannot afford to let this happen again on GNIRS.

10.4 Software Design Schedule

The software schedule will be in place in April 2000. Until then, it is impossible to determine if software is on the Critical Path.

10.5 Software Design Milestones, Cost, and Manpower Charts

These will not be known until the detailed software plan is delivered in April 2000.

11. Software Fabrication

No software fabrication was scheduled to begin before the Quarterly Review.

12. Subsystem Integration

No subsystem integration was scheduled to begin before the Quarterly Review.

13. System Integration

No system integration was scheduled to begin before the Quarterly Review.

14. Test and Checkout

No test and checkout tasks were scheduled to begin before the Quarterly Review.

NOAO Intranet Services

NOAO Copyright  Statement