Weekly Report for Week Ending November 16, 2000

The LIGO Executive Committee Agenda for Monday  November 20, 2000 will be:
 (Meeting time: 10:30 am Pacific Time)

Open meeting 10:30 - 11:30

1. Announcements
2. LSC Issues (Weiss)
3. Comments on Weekly Report
4. WBS 1 LIGO I Construction (Lindquist)
• Field Change Orders/Contingency Liens/Change Requests
5. WBS 2 LIGO Lab Operations
• Administration (Lindquist)
• Sites (Raab, Coles, Shoemaker, Sanders)
• Detector (Whitcomb, Coyne)
• Campus Research Facilities (Weinstein (40 Meter), Libbrecht (TNI), Zucker(LASTI))
• Data Analysis and Computing (Lazzarini)
6. WBS 3 and 4  Advanced R&D and LIGO II (Sanders)

Special Items: Status of Budgets and Proposal for FY 2002-2006

Special Announcements:

Weekly Report Highlights
 

LSC Issues (Weiss)

no report

LIGO I Construction/LIGO Laboratory Administration (Lindquist)

WBS 1.2 LIGO Operations--Administration

LIGO Weekly Site Telecon (Lindquist)

The list of current actions revised to reflect open actions assigned through November 9, 2000 may be found at ACTION LIST.

From: Ed Chargois <chargois_e@ligo.caltech.edu>

• Assisted the Optic Engineer (G. Billingsley) with packing and shipping of a Suprasil 311 (glass) to the Ginzton Laboratory (A. Alexandrovski), Stanford University , scheduled arrival date 11-17-00.  Account Number P1157746
• Received confirmation from Gerry Lane, Chevrolet (Baton Rouge LA), that all repairs are completed on the GSA Chevrolet Corsica, and it is ready for pick-up.  LLO (B. Wascom) was notified on 11-14-00.
• Received request on 11-14-00 from the General Services Administration Fleet Management Office (J. Williams) that the repaired Chevrolet Corsica be returned to GSA for a New Vehicle, Make and Model is unknown.  This information was relayed to the LLO (B. Wascom).
• Assisted the Detector Group (L. Jones) with the packing and shipping of two (2) Bellows to Allied Engineering & Production (M. Kahn), Alameda CA, via Federal Express, expected delivery date 11-16-00. Account number P158082.

>From: Linda Turner - turner@ligo.caltech.edu>

Web pages for the DCC give simple how-to's for document numbering, easy access to the latest on-line documents, and search capabilities for the DCC database. Take a look. . .

ACCOMPLISHMENTS

• A milestone was reached this week with the completion of the auditing and sorting of construction drawings for Livingston.  A similar process has begun with the Hanford construction drawings.
• Christmas party planning continued in securing various restaurant/caterers for main entrees and completing the floor plan for Physical Plant's setup of Dabney.  Budget estimates were collected from Smart & Final and Costco for paper goods, wine, and our party supplies.
• Worked through a tremendouse number of electronic documents.  Over 120 electronic documents were processed and transferred into the dcc directory and database.  This was one of our most productive weeks to date.
• Prepared several sheets of labels for past revisions of Livingston drawings needing to be filed and for drawings sent out as addendum 4 to the staging buidling IFB EJ-319.
• Met with Barbara and Larry to set up a schedule for transitioning the DCC to Office 2000.  The new server will be loaded and ready for Barbara new database version by December 8th.  From December 11-13, I will enter data on a test basis to ensure that everything is working without any glitches. On December 14 and 15, the DCC database will be unavailable for number assignments or data entry so that Barbara can transition the workstations into production capability.  LIGO staff needing DCC numbers should plan ahead for the two-day shutdown.  A memo will go out alerting all staff prior to the actual shutdown.
• Work continued on the large pile of process travelers.

• In discussion with a staff member this week, I discovered that while drawings were being electronically filed in .pdf format with the DCC, the .dwg format was being accessed from his personal directory by those needing a specific drawing.  This raised several concerns:
• Revision control,
• The need for parallel electronic formats for any given document, and
• The integrity of the DCC for providing the latest rendition of a document
• Addressing the above concern, changes were made to web pages, data entry screens, etc. to facilitate multiple formats of the same document.  Although other formats of a document may be submitted, the .pdf version is still required for an electronic submittal.  This flexibility would allow a .dwg format to be contained in the directory along with the .pdf.   Hence, it is recommended that access for a drawing be made through the DCC directory rather than from someone's personal folder.
• Secondly,  an review of internal bulletin boards will begin to see that all linked documents are:
• in the DCC, and/or
• links are to the dcc directory for a document rather than to a personal folder.

Progress on auditing of drawings continues.  Livingston Facility/Construction drawings are just about done with a few loose ends to tie up.  Auditing of Hanford drawings have begun.  With the addition of numerous full-size drawings being filed away, there was some need for considerable reshuffling of files to accommodate for these drawings.

Prepared, packaged, and shipped Addendum #4 to IFB EJ-319.

Processed large number of electronic documents as well.

Press here to access the DOCUMENT CONTROL CENTER WEB PAGE.

From: Esther Cunningham <esther@ligo.caltech.edu>

Press here for ACCOUNTS PAYABLE HISTORY DATA.

From: "Brambila, Ruth" <Ruth.Brambila@caltech.edu>

The Washington State Auditor returned yesterday and today to continue the audit of LHO.

Received a request today for the DCAA auditor to provide more detailed information on the "Level of Effort" contracts for LIGO. I will be obtaining that information today to respond.

Working on internal modifications as requested, Hereaus-Amersil, Corning, setting up the file for Manpower, and completing University of Michigan.

Working on resolving the one time PO to VWR and Thor. Continue to place and reconcile pcard transactions.

Requested special permission from Purchasing Services to have all LIGO's POs processed as "three-way" rather than "two-way" to ascertain that on-line receiving will be completed. Should be getting a response soon.

The class action lawsuit against Airgas was forwarded by the main Purchasing Office to Sandy Pool. Awaiting to hear a response.

Have been scheduled for two 2-hr Webster/Oracle Discover training sessions later this month.

Electronic e-mail notifications are now being generated as notification to complete and approve POs after a certain period of time. In addition, notifications are now being sent on blanket POs when funds are more than 85% expended.

• Met with property accounting department received and reviewed the "NSF Property in the Custody of Awardees" report as of September 30, 2000.  A complete package consisting of supporting documents indicating how the various amounts on the report were derived, and a statement of the methodology and assumptions made in preparing the report will be submitted to the Document Control Center.
• Completed and sent out Integrated Report for MIT for October.
• Completed and sent out Integrated Report for Data Analysis Group for October.
• Submitted a list of FY2001 account numbers to Sponsored Research Department for approval.
• Prepared a draft of "Classification Guidelines for Equipment" which attempts to clarify the distinction between equipment and supplies and the distinction between "Government-Owned", "Caltech-Owned" and "Other Owned", and the proper uses of Equipment Fabrication accounts.
• Notified Ruth Brambila of encumbrance balances that need to be closed.
• Provided Alan Weinstein with additional information for expenditures Campus Research.
• Provided Federal Accounting with information for required realignment of budget for R&D accounts.

Prepared and issued report for October for two Visitor accounts.

From: irena@ligo.caltech.edu (Irena Petrac)

• CSIRO: Issued CO No. 08 for teflon polishing of Caltech/LIGO provided sapphire test pieces.
• TAMA-SAS prototype/Galli&Morelli: Processing of the Purchase Order for TAMA-SAS hardware from Galli&Morelli is pending receipt of order from National Astronomical Observatory (NAO), Japan for same.
• Univ. of Florida:  Payment of invoices for August and September is pending receipt of additional costs incurred detail (from the Univ. of Florida) to allow allocation to appropriate LIGO accounts.
• CO No. 10, adding the second installment of funds (October 1, 00 through September 30, 01) for Sany Yoshida and Heisheng Rong, is pending receipt of an updated budget from the Univ. of Florida.

From: Ed Jasnow <jasnow@ligo.caltech.edu>

• No report this week (See section addressing Construction below.)

Advanced LIGO (Frey, Petrac)

WBS 1.4.1.2   Project Controls (LIGO Construction)

Annual Report:  There will be an Annual Report due as of the end of November.  I will be requesting inputs from the usual bunch mid-December after the dust from the proposal has a chance to settle a bit.  I propose to keep this report moderate in size, about the size and complexity of a Quarterly Report, to minimize the burden.

The following Change Requests have been submitted:
 

Press for the latest Contingency Needs Projection.

From: Kris Duncan <kris@ligo.caltech.edu>

• Completed first draft of the October 2000 CSSR. I will distribute and hope to receive comments and review by Wednesday November 22.
• Continued with construction contract/work package completion close out.
• Initiated work on the RAM (Responsibility Assignment Matrix) for month ending October 2000.

• I updated the details for encumbrances and actuals for the Construction and Advanced R&D accounts for October 2000 and emailed them to the Task managers.   When looking over the details, I noticed several errors and I have initiated corrections.
• I have just received the MIT TWONK (COST ALLOCATION) report for the month of October 2000.
• I continue to reviewed the open encumbrances on the Construction and Advanced R&D accounts to see if the requested removals have been made. I will continue to review these accounts weekly until all encumbrances have been removed (Remember, I can only get these errors corrected if the Task Managers point them out to me).
• The financial reports on the web provide supporting detail.
• http://docuserv.ligo.caltech.edu/~fireport
• http://docuserv.ligo.caltech.edu/~finance
• I have given an updated list of all OPEN LIGO account numbers to Barbara Kratochwill and she has put them on the LIGO internal bulletin board.  Please use these lists when you need a LIGO account number (or make yourself a printed copy for quick references).  Also, please note that the FY00 LIGO Operation account numbers expire on November 30, 2000.  A new list of LIGO Operation account numbers will be issued as soon as I get them.
http://docuserv.ligo.caltech.edu/

From: Ed Jasnow <jasnow@ligo.caltech.edu>

• Revised bid packages for the Livingston Staging Building were sent out to the four original bidders on Tuesday, November 14. A pre-bid conference has been scheduled for Friday, November 17, to discuss all of the revisions. The bid opening is scheduled for Thursday, November 30.
• Six architect/engineering firms have responded to our request for interest in designing the new Staging Building at Hanford.  These will be reviewed by a committee to narrow this number down to the two or three most qualified firms. They will then be invited to make presentations to the committee.

• No report this week; see Operations Section above.

Quality/Safety (Tyler)

No report.

LIGO Hanford Observatory (LHO) Operations (Raab).

General Items (F. Raab)

E2 has been completed as scheduled and tasks that were on hold to allow unmolested running are getting attention. A great job was done by all. The overall run looked successful, but will be reported more thoroughly elsewhere. My major observation is that an idea we adopted early on in planning the facilities - that users would be able to conveniently access data from the computer users room and elsewhere on site - is not yet reality. Consequently, it is a hardship to get at diagnostics without being at a control-room computer, so it is difficult to control control-room traffic when we have many visitors. This is not urgently an issue, but it will need attention before runs become "routine."

LIGO Livingston Observatory (LLO) Operations (Coles).

OPTICS/COC/SEI INSTALLATION: Realignment of ETM-y is in progress. Arm cavity baffle has been installed, and will be aligned today. We measured a 1 cm error in the +Y position of the ETM (which also is reflected in the PSI monuments on the slab) referencing BT-VE5. We'll move the LOS to correct this before final alignment. We plan on completing the in-vacuum work and pumping down early next week. (Jonathan Kern)

Computing: Shannon is working on the control room checklist, and is adding functionality to it. (Data entry is working, but data retrieval
forms are still being written.)We have ordered extra SCSI drivers and a disk to allow us to move the web and mail services to separate machines, and to provide some redundancy for them. We are expecting an evaluation copy of "Quick Restore" to help automate backups on our new tape robot.We are configuring a workstation to be a standalone Solaris 7 machine for running the E2E software, which seems to be having problems with Solaris 8.

ISC: Sany and I continued ISC prep. by fine aligning Faraday port beam into ISCT1; the beam now looks fairly gaussian in both x and y axis: appears to be no clipping now. ITMX, ITMY, and BS seem to have suspension control issues; we are troubleshooting these. (Anthony Rizzi)

Detector/Technical Support (Whitcomb, Coyne).

See also the daily electronic logs for the installation and commissioning activities:
Hanford Detector Log
Livingston Detector Log

1.1 LHO INSTALLATION & COMMISSIONING

Daniel Sigg, John Zweizig, Stan Whitcomb and a cast of thousands....

The system performed pretty well.  Typical lock durations were  30-60 minuted in the recombined configuration,  2-6 hours in the single arm configuration.  Reacquisition of lock typically took less than 1 minute. There were no major hardware failures.  The data acquisition system complained about the heavy load, but hung in for the duration.  We collected 2 terrabytes bytes of data, performed some on-line analysis, and educated a number of LSC scientists.

E2 run (Daniel Sigg)

The second engineering run was a great success. Data taking started on Wed., 11/8, at 12:00am and ended a week later on 11/15, 12:00am.  Without making an exact analysis the estimated up-time was at least 80%---probably around 90%. The first 5.5 days were run in the recombined Michelson configuration with typical lock stretches between 30 and 60 minutes; with the longest one approximately 3 hours. The remaining 1.5 days were dedicated to single arm operations, equally split the y and the x arm configuration.

No major hardware or software problem was encountered during the run: the SUS, LSC and ASC systems worked reliably over days. The PSL and IOO showed excellent performance---not a single lock was lost due to laser or mode cleaner. The data acquisition system was handling the load and all data was saved to disk and broadcasted to the data monitoring tools. A couple of glitches which have to fixed in the next few weeks showed up: reliability was not 100% and the system (or part of it) had to be rebooted occasionally, the number of EPICS channels could not be scaled up to the full set, there were occasional timing glitches and the lack of anti-aliasing

The organization of the shifts and check list tasks worked out fine. Even so, we had do a lot of self-education at the beginning, the excellent motivation of people on shift made it work. Clearly, a couple of  improvements need to be implemented for the next go around: in particular the e-log and the check lists need more attention and fine tuning,

The scientific shift group also undertook a set of investigation tasks looking at the cause of lock, various stationary and non-stationary noise sources, transient, angular fluctuations, and data integrity. We also performed extensive calibrations of both amplitude, phase and absolute timing and frequency noise propagation. First preliminary results are already available for cause of lock (length-angle coupling twisting the optics), tidal trends (both differential and common mode over a total of 130 hours), calibration of the sensitivity (a few times 10^-18 m/m/rtHz at 1kHz) and the stability of  lines in the spectra (including calibration lines).

More as it becomes available on the E2 web site

Luca Matone, Micheal Landry

Worked on the Calibration task for E2. We have made a few measurements, in particular we have sent sinusoid sweeps to the masses to get transfer functions of the response of the system and we have injected calibration lines to the ETMs to obtain their calibrations when only one arm was locked).

Input Optics (Dave Reitze )

Preparation of the small optics continues.  A mishap with one of the MC flat mirrors occurred; it looks like it will have to be repolished and recoated.  Fortunately, spares were available and have been sent to LHO for gluing. In addition, sufficient magnet standoffs and wire standoffs for 10 additional mirror glue cycles have been manufactured and will be

1.2 LLO INSTALLATION & COMMISSIONING

Commissioning (Peter Saulson )

We have done careful measurements of the mode cleaner's loop gain, confirming that a unity gain frequency of about 20 kHz can be achieved with good behavior, but that turning the gain up leads to oscillation around 70 kHz. A new calibration was also made of the MC length error signal. In addition, the MC length/frequency crossover was mapped as a function of MCA gain.

Warren Johnson continues to help us learn about the anomalous damping of  the small optics. In addition to low Q's of many pitch modes, there are other modes that show Q that gets dramatically worse at the low amplitudes characteristic of operating conditions, even though they show reasonable Q when driven to large amplitudes.

Nergis is updating our CDS software. Anthony and Sany are making good progress on the alignment of ISCT1. We also successfully completed the renaming of the IOO controls channels. New IOO DAQ channels have been hooked up and Rolf has added them to the master configuration.

We reinstalled some IOO WFS software and got the mode cleaner WFS working again. Since installation of the RF distribution system, the RF phases have changed, so the optimal WFS matrix will have to be remeasured, but the loops work even with the old matrix.

Input Optics (Dave Reitze)

• Work of MC commissioning continues: loop gains have been measured by Nergis and David S., length error signal has been calibrated (1.21 KHz/V at DC; see 11/13 LLO e-log for more details).  Length/frequency crossover has been measured; results as a function of MCA gain with and without the elliptic filter have been recorded (see 11/13 e-log)
• Sany and Warren continued Q measurements. Thus far, they have made measurements for MC1, MC2, MMT1, MMT2, SM and RM. Detailed analyses are ongoing.
• Continued the preparation of ISC1 table. Now the beam reflected from RM and rejected by Faraday isolator hits the center of the top periscope miror on ISC1 Table. The power coming out of FI is ~20% lower than the expected value (calculated from SM transmission and measured MC transmitted power); although there is some uncertainty in this number. Checked the profile of the beam coming out of FI with a beam scanner. It was not a perfect Gaussian profile, but did not show that the beam was clipping. (Because RM's damping controller is not working properly for some reason, the beam coming out of FI is swinging slightly. This causes the beam profile unstable.)

2.0 Other Engineering and Scientific Activities

2.1 Design/Analysis/Fab

Optical Metrology

FM04 and  SPETM04 (uncoated) were sent to S. Rowan at Stanford to investigate the effect of coating on Q.

PSL
Note: Gary Sanders, Peter King, Rich Abbott, Rick Savage and Joe Kovalik are visiting the Univ. of Hanover to discuss collaboration on advanced LIGO laser development.

Rick Karwski

We continue to test the pre-mode cleaner circuit boards.
The test fixture for the Photo-diode boards has been built, tested and is in use.
We have completed an on-the-bench functional test of one of the photo-diode boards.
We have discovered circuit discrepancies on the PD and we are in the process of resolving issues with the LHO crew.
We have begun testing the pre-mode cleaner servo card.

Lee Cardenas

• I have Finished the Resonance frequency measurements of most of the 1.00" mirror mounts used in the PSL Lab and the external IO.  First resonances:
Solid Aluminum block mount from LLO: 1248 Hz (no fine angular adjustment capability)
Lees make, mounted on top of a New Focus post model 9954 (1.0" Dia x 2.0" Long S.S post): 956 Hz
New focus  model 9807: 854 Hz
NRC model U100: 764 Hz.
The 2.0"  steering mirror mount from COS measures 432 HZ.
The rest are bellow 500Hz. Graphs are available for all these.
• The design and drawing for the new base of the old version PMC is completed and Fabrication is on going.
• I have received a LIGO 10 W laser (MOPA S/N 110) which was repaired and have installed it in the PSL Lab.  I have sent MOPA S/N 107 to LLO along with a new Reference Cavity for the 4K Interferometer.  Performance data check  for the MOPA S/N 110 is on going.
• I have thoroughly cleaned  the laser chiller and the hoses and  put new DI water mixed with Ethylene-Glycol (20% & 80% DI) along with a Chloramine-T (1gram/gal) to restrict the growth of Algae as recommended by Neslab.

Peter Fritschel, Myron McInnis

• Myron and Peter have begun production of components for the new osem assemblies; current production is a batch of about 80 heads for the LHO 4k interferometer. We have bonded the spring clips (which hold down the 1.06 micron block filters) to all the photodetector boards we have (need a few more long boards to complete). We also began producing the rear connector pin plates.
• A test run of a zirconium nitride plating on alumina heads should be completed by RPC late this week.

Visited a potential OSEM head coater (Macro Metalics) and examined their TiCr eletrostatic discharge coating. It looks promising, but another set of samples need to be fired to get the correct resistance. Also met a representative of one of the alumina head manufacturers (Progressive Technologies): they reported a delay of a few days in completion of the production lot of 300 heads.

2.2 Issues Concerns

40 Meter Interferometer (Weinstein).

STACIS active seismic isolators:

• Larry Jones and Steve Vass have grouted one of the (3 times 4) plates on which the isolators will sit. The grout expanded, bringing the plats out of flatness and levelness.  The search is on for a more appropriate grout.
• The grout plates and support beams for the test mass seismic stacks were surveyed by Larry with an autolevel that has since been discovered to be out of calibration.  Larry and Steve have re-calibrated it.  They don't believe that things have moved so much as to make this a real problem.
• All the support dradles have been manufactured and delivered.
• We still await the remaining 3 of 4 STACIS systems. Expected any day now!
• Expect installation of STACIS to be complete by mid-December.

• Allied is cleaning and baking the parts it has.  Replacement spares for the vacuum bellows damaged in shipment to Allied have been sent up there,  Larry Jones is soliciting bids on adjustable bellows for the MC beam tube.  CES is making progress in manufacturing MC chamber parts. The table on which the MC chamber sits is being painted and plated.
• 40m building rehab: Fred Asiri will put pressure on the contractor to finish up work, and go through a walk-through of the job.  Still some cable tray work, wall patching and painting, electrical runs, floor tiling, etc., to be completed.
• Caltech says it will re-roof the 40m building (including North and South Annexes), with due attention to joints and penetrations, during December (assuming it doesn't rain!). Should take a couple of weeks.
• SO, by mid-December, we should be finished with STACIS installation, contractor work, and re-roofing. At that point, we should move back into clean-room mode. We can then install the PSL enclosure, and begin to assemble the PSL, with the goal of having a working light source by March. Installation of the output optic chamber and 12m mode cleaner will happen in the spring.
• Vacuum control system: Ben Abbott has all the cables for the  remote vacuum valve controls. He'll start wiring on Friday.  Should have all the valve control boxes wired and deployed by the end of next week.
• Ben Abbott has made drawings of the computer networking for the 40m, including the computers for DAQS, and for operator EPICS, GDS, DMT screens.  He has another drawing laying out our operator control room  (a mini version of the site control rooms).  He's looking into furniture (desks, tables, chairs, carts) and video monitors.
• Rick is testing PMC control boards for the PSL for H4K, 40m, and LASTI.
• Garilynn has prepared the PO's for the Corning and Heraeus glass blanks for the 40m. They should go out soon.  She needs wedge angles, and tolerances for the ROC's.  Mike Smith will work on the wedge angles when he returns next month, and AJW will establish the ROC tolerances after talking with Bill Kells and Dennis Coyne.
• Janeen met with LIGO SOS experts on updating the design of the 40m SOS suspension parts, fixtures, and balancing specs.  She's preparing drawings for manufacture by CES.  And, she's still ordering parts and components for 40m OSEMs and pigtails.
• Tom Frey reports that the 40m contruction is scheduled to be complete  in Jan 2003. He's working with AJW on a schedule for the experiment.
• AJW has spent much of the last week trying to understand the origin of  many of the big charges made against the 40m accounts (much help from Florence Kaufman, Steve Vass, and others).  PLEASE, is you charge anything to the 40m accounts, even Pcard charges, inform AJW, and get copies of the PO's  to AJW and Steve Vass!
• Dennis Ugolini is at Hanford for E2.

Thermal Noise Interferometer (Libbrecht).

This week we finished and installed cables and electronics for one of the test cavities.  Active damping appears to work nicely on both mirrors.

LASTI (Zucker).

LASTI (MacInnis, Mason, Smith, Kruzel, Shoemaker, Zucker)

SEI installation: Still waiting on some cleanroom parts for HAM21, so we are drilling for the BSC pier anchor bolts and roughing up concrete before finishing HAM21 SEI installation.

Cleanrooms: KM and MZ are working on a BSC custom cleanroom solution compatible with Larry Jones' Cartridge SEI/SUS installation scenario; headroom is tight but with proper attention to low headroom in designing lift fixtures etc. it should be feasible.  We are now looking at having the roof of the main cleanroom double as a work platform for installation by making its section deeper, adding reinforcements for man-rating, and mounting the fan units on edge as side-blowers. There isn't enough clearance for a standard bolt-on  work platform.

Vacuum system: ELK is completing wiring and chilled water installation in preparation for test of the new roughing pump today or tomorrow.

LASTI (Janeen Romie)
 
 

Data Analysis and Computing (Lazzarini)

Simulation and Modeling (Yamamoto)

LHO E2 run

• Hiro worked from Nov. 6 to Nov. 12, and Biplab worked from Nov. 11 to 15th, and they attended two owl shifts each.
• Biplab studied correlations of angular fluctuations   of mirrors in both locked and unlocked states with environmental variables like the building tilts. Frequency ranges of strong correlations with seismic noise and control loops have been identified.
• Hiro studied the status just before the lock breaks, both the asy DC power and pitch and yaw of cavity mirrors. For this study, 3D mirror implementation is given the highest priority.

• Biplab cleaned and tested some parts of E2E code that handle misalignment calculations in summation cavities.
• Hiro studied the power spectrum whitening and dewhitening using e2e digital filters.
• Curvature mismatch calculation is almost done, but there still is a point to be validated.
• Malik of UFL is going to start working on the modeling of the thermal effect and the dual recycling configuration.

• Tavio and Matt worked designing the data structures for thread synchronization within the the first version of concurrent multithreading e2e simulation have been designed and sanity-checked.  These are aboutto be verified using a test bed designed to closely resemble their use within modeler.
• As soon as the synchronization mechanism is verified it will be surgically   implanted within modeler, and then modeler's performance on Sargas will be tested.
• Assuming all goes well with the above two items, the next two tasks should be:

1. Tuning the queueing structures to allow a mixture of dynamic and precomputed queueing over the evolution topology.  Groups of contiguous & interdependent, less time-consuming primitives will be precomputed: written into several small tick queues as before, so that those areas where the overhead of dynamic queueing could slow down the simulation are reduced while still allowing concurrent threads to operate.
2. Separating out the intermittent disk writes of data_out modules within a seperate thread of execution running at lower priority, using buffering in such a way as to ensure the disk-write operations do not slow down the time evolution iterations.

Ed and Bruce fixed more bugs in the bug list. alfi is quite stable now. alfi is also capable of opening box files which were editted without using alfi. The code change strategy is to make the whole system more robust in the future modification. The program report is reorganized so that it is easier to make a strategy to completely wipe out the bugs in the list.

LIGO Data Analysis System

Software Systems (Blackburn)

During the second engineering run at Hanford, LDAS was able to ingest over 700,000 trigger events into its database from the DMT. This averages out to a 1 Hz trigger rate. However, there were epochs during which the DMT was generating and we were logging triggers at 100 Hz(!) and LDAS was able to handle this extreme load.

A couple of days prior to the run it was determined that bad frames were being written with duplicate channel names. This caused a segmentation fault in the frameCPP which caused the frameAPI to go into a cyclic restart-fail loop. The frames were fixed and a patch to the start-up macro for the frameAPI was installed. This fixed all the problems seen with the frames and the frameAPI.

The writing of the frame data to tapes proved to be much more problematic, although no data was lost and we expect the final set of tapes to arrive here at CIT tomorrow. There were tremendous problems with the tape robot and the drivers used to communicate with it. The manufacturer of the unit was contacted several times and in one instance spent several hours on the phone with us trying to sort out the problems. For reasons still unknown to us, the writting of data to tape would occasionally result in scsi resets and other problems such as lose of communications with the unit. The server for the tape robot had to be rebooted on several occasions to test out a new configuration or different hardware component. During these reboots the LDAS software had to be shutdown and later restarted. During these down periods, DMT triggers were not ingested into the database. The lesson that was learn from this was to separate the server for the tape robot from the rest of LDAS so that LDAS could continue to run when tape problems occurred.

There have been a couple of issues associated with the migration of the wrapperAPI over to LAM. These are still being worked out. However, we did learn that "collective" calls can be very dangerous to use and have hence replaced all these with point-to-point communications.

Work on the mpiAPI has begun again using the LAM implementation of MPI as the standard. There is still a significant amount of code development needed to integrate the mpiAPI with the wrapperAPI.

Several bugs in the dataConditionAPI call chain were identified and fixed this week. The focus of work on the dataConditionAPI is now on formalizing the implementation of the new unified data type (UDT). A substantial portion of the dataConditionAPI has been reworked to use this new UDT. This new code will be appended to the cvs repository later this week.

A new user interface was developed to exercise user commands which condition frame data using the dataConditionAPI. This interface is graphical and is called "chex" for CHannel EXerciser. It is in the LDAS repository and has been made available to the development team for the dataConditionAPI for evaluation and feedback. In addition chex was used during the engineering run to perform statistical analysis of various frame channels in 60 second segments.

The new controlMonitorAPI which appeard in the 0.0.12 release of LDAS used during the engineering run was successfully used to monitor the state of LDAS and the tape controller software. It also was used to easily start and stop the LDAS system when hardware maintenance was performed.

All problems identified with the frameCPP i/o library during the engineering run have been fixed and added to the repository, as well as to the framebuilder.

In closing I would like to say that the performance of the LDAS software during the second engineering run was outstanding. The software was very stable and able to endure stresses outside of those used in routine validation test. The software team did an excellent job of preparation getting ready for this run.

Hardware Systems (Anderson)

The memory problems with the new 16-node LDAS beowulf computer have been fully characterized and the system is close to being released for software development.

Problems with the AIT-2 tape drives during the 2nd LHO Engineering run have been diagnosed and a candidate fix will be tested now that the run is over.

The first price quotes for a 6000 slot LDAS tape robot has been obtained.

An investigation into an alternative tape storage system to HPSS is commencing.

----- Forwarded message from Greg Mendell -----

1) LDAS successfully archived all of the raw frame data from the LHO E2 run onto tape. Seventeen tapes have already arrived at Caltech.  Another eight tapes are in transit and three more are waiting to be shipped today. All tapes should be at Caltech by Friday Nov. 17, 2000.  A backup copy of each of these tapes is at LHO.

2) SCSI reset problems with the tape drives were documented in the elogs at LHO.  These types of resets did not occur during testing on the ldas system prior to the run.  The resets were overcome by restarting the tape control script as needed, and by enhancing the script to handle resets. They did not cause the loss of any files from the entire E2 run. Reconfiguration of the drivers for the tape drives should fix the problem in the future.  Test of this are now underway.  Also, it should be noted that the actual enterprise level tape library system to be used at LHO has not yet been purchased.  The system used for the E2 run was based on the hardware currently on hand.

3) The tapes contain tar balls with 200 one-second frame files per tar ball. The frame size during the E2 run was 3.183092 MB, thus each tar ball is just over 636.6 MB.  Each tar ball will have a name beginning with the first frame file in the tar ball, followed by .n, followed by the number of frame files in the tar ball, followed by .tar, e.g.: H-657305418.F.n200.tar

3) The tapes are currently being ingested in HPSS at Caltech.

Data Analysis Activities (Charlton)

Worked on Matlab code for collating and testing statistics of the FCT.  Tom would like to see how the values of the SNR are distributed from a large number of trials of a (fixed) chirp + noise (where new noise is generated for each trial). We anticipate that the distribution should be Ricean (that is, non-central chi²) when a chirp is present and chi² degree 2 when no chirp is present. I'm writing Matlab routines for generating chi² prob. distribution function and cumulative distribution function. To test the goodness-of-fit I'm writing code for a chi² test and the Kolmogorov-Smirnoff test.

General Computing

MIT:

• Getting things ready to setup ILOG service on one of the servers. The disk for this service should arrive the first of next week if not before.

• Shannon is working on the control room checklist, and is adding functionality to it. (Data entry is working, but data retrieval forms are still being written.)
• We have ordered extra SCSI drivers and a disk to allow us to move the web and mail services to separate machines, and to provide some redundancy for them.
• We are expecting an evaluation copy of "Quick Restore" to help automate backups on our new tape robot.
• We are configuring a workstation to be a standalone Solaris 7 machine for running the E2E software, which seems to be having problems with Solaris 8.

• Larry spent a few days assisting Christine during the E2 run. Mostly, taking care of minor problems users where having with e-mail and moving data around.
• Albert and Larry are working on some of the logistics to start getting a new network line up and working.

(Lisa)

• Worked out the kinks on the sunpci card.  Got P-card to install on it.  Wrote up an html page with everything I know about sunpci.  Did I mention that I got p-card running on it?  It was kind of a big deal at the time. The SUNPCi card in the past was tolerable at best for use on the SUN box with this new s/w installation it proves to be quite useful.
• Installed a new scanner in DCC.  Now both the ricoh and the new HP scanner can interface directly through Acrobat 4 and Photoshop.  Gave Linda a small tutorial on using both.  Installed HP jetadmin to get the printers up.  For some reason it wouldn't recognize them off of sirius.
• Swapped out some old sun workstations from the 3rd floor of Bridge.
• Larry and I have decided to go with Sun Solstice Backup as our backup software. The price on it was about $10K less than Quick Restore.  We will use this in conjunction with some home-brew scripts.

Worked a few PC issues. Basically, busy with school.

(Suresh)

• Restarted network time protocol (NTP) server in sirius synchronizing the time with LIGO router and three NTP servers at Caltech. Most of the LIGO servers and some workstations are already the client of this NTP daemon, work is continuing to complete with remaining systems.
• Installed and testing a popular network security package called "SAINT" in system homam. It reports security related bugs in system and suggest solutions to fix them. I have already tried in some systems and proved to quite informative.
• Work with Larry to unpack, install and configure a high end HP plotter (HP DesignJet 1055CM). It has been assigned with IP address  and named hpplotter1. It will be moved to second floor west bridge.
• Create a symbolic link to correctly point to Acrobat Reader version 4.

• Reworked all the undergraduate student pages.  Added ppt documents to LSC transparencies pages.  Created MG9 page.  Updated LIGO accounts pages.
• Created bibliographic reports for Liz from Pubs database.  Worked with Ed C. to resolve issues about the equipment database.  Met with Linda and Larry to plan DCC conversion to Office 2000.

• Spent a number of hours working out procurement issues. The worst being a order placed with SUN that got lost. It took us quite a bit of time to get the problem cleared up. Placed a few orders for upgrading the DCC equipment and have started receiving some of the items. Ordered a number os single external disk drives for different locations needing them for specialty use.
• Had an enjoyable experience assisting others for a few days on the E2 run.
• Item of note; the navidad.exe virus has hit a number of people on Campus. If  you receive it be sure to delete it, do not run it.
• Worked a number of financial issues, mostly with old accounts.
• Working on upgrading a couple of machines PC and Unix workstations.
• Spent time working some DCC and pdf issues. The conversion from PDF version 4 to version 3 is not as clean as it should be. There are a number of people working on getting the issues worked out.

LIGO II/Advanced R&D (Sanders)

(Helena Armandula)

Phil Willems is in Glasgow this week.  Nothing significant to report.  Working on the SUS schedule.

Seismic Attenuation (Riccardo DeSalvo)

No report.

For additional information about this report, contact sanders@ligo.caltech.edu