Report of the HAM-SAS Evaluation Committee

LASER INTERFEROMETER GRAVITATIONAL WAVE OBSERVATORY LIGO Laboratory / LIGO Scientific Collaboration LIGO LIGO- M070061-00-W April 20, 2007 Report of the HAM­SAS Evaluation Committee Rana Adhikari, Mark Barton, Doug Cook, Peter Fritschel, Joe Giaime, Richard Mittleman, Norna Robertson, Fred Raab (chair) Distribution of this document: LIGO Science Collaboration This is an internal working note of the LIGO Project California Institute of Technology LIGO Project – MS 18-34 1200 E California Blvd Pasadena, CA 91125 Phone (626) 395-2129 Fax (626) 304-9834 E-mail: info@ligo.caltech.edu Massachusetts Institute of Technology LIGO Project – NW17-161 175 Albany St Cambridge, MA 02139 Phone (617) 253-4824 Fax (617) 253-7014 E-mail: info@ligo.mit.edu LIGO Hanford Observatory P.O Box 159 Richland WA 99352 Phone 509-372-8106 Fax 509-372-8137 LIGO Livingston Observatory P.O Box 940 Livingston, LA 70754 Phone 225-686-3100 Fax 225-686-7189 http://www.ligo.caltech.edu/ LIGO LIGO-T03xxxx-00-D Introduction This is the report of the HAM-SAS Evaluation Committee, consisting of Rana Adhikari, Mark Barton, Doug Cook, Peter Fritschel, Richard Mittleman, Norna Robertson, and Fred Raab (chair) The committee was requested by David Shoemaker to evaluate an experimental test of a HAMSAS installation at LASTI The full charge to the committee is reproduced here as Appendix Essentially the committee was asked to evaluate the capability of the HAM-SAS system to meet the HAM seismic isolation system (SEI) requirements on basis on prototype test results and simulations/analyses Fourteen questions were posed to guide this evaluation The general consensus of the committee is given in section below and each of the questions is separately addressed in a corresponding subsection of section A very tight schedule for the experiment and committee evaluation was dictated by the need to proceed with orders for vibration isolation for the Enhanced LIGO upgrade and for Advanced LIGO The experimental team provided us with three reports We received “Report to the HAM SAS Evaluation Committee” (T070079-00) and “HAM-SAS Prototype Installation and Commissioning Experience” (T070080-00) on April Based on a reading of these reports, the evaluation committee prepared a list of questions to clarify information or issues raised by the first report, which it issued on April 11 Because of difficulties experienced by the experimental team and a tight and unforgiving schedule for obtaining results, the team was unable to achieve the desired state of completeness in the experimental work The committee also presented to the team a prioritized listing of results that the team might be able to pursue in the remaining time, which would still allow time for the committee to incorporate into its evaluation, without seriously slipping the required reporting date The committee discussed this with the experimental team in a telecom on April 13 and received the team’s “Second Report to the HAM-SAS Evaluation Committee” (T070087-00) on April 19 Findings The committee acknowledges the very hard work of HAM-SAS experimental team, which resulted in major accomplishments on a short time scale The accomplishments of the HAM SAS, LASTI and CDS teams, supported by external contractors and other LIGO personnel, are impressive This sort of work is exactly the multi-disciplinary engineering and commissioning that we are expected to for many aspects of Advanced LIGO, and this effort showed it could be done within a highly compressed schedule and using facilities shared by several other projects Much was learned in the processes for construction of the HAM-SAS system: construction techniques, installation practices, vacuum prep techniques Unfortunately, the major experimental points that needed to be tested or demonstrated were not achieved in time The team was able to demonstrate successful installation of HAM-SAS into a HAM chamber and tuning of the resonances to low frequencies with a reasonable degree of stability However, the team was not able to demonstrate that isolation to the HAM requirement (T060075-00-D) could be achieved This should not be interpreted as showing that the required isolation is impossible to achieve, but rather as having serious performance shortcomings that might or might not be overcome by further work LIGO LIGO-T03xxxx-00-D 2.1 Vacuum Compatibility The HAM-SAS assembly materials and handling procedures adhered to the LIGO specifications for in vacuum components currently in force as detailed in section 3.1 of T070079-00-D 2.2 Form and Fit The system has been shown to fit correctly in a LIGO HAM at LASTI, attaching to the standard support tubes The report states that for HEPI-equipped HAM chambers, some tall assemblies (such as the triple pendulums) would have to be installed after the HAM SAS were inserted 2.3 Effectiveness and Ease of Installation/Tuning This first installation went reasonably well for a first installation, but the tuning and troubleshooting showed the delicate nature one expects of any low-frequency, multiple-degree of freedom isolation system There was no evidence in the installation or tuning to indicate that this system is easier to work with than the baseline system 2.4 Seismic Attenuation The team was able to demonstrate compliance with the HAM requirements without use of HEPI for the motion in horizontal degrees of freedom (transverse translations and yaw) from 0.8 to Hz, even in the LASTI environment which is noisier than the observatories at these frequencies Vertical motion was not shown to be as good However the data below 0.2 Hz – especially in the horizontal degrees of freedom – showed serious excesses by orders of magnitude from the required motion, even though the LASTI environment is similar to that found at the observatories in that frequency range The data indicate that the table surface moves by tens of microns of RMS Because of further experimental difficulties it was not resolved whether this was caused by horizontal or tilt couplings or whether this could be mitigated sufficiently by retuning of the resonance or by electronic controls (However, given the magnitude of the motion one would expect that a purely electronic solution would inject intolerable noise at higher frequencies.) There were also serious deviations from the HAM requirement above Hz that appeared to be due to a multitude of mechanical resonances These might be able to be damped successfully and we note that effective damping of some mechanical resonances at these frequencies has been achieved in past experimental work Unfortunately, there was not enough time to attempt damping Until such damping is achieved one never knows whether there are additional resonances, due to differences from past experiments, which may not be amenable to simple damping methods used before Given the current state of knowledge, the committee cannot say that the HAM isolation requirements are achievable using the HAM-SAS system 2.5 Long Term Pointing and Control Stability Data were collected using an optical lever with the triple suspension mirror locked down Unfortunately the data were not analyzed at the time of this report LIGO LIGO-T03xxxx-00-D 2.6 Efficacy of Safety Locks and Limiters A convenient method has been provided to lock down the table when necessary and this was used prior to evacuation of the chamber From the description of page 13 of T070080, it appears to provide a robust method of locking down the table However, it was not tested whether one could align an optic to a particular orientation (for instance using an optical lever), lock and unlock the table and preserve the original orientation to within reasonable limits With regard to earthquake safety, it appears there is sufficient clearance around the stops to prevent a collision with the stops in an earthquake, if one assumes that the isolation at low frequencies is able to achieve the requirement However the experimental data showed amplification of motion at low frequencies Since it is not known how this would be mitigated, we cannot assert that the earthquake safety stop strategy will work as planned In addition, the stop gaps are so large that it is likely, in the event of earthquake induced tilts, that the triple pendulum would run into its stops repeatedly This is a failure mode which is amenable to mitigation by use of proper watchdog software and failure analysis 2.7 Margin Against Instability We not have the relevant experimental data to say much about long term stability The system showed no instability at the tuning used for the tests that were reported, but this tuning was not sufficient to achieve low-frequency requirements Instability becomes more likely as the system is tuned to lower frequency, since the tuning essentially subtracts two large spring constants, allowing a small drift to drive the system from stability to instability To test this would have required that the tuning and controls needed to achieve the HAM requirements be used and then the range of tolerance to errors be assessed 2.8 DC Compatibility of HAM-SAS and Triple installation The triple pendulum was locked down for the duration of the testing and so there is no experimental data relating to DC compatibility 2.9 Coupled Dynamic Interactions of HAM-SAS and Triple Pendulum A long-time concern with the HAM-SAS concept has been the potential for complicated interactions between HAM-SAS and suspension payloads, or between suspensions mounted on a HAM-SAS Unfortunately no tests have been performed on the HAM-SAS to investigate any such interactions, however, T070087 does contain two simplified calculations on the issue Both calculations suggest there could exist a large coupling (of order unity or even greater) between multiple suspended optics on a single HAM-SAS platform These calculations are surely oversimplified, but nonetheless the committee is extremely concerned that such couplings could greatly complicate the global interferometer control, or even lead to an unworkable control situation For this system to be a viable candidate for LIGO, this concern would need to be successfully addressed (either solved or shown not to be a significant problem) with further modeling as well as direct testing LIGO LIGO-T03xxxx-00-D 2.10 Ease of Integration into a Hierarchical Control System Given the committee's goal to evaluate HAM SAS for use in Advanced LIGO, we need to consider both types of HAM SAS use, with and without HEPI present In Advanced LIGO, the payloads are planned to be supported on HEPI systems At LLO, these are already in place and are used in a “sensor correction” scheme to reduce noise in the 0.1 - Hz band The in-vacuum payloads, particularly the core optics, are commanded by the sensor correction signal There are two options; the first is to eliminate HEPI, requiring that the HAM SAS system be used to command its payload to track the HEPI sensor correction signal (and therefore the core optics) The committee has not seen evidence that the response of the HAM-SAS to those commands is compatible with this function; in particular, performance can be compromised by RHP zeros near the active band, if they are present We just don't know Alternatively, one can consider a second option installing HAM SAS on top of HEPI, letting HEPI its present job and using HAM SAS to isolate at higher frequencies By design, the analysis or demonstration of this option was not part of the HAM SAS effort that we have been asked to evaluate As stated in 2.9, there are possibly problematic areas with regards to control integration Three examples are listed here:  On the HAM2 table, there will probably be two triple pendula; MC2 and a PRM As shown on pg.12 of T070087, there is a significant coupling between two such suspended optics It is probably true that this can be mitigated at low frequencies by implementing a crosscoupling matrix to link the MC and LSC sub-systems Since the Common mode servo puts the damped motions of the arm cavity common mode onto the MC length, the control signals on MC2 are roughly equivalent to the arm length motion divided by that length ratio (~4000/16) This is probably enough to ensure that the cross coupling between the BSCHEPI and the PRM HAM-SAS via the cross talk of the PRM / MC2 suspensions is not too large to produce instabilities  The passive initial LIGO stacks are approximately rigid below Hz and so the common mode rejection of low frequency seismic noise is > 10 At LLO this is preserved with the HEPI system by tuning the STS based feed forward to reduce differential tank motions If the low frequency noise in the HAM-SAS comes from something not common to the tables at the level of 10%, then the excess motion is likely to require some bandwidth increase in the ASC loops and actuation authority re-allocation in the SUS electronics  The design of the BS/PRM/SRM/MC SUS is made easier by assuming that there will be control re-allocation between the SUS and SEI systems in the f < Hz band as is done at LLO in initial LIGO The HAM SAS system has enough control force available to track the low frequency (f < 0.2 Hz) motions What is not clear, however, is the level of diagonalization of the actuators above the bandwidth of the LVDT loops In addition, the complexity of sending global correction signals into a MIMO system with unity gain frequencies that are in the correction band is not a negligible delta to the commissioning complexity of the ISC/SEI sub-systems LIGO LIGO-T03xxxx-00-D 2.11 Effect of Cable Compliance on Performance The Report states that the system exhibited "no problem" with a number of cables present It is difficult to analyze this, especially in the case where a lower passive resonance is used for the inverted pendulums On the other hand, the committee doubts that any cable stiffness problems would be a show-stopper, especially if one allows the possibility of cable redesign 2.12 Adaptability to Handle Payload Variations The system design appears to be able to handle the maximum payload requirement and payload shifts although the LASTI tests may need additional weights to meet the horizontal modes eignenfrequency goals The LASTI HAM chamber was set up to represent one of the worst case Advanced LIGO scenarios Even with the loose ‘roll bar’ the table was balanced without steppers and maintained balance for days utilizing only passive components This was performed during ambient temperature swings of ~8 degrees F and with the chamber vented A uniform gap was maintained on all four corners stops to ~ 0.5mm The final balance was achieved using a handful of 3/8” machine nuts and other very light counter weights to fine tune it mechanically With the discovery of the loose roll bar fixed and tuning of the spring tension, the ease of balancing the table should improve 2.13 Reliability/Serviceability Some actuators are difficult to reach, but failure is unlikely Pre-installation testing should eliminate most issues Mechanical make up is simple so failure should be minimal Over current protections could be handled in the same way as the initial LIGO SUS coil drivers Other components are mechanically robust, but some have room for minor improvements 2.14 Cost and Schedule We note that costs for the systems proposed for use in Enhanced LIGO have been provided in section 3.14 of the first report to the committee (T070079) We are confident that these costs have been estimated reliably for what has been included, and could be scaled for the number of units required in Advanced LIGO However, we note that if feedforward techniques are required in the HAM-SAS system to achieve the required performance then further costs need to be added to the unit price of a system A very rough estimate of what might be the added cost per system is the sum of 1) to 4) below minus savings of 15.5K on LVDTs = 57 - 15.5 = $41.5K 1) Fraction of STS-2 = $9.0K ( based on $17K per STS-2 scaled by number needed over and above those already in place + spares = 8, divided by number of HAM chambers to be fitted = 15) 2) ADC boards $4.2K (based on worst case per building at $10.5 K times buildings divided by 15) 3) Fraction of Interface module for STS-2 $0.8K (based on cost =$1.58K times divided by 15) 4) Low noise capacitive displacement sensors (ADE) to replace LVDTs = $43 K (based on $32K for times 8/6) LIGO LIGO-T03xxxx-00-D If inductive sensors (such as used in HEPI) rather than capacitive sensors could be used, cost is similar to LVDTs, and hence extra cost is sum of 1) to 3) = 14K per system The production schedule for Enhanced LIGO is addressed in Section 3.14 of the first report to the committee (T070079) We believe that a reasonable estimate of production time to obtain an assembled system could be inferred from this information However, there are uncertainties in what the commissioning schedule would look like given that there are open questions as noted earlier in our report (2.4 – 2.10) In particular, we not have enough information to estimate the time needed for any control system developments and adjustments needed to satisfy performance requirements Thus we are not able to make a statement on how the schedule for HAM-SAS would compare to the baseline schedule assumed in the current Advanced LIGO project Appendix 1: Charge to the HAM-SAS Evaluation Committee The following is the charge to the HAM-SAS evaluation committee: The baseline approach employs the HAM-ISI (internal seismic isolation) system coupled with HAM-HEPI (hydraulic, external pre- isolator) HAM-SAS (seismic attenuation system) is an alternative seismic isolation concept for the AdL HAM chamber (Note that HAM-SAS can be used with or without the HEPI system.) This committee is asked to evaluate the capability of the HAMSAS system to meet the HAM seismic isolation system (SEI) requirements on basis on prototype test results and simulations/analyses The HAM-SAS team will present the results of the prototype effort to the committee in a written report by April 2nd The committee is asked to review the report (as well as previous reviews and evaluations of HAM-SAS) and provide written questions to the HAM- SAS team by April 9th The HAM-SAS team will then meet with the HAM-SAS evaluation committee, on or about April 16th, to respond to the committee's questions In evaluating the HAM-SAS compliance with the HAM SEI requirements, the committee is explicitly asked to consider the following factors: Vacuum compatibility of the HAM SAS system, including materials compatibility, compensation for buoyancy, and implications of floor tilt Form and fit of the HAM SAS system in a LIGO HAM chamber, including compatibility with the existing support structure (piers, scissor tables, crossbeams, support tubes and (at LLO only) the HEPI system) Effectiveness and ease (duration, skills required, repeatability/predictability) of the HAM SAS installation and 'tuning', characterization, and integration/commissioning Seismic attenuation factor of the HAM SAS, including total rms motion (without a HEPI system) and ground tilt sensitivity and any indication of self-generated noise (electronic/controls, and/or mechanical) Long term pointing stability and long-term control stability (as a function of temperature, floor tilt, other accessible and likely variables) Efficacy of optics table mechanical "locks" and limiters for safety (integrating optics table payloads) and for earthquakes Margin against instability LIGO LIGO-T03xxxx-00-D DC compatibility of the HAM SAS and triple installation (mechanically locking HAM-SAS and aligning to an optical reference) Controllable or negligible coupled dynamic interaction of the HAM-SAS and a triple pendulum suspension 10 Judgment on the ease of integration into a hierarchical control system (tidal/microseism tracking, quasi-DC suspension force minimization, etc.) 11 The effect of cable compliance on the isolation performance 12 Adaptability, extensibility, reconfigurability to handle payload variations (position and range in mass properties) 13 Reliability and reparability/serviceability of the in-vacuum sensors and actuators 14 Cost and schedule The committee is not asked to review management, staffing, or contingency The committee is also not asked to make a comparison with the baseline The committee's findings should be provided in a written report to AdL management on or about April 20th AdL management will consider the committee's HAM-SAS evaluation and an implementation plan In addition AdL management will consider the status of the Baseline (HAMISI + HAM-HEPI) by getting a report from the SEI leadership, especially any news on meeting requirements, ability to fabricate, costs, reliability, technical updates/issues from the BSC and HAM variants David Shoemaker, AdL Project Leader, will then make a decision to either continue with the baseline approach or switch to the HAM-SAS approach, by about April 23rd ... simulations/analyses The HAM-SAS team will present the results of the prototype effort to the committee in a written report by April 2nd The committee is asked to review the report (as well as... previous reviews and evaluations of HAM-SAS) and provide written questions to the HAM- SAS team by April 9th The HAM-SAS team will then meet with the HAM-SAS evaluation committee, on or about... with the experimental team in a telecom on April 13 and received the team’s “Second Report to the HAM-SAS Evaluation Committee? ?? (T070087-00) on April 19 Findings The committee acknowledges the