From: Juna Kollmeier Date: Wed, 4 Aug 2021 15:15:22 -0700 Subject: Re: [infrastructure 20885] Re: [sloan-25mlog 5586] 2.5m obslog --000000000000170c5305c8c3248d Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Many thanks for this update and detailed explanation! What is a possible solution for fixing our pocked rollers? Juna On Wed, Aug 4, 2021 at 11:20 AM Katie wrote: > More details regarding the ATF tests yesterday below: > > >> > >> Motivation: A new alignment telescope fixture (ATF) built by JHU for > SDSS-V has not performed as intended. When attached to the cartridge > interface, the secondary fiducial focused through the alignment telescope > should trace out a circular orbit as the telescope rotator is rotated. > Fred M. determined some months back that the orbit shape was triangular, > which indicates that the alignment scope tilt is changing with rotator > angle. Today the problem was investigated by JHU staff Robert B. Aidan G= . > and Steve S. with help from APO staff Ryan W. and Katie G. > >> > >> The ATF Interface: > >> > >> The ATF uses the SDSS cartridge interface, as does the FPS, a > pseudo-kinematic interface comprised of three v-groove blocks at the > cartridge outer diameter, which engage three v-roller assemblies mounted = to > the rotator flange. A pneumatic clamp latches the cartridge when actuated > and pulls the v-groove tight against the v-roller assembly, > >> > >> > >> Theory 1: Excessive clamping force from pneumatic clamps was the > problem. > >> > >> Were the pneumatic clamps somehow causing the problem, possibly > inducing a variable distortion in the rotator? No. > >> > >> To determine this, a set of mechanical screw clamps were implemented t= o > allow the ATF to be clamped in place with light preload, The alignment > scope was trained on the secondary fiducial and the centroid of that > fiducial was tracked w/to the alignment telescope crosshair as the rotato= r > was rotated through 360 degrees. Data was taken every 45 degrees. The > pattern recorded during this exercise was a triquetra, not a round orbit = or > even a triangular orbit. The radial deviation of the pattern was > approximately 10 arcseconds. This procedure was repeated and the second > run produced near identical results. It was repeated a third time with t= he > pneumatic clamps engaged, with ostensibly the same results. Dial > indicators positioned to measure clamp body motion showed no change when > the pneumatic clamps were engaged, relative to the mechanical screw clamp= s. > >> > >> > >> More Investigation: > >> > >> The results from the testing just described led to further > investigation into the motion of the ATF as the rotator slews from 0 - 36= 0 > degrees. Of particular interest here was an answer to the question: Is > the ATF base plate (the large violet plate) moving relative to the rotat= or > with slew angle? The answer is yes. And this is a problem for the ATF an= d > may be a problem for the FPS as well. > >> > >> Two sets of measurements were taken. > >> > >> First. Three dial indicators were set to measure piston motion of the > ATF base plate (BP) at the clamp locations. To take these measurements, t= he > indicator stand base was magnetically clamped to the rotator flange and t= he > dial indicator tip was positioned on the bottom surface of the BP. Once = at > each of the three v-groove interfaces. Motions were monitored at each > clamp location over a 360 slew of the rotator. Motions varied depending = on > the interface; from roughly 0.0003=E2=80=9D to 0.0007". It was determine= d from the > measurements that the maximum BP tilt was approximately 8 arcseconds, ver= y > nearly the motion seen in the alignment scope. > >> > >> Second. The three dial indicators were repositioned to measure the > radial motion of the BP. Motions varied depending on the interface. A > maximum motion of 0.002=E2=80=9D (min to max) was recorded over a 360 sle= w of the > rotator. > >> > >> In short, the ATF base plate is moving both axially and radially and > the amounts vary from one interface location to the next. The tilt induc= ed > is very likely responsible for the change in angle seen in the alignment > scope. > >> > >> > >> Further Investigation and Theory for Root Cause: > >> > >> A close visual examination of the "v-groove" block interfaces reveal > smooth interface surfaces. No cause for concern there. However, a visua= l > examination of the "v-roller" contact surfaces reveals what looks to be > dimpling on the contact surfaces. Inspection is difficult with the rolle= r > assembly attached. A more careful inspection of these surfaces will > require removal of the roller bearing cage, and we recommend this be done= . > >> > >> Dimpling of the v-roller surfaces could be the root cause for the > ATF/FPS tilt. If rollers are seated in a dimple, the roller will lift ou= t > of the dimple with radial motion and that lift will tilt the ATF/FPS. Th= e > radial motion seen here is likely the results of slight out of roundness = in > the very large rotator bearing. It should be said that the cartridge > interface is designed to accommodate small radial motions, e.g. out or > roundness or thermal excursions and material CTE mismatches. It is a sol= id > design concept. However, if either of the contact surfaces that make up t= he > v-groove interface are dimpled then radial motions will change the > piston/tilt of the attached instrument. > >> > >> It is very likely that the v-roller assemblies are original, dating > back more than twenty years. Dimpling of the contact surfaces would not = be > surprising given the 15,000 plus cartridge exchanges and thousands of > camera exchanges that have taken place over all these years. > >> > >> > >> > >> > >> > >> > >> > >> ---------------- > >> Stephen Smee, Ph.D. > >> Principal Research Scientist > >> Instrument Development Group > >> Department of Physics and Astronomy > >> Johns Hopkins University > >> 3701 San Martin Drive > >> Baltimore MD, 21218 > >> > > > On Aug 4, 2021, at 3:54 AM, kinemuchi@apo.nmsu.edu wrote: > > > > > > APACHE POINT OBSERVATORY SDSS 2.5M > > > > OBSERVING LOG > > Tuesday August 03, 2021 > > > > (SJD 59430) > > > > ---- Observing Team ---- > > > > Viktor Malanushenko > > Karen Kinemuchi > > > > > > > > Katie Grabowski > > > > JHU Team > > =3D=3D=3D=3D=3D=3D=3D=3D > > Steve Smee > > Robert Barkhouser > > Aidan Gray > > > > ---- Observing Plan ---- > > > > Finish optical collimation, if the weather permits. > > > > > > > > ---- Observing Summary ---- > > > > > > Entire night lost to weather. > > > > JHU tested the ATF until 6:00Z. > > > > ---- Observing Log ---- > > > > Afternoon: > > =3D=3D=3D=3D=3D=3D=3D=3D=3D > > 23:52Z ATF on the telescope, Steven Smee is working with the telescope. > > > > 01:30Z Restarted the TPM to get weather information to populate the > > weather.py page. Got this error: > > > > FIXME reply cw multi OK > > comm unknown TPM_MULTI > > FIXME reply cw multi OK > > comm unknown TPM_MULTI > > > > Which apparently killed the MCP. > > > > Followed the MCP restart procedure and it did not appear to restart > anything. > > Reconnected TCC with "tcc axis connect" and TCC is fine. > > > > With some help from DanO, we redid the procedure for MCP restart after > > killing duplicated processes. Restarted TPM again and reconnected tcc. > > > > MCP, TPM, TCC, and the weather information are all back to normal. > > > > 01:57Z JHU crew is still working at the telescope, but humidity has > risen. > > Dewpoint differential is at 1.5C now. 3.5m is also closed, so we will > wait > > for the humidity to drop. > > > > 04:20Z Cleared up and humidity is dropping, but the JHU team is still > working. > > More radar and storms on the way. Seeing lightning in the distance. > > 06:00Z JHU team has given up for the night. They had been testing the > ATF > > until now. > > > > Engineering camera is mounted on the telescope, but not connected. > > > > Radar and lightning approaching. > > > > 07:56Z Storm approaching. Turned off amplifiers. Lots of lightning. > > 09:53Z Giving up. > > > > ---- Mirror Numbers ---- > > > > > > No mirror numbers taken due to work on the telescope. > > > > ---- Telescope Status ---- > > > > Never opened. > > > > > > ---- Software Used ---- > > > > Get Versions, 08/03/21, 23:43:32 > > -- STUI software: > > STUI : 1.7.13 > > alerts : trunk > > apo : trunk > > boss : unknown > > -boss.specMec : 3.0.4 > > gcamera : FAILED > > guider : unknown > > hartmann : 1.8.0 > > hub : Tag: v4.2 > > mcp : 2.0-7-g8799 > > -mcp.plc : 34 > > -mcp.azFiduci : v1_145 > > -mcp.altFiduc : v1_145 > > -mcp.rotFiduc : v1_145 > > platedb : 2.8.0 > > sop : 4.0.8 > > tcc : 1.4.0 > > apogee : not availble > > apogeecal : not availble > > apogeeql : unknown > > STUIAdditions : 3.1.7 > > > > -- Other versions ---- > > idlspec2d : v5_7_12 > > plate-mapper3 : NOCVS:v7_0_9 > > petunia : 2.2.3 > > autoscheduler : git > > sdss_python_module: trunk > > mangadrp : trunk exported > > > > > > > > ____________________________________________________________ > > Logs of Nightly APO Activities > > > https://trac.sdss.org/tracmailman/browser/private/sloan-25mlog/index.html > > ____________________________________________________________ > Common APO Hardware and Software > https://trac.sdss.org/tracmailman/browser/private/infrastructure/index.ht= ml > --000000000000170c5305c8c3248d Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Many thanks for this update and detailed e= xplanation!What is a po= ssible =C2=A0solution for fixing our pocked rollers? =C2=A0Juna=C2=A0On Wed, Aug 4, 2021 = at 11:20 AM Katie wrote:More details regardi= ng the ATF tests yesterday below: >> >> Motivation: A new alignment telescope fixture (ATF) built by JHU f= or SDSS-V has not performed as intended.=C2=A0 When attached to the cartrid= ge interface, the secondary fiducial focused through the alignment telescop= e should trace out a circular orbit as the telescope rotator is rotated.=C2= =A0 Fred M. determined some months back that the orbit shape was triangular= , which indicates that the alignment scope tilt is changing with rotator an= gle.=C2=A0 Today the problem was investigated by JHU staff Robert B. Aidan = G. and Steve S. with help from APO staff Ryan W. and Katie G. >> >> The ATF Interface: >> >> The ATF uses the SDSS cartridge interface, as does the FPS, a pseu= do-kinematic interface comprised of three v-groove blocks at the cartridge = outer diameter, which engage three v-roller assemblies mounted to the rotat= or flange. A pneumatic clamp latches the cartridge when actuated and pulls = the v-groove tight against the v-roller assembly, >> >> >> Theory 1: Excessive clamping force from pneumatic clamps was the p= roblem. >> >> Were the pneumatic clamps somehow causing the problem, possibly in= ducing a variable distortion in the rotator?=C2=A0 No.=C2=A0 >> >> To determine this, a set of mechanical screw clamps were implement= ed to allow the ATF to be clamped in place with light preload,=C2=A0 The al= ignment scope was trained on the secondary fiducial and the centroid of tha= t fiducial was tracked w/to the alignment telescope crosshair as the rotato= r was rotated through 360 degrees.=C2=A0 Data was taken every 45 degrees.= =C2=A0 The pattern recorded during this exercise was a triquetra, not a rou= nd orbit or even a triangular orbit. The radial deviation of the pattern wa= s approximately 10 arcseconds.=C2=A0 This procedure was repeated and the se= cond run produced near identical results.=C2=A0 It was repeated a third tim= e with the pneumatic clamps engaged, with ostensibly the same results.=C2= =A0 =C2=A0Dial indicators positioned to measure clamp body motion showed no= change when the pneumatic clamps were engaged, relative to the mechanical = screw clamps. >> >> >> More Investigation: >> >> The results from the testing just described led to further investi= gation into the motion of the ATF as the rotator slews from 0 - 360 degrees= .=C2=A0 Of particular interest here was an answer to the question: Is the= =C2=A0 ATF base plate (the large violet plate) moving relative to the rotat= or with slew angle? The answer is yes.=C2=A0 And this is a problem for the = ATF and may be a problem for the FPS as well. >> >> Two sets of measurements were taken.=C2=A0 >> >> First. Three dial indicators were set to measure piston motion of = the ATF base plate (BP) at the clamp locations. To take these measurements,= the indicator stand base was magnetically clamped to the rotator flange an= d the dial indicator tip was positioned on the bottom surface of the BP.=C2= =A0 Once at each of the three v-groove interfaces.=C2=A0 Motions were monit= ored at each clamp location over a 360 slew of the rotator.=C2=A0 Motions v= aried depending on the interface; from roughly 0.0003=E2=80=9D to 0.0007&qu= ot;.=C2=A0 It was determined from the measurements that the maximum BP tilt= was approximately 8 arcseconds, very nearly the motion seen in the alignme= nt scope. >> >> Second.=C2=A0 The three dial indicators were repositioned to measu= re the radial motion of the BP.=C2=A0 Motions varied depending on the inter= face.=C2=A0 A maximum motion of 0.002=E2=80=9D (min to max) was recorded ov= er a 360 slew of the rotator. >> >> In short, the ATF base plate is moving both axially and radially a= nd the amounts vary from one interface location to the next.=C2=A0 The tilt= induced is very likely responsible for the change in angle seen in the ali= gnment scope. >> >> >> Further Investigation and Theory for Root Cause: >> >> A close visual examination of the "v-groove" block inter= faces reveal smooth interface surfaces.=C2=A0 No cause for concern there.= =C2=A0 However, a visual examination of the "v-roller" contact su= rfaces reveals what looks to be dimpling on the contact surfaces.=C2=A0 Ins= pection is difficult with the roller assembly attached.=C2=A0 A more carefu= l inspection of these surfaces will require removal of the roller bearing c= age, and we recommend this be done. >> >> Dimpling of the v-roller surfaces could be the root cause for the = ATF/FPS tilt.=C2=A0 If rollers are seated in a dimple, the roller will lift= out of the dimple with radial motion and that lift will tilt the ATF/FPS.= =C2=A0 The radial motion seen here is likely the results of slight out of r= oundness in the very large rotator bearing.=C2=A0 It should be said that th= e cartridge interface is designed to accommodate small radial motions, e.g.= out or roundness or thermal excursions and material CTE mismatches.=C2=A0 = It is a solid design concept. However, if either of the contact surfaces th= at make up the v-groove interface are dimpled then radial motions will chan= ge the piston/tilt of the attached instrument.=C2=A0 >> >> It is very likely that the v-roller assemblies are original, datin= g back more than twenty years.=C2=A0 Dimpling of the contact surfaces would= not be surprising given the 15,000 plus cartridge exchanges and thousands = of camera exchanges that have taken place over all these years. >> >> >> >> >> >> >> >> ---------------- >> Stephen Smee, Ph.D. >> Principal Research Scientist >> Instrument Development Group >> Department of Physics and Astronomy >> Johns Hopkins University >> 3701 San Martin Drive >> Baltimore MD, 21218 >> > On Aug 4, 2021, at 3:54 AM, kinemuchi@apo.nmsu.edu wrote: > > >=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 = =C2=A0 =C2=A0APACHE POINT OBSERVATORY SDSS 2.5M > >=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 = =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0OBSERVING LOG >=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 = =C2=A0 =C2=A0 =C2=A0 Tuesday August 03, 2021 > >=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 = =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 (SJD 59430) > > ----=C2=A0 Observing Team ---- > > Viktor Malanushenko > Karen Kinemuchi > > > > Katie Grabowski > > JHU Team > =3D=3D=3D=3D=3D=3D=3D=3D > Steve Smee > Robert Barkhouser > Aidan Gray > > ---- Observing Plan ---- > > Finish optical collimation, if the weather permits. > > > > ---- Observing Summary ---- > > > Entire night lost to weather. > > JHU tested the ATF until 6:00Z. > > ---- Observing Log ---- > > Afternoon: > =3D=3D=3D=3D=3D=3D=3D=3D=3D > 23:52Z ATF on the telescope, Steven Smee is working with the telescope= . > > 01:30Z=C2=A0 Restarted the TPM to get weather information to populate = the > weather.py page.=C2=A0 Got this error: > > FIXME reply cw multi OK > comm unknown TPM_MULTI > FIXME reply cw multi OK > comm unknown TPM_MULTI > > Which apparently killed the MCP. > > Followed the MCP restart procedure and it did not appear to restart an= ything. > Reconnected TCC with "tcc axis connect" and TCC is fine. > > With some help from DanO, we redid the procedure for MCP restart after= > killing duplicated processes.=C2=A0 Restarted TPM again and reconnecte= d tcc. > > MCP, TPM, TCC, and the weather information are all back to normal. > > 01:57Z=C2=A0 JHU crew is still working at the telescope, but humidity = has risen. > Dewpoint differential is at 1.5C now.=C2=A0 3.5m is also closed, so we= will wait > for the humidity to drop. > > 04:20Z=C2=A0 Cleared up and humidity is dropping, but the JHU team is = still working. > More radar and storms on the way.=C2=A0 Seeing lightning in the distan= ce. > 06:00Z=C2=A0 JHU team has given up for the night.=C2=A0 They had been = testing the ATF > until now. > > Engineering camera is mounted on the telescope, but not connected. > > Radar and lightning approaching. > > 07:56Z=C2=A0 Storm approaching.=C2=A0 Turned off amplifiers.=C2=A0 Lot= s of lightning. > 09:53Z=C2=A0 Giving up. > > ---- Mirror Numbers ---- > > > No mirror numbers taken due to work on the telescope. > > ---- Telescope Status ---- > > Never opened. > > > ---- Software Used ---- > > Get Versions, 08/03/21, 23:43:32 >=C2=A0 -- STUI software: > STUI=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 : 1.7.13 > alerts=C2=A0 =C2=A0 =C2=A0 =C2=A0 : trunk > apo=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0: trunk > boss=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 : unknown > -boss.specMec : 3.0.4 > gcamera=C2=A0 =C2=A0 =C2=A0 =C2=A0: FAILED > guider=C2=A0 =C2=A0 =C2=A0 =C2=A0 : unknown > hartmann=C2=A0 =C2=A0 =C2=A0 : 1.8.0 > hub=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0: Tag: v4.2 > mcp=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0: 2.0-7-g8799 > -mcp.plc=C2=A0 =C2=A0 =C2=A0 : 34 > -mcp.azFiduci : v1_145 > -mcp.altFiduc : v1_145 > -mcp.rotFiduc : v1_145 > platedb=C2=A0 =C2=A0 =C2=A0 =C2=A0: 2.8.0 > sop=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0: 4.0.8 > tcc=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0: 1.4.0 > apogee=C2=A0 =C2=A0 =C2=A0 =C2=A0 : not availble > apogeecal=C2=A0 =C2=A0 =C2=A0: not availble > apogeeql=C2=A0 =C2=A0 =C2=A0 : unknown > STUIAdditions : 3.1.7 > > -- Other versions ---- > idlspec2d=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0: v5_7_12 > plate-mapper3=C2=A0 =C2=A0 =C2=A0: NOCVS:v7_0_9 > petunia=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0: 2.2.3 > autoscheduler=C2=A0 =C2=A0 =C2=A0: git > sdss_python_module: trunk > mangadrp=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 : trunk exported > > > > ____________________________________________________________ > Logs of Nightly APO Activities > https://trac.sdss.org/= tracmailman/browser/private/sloan-25mlog/index.html ____________________________________________________________ Common APO Hardware and Software https://trac.sdss.org/tra= cmailman/browser/private/infrastructure/index.html --000000000000170c5305c8c3248d--