From: William A. Podgorski Date: Tue, 10 Feb 1998 14:05:24 -0500 To: (list) Subject: FACT - Revision 1 Cc: (list) Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-UIDL: 58e52c5b97cb51687424b7a89e7e3f18 John Polizotti gave me some updated information on the location of the HRC retro on the HARM. It is 9/16 in on the -Y side, or -14.3 mm in Y. I have revised the calculations to show this. The results only change by at most 0.2 mm. Bill Introduction The Focus Alignment Cross-Check Test (FACT) was performed at TRW on Feb 5 and 6. The procedure is documented in TRW's Special Task Procedure STP-L7-135, by Scott Texter. The essentials of it are as follows: 1) Set up the Sokkia NET2 laser rangefinder and an adjustable fold flat on an optical table below AXAF. 2) Adjust the fold flat and NET2 such that the flat is directly below one of the CAP gauge blocks and the NET2 is aligned with its' LOS normal to the surface of the block. This is done by viewing the image of the front of the NET2 thru the NET2 telescope and is quite difficult to do. The fold flat and NET2 are also aligned such that the line between them lies on a HRMA diameter, so that adjustment of the flat (or NET2 telescope) in elevation moves the LOS radially wrt the HRMA. 3) Range off the gage block using the NET2. 4) Adjust the fold flat and NET2 telescope so as to view the HRC alignment reference mirror (HARM). This is done thru apertures in two baffles, one at the end of the Aft HRMA structure and one at the end of the aluminum part of the AFT Inner cylinder. (These apertures, 24 in each baffle, were designed such that the centroid detector could penetrate to the HRMA focus). This requires some adjustment of the position of the HRC and careful lighting. Place the NET2 LOS on the HRC retro-reflector, which is built into the HRC HARM 5) Range off the HRC retro 6) These two range mesurements can be analyzed to give a measurement of absolute focus position which can be compared to measurements taken during telescope alignment, thus providing the cross-check. Data The measurement was made on two different gauge blocks. The first (#1) was on the block located in the (+Y, +Z) quadrant of the HRMA. The second (#2) was taken on the block in the (-Y, -Z) quadrant. Following is the data taken: Meas# SIM TT POS Gage Block Range(mm) Retro Range(mm) Raw Delta(mm) ----- ---------- -------------------- --------------- ------------- 1 -65360 3193.7 13138.0 9944.3 2 -22360 3195.9 13137.2 9941.3 Analysis I have analyzed the data with respect to Kodak's focus alignment data, supplied via a January 26, 1998 E-mail from Charlie Atkinson. The information below has been extracted from this E-mail. ARM Focus location -2.268 From HRMA final tests HRMA FOCUS 0 CCRF surface 360.4209 Based on ISIM and IAS tests PLANO surface 406.45834 Based on ISIM and IAS tests ISIM INTERFACE LOCATION 723.8085 Based on ISIM and IAS tests CAP DATUM -D- 10029.63 3/4/97 e-mail from me CAP CENTERLINE 10054.58 CAP DATUM -A- 10079.53 CAP thickness of 49.9mm ******Podgorski had 10079.2, using H6 end as datum******** ARM -X surface 10075.932 ARMF focal length & HRMA data ARM +X surface 10126.732 2.000" thickness of ARMF The key data above is the distance from CAP Datum A (P side of CAP) to the focus, which is 10079.53 mm according to Charlie. I had also calculated this number independently to be 10079.2 mm. In order to compare the Kodak focus data (10079.53 mm) to the NET2 data several adjustments and corrections have to be made. First, there is a common path between the NET2 and the fold mirror. Even though there are slight differences in path length due to adjuatments of the fold mirror and NET2 pointing direction, the diffetences are small. We can therefore directly subtract this common path from all measurements. It was not measured at the time, but a good estimate would be 500mm. Ranges to the fold flat from the gauge block or HARM retro then become: Raw(mm) Corrected(mm) -------- ------------- Meas 1 to Block 3193.7 2693.7 Meas 1 to HARM retro 13138.0 12638.0 Meas 2 to Block 3195.9 2695.9 Meas 2 to HARM retro 13137.2 12637.2 The two range measurements also differ in that the gauge block measurement is almost vertical, along the telescope X (or optical) axis, while the HARM measurement is at an angle to the X axis. This angle can be calculated since we know the the Y, Z co-ordinates of the fold mirror (under the gauge block) and the HARM retro, and the distance along the hypotenuse between the two. The blocks are just inside the edge of the hole in the CAP, which has a diameter of 17.38". Allowing for block placement, the radius is about 9", or (approx) 228.6 mm. Taking the Y and Z components for the two blocks used we get: Y Pos(mm) Z Pos(mm) --------- --------- Block #1 161.6 161.6 Block #2 -161.6 -161.6 The retro on the HARM is at Y=-1.0" . The position of the HARM during the two measurements is as follows: HARM Center (approx co-incident with HRC-I center) = -50360 steps Meas 1 TT step count = -65360 setps, delta = 15000 steps ~= 1.5" in +Z or Z = +38mm. For meas 2, step count was -22360, delta = 28,000 steps or about 2.8" in -Z, approx 70 mm in -Z. So the retro co-ordinates and deltas are: Y Pos(mm) Z Pos(mm) Retro Block Delta Retro Block Delta ----- ----- ----- ----- ----- ----- Meas #1 -14.3 161.6 175.9 38.0 161.6 123.6 Meas #2 -14.3 -161.6 -147.3 -70.0 -161.6 -91.6 Meas# del_Y del_Z del_R Range(mm) Angle(deg) X Dist(mm) ----- ----- ----- ----- --------- ---------- ---------- 1 175.9 123.6 215.0 12638.0 0.974773 12636.2 2 -147.3 -91.6 173.5 12637.2 0.786656 12636.0 where del_R = SQRT(del_Y**2 + del_Z**2) Angle = arcsin(del_R/range) X Dist = Range * cos(angle) Finally, to calculate the distance from CAP datum A to the focus from the cross check measurements we must make the following adjustments to the (X-distance - HARM retro) value: 1) Subtract the block thickness (0.5" or 12.7 mm) since the gauge block extends in the +X direction from CAP datum-A and makes the (X-dist - block dist) too large. 2) Subtract the effective retro depth of 17.1mm (supplied by Scott Texter based on calibration in lab) since it makes the (X-dist - block dist) too large. 3) Add the HARM surface dist to Focal plane diatance of 169.2 mm to the (X-dist - block dist) since the focus is furthur away from the rangefinder than the HARM surface. The 169.2 mm number (6.661") was obtained from Randy Moore by Scott Texter and verified by John Pollizotti and Gerry Austin. (The 7.339" number in Mike Duncan's spreadsheet is the diatance from the SIM I/F plane to the HARM surface, not the HARM to focal plane surface. This number was not used, however). Distance(mm) Meas#1 Meas#2 -------------------------- ------- ------- X Dist 12636.2 12636.0 (-) Block Range(corrected) -2693.7 -2695.9 -------------------------- ------- ------- X Dist - Block Range 9942.5 9940.1 (-) Block thickness -12.7 -12.7 (-) Retro effective depth -17.1 -17.1 (+) HARM to Focal Plane Dist 169.2 169.2 ---------------------------- ------- ------- Cross-check Datum-A to focus: 10081.9 10079.5 Kodak Datum-A to focus: 10079.5 10079.5 ---------------------------- ------- ------- Difference: 2.4 0.0 The two measurements agree to 2.4 mm for #1 and 0.0 mm for #2. We have two other measurements which can be compared. Scott Texter made measurements at XRCF using the HRC at best focus and calculated that the distance from the +X gauge block surface and HARM surface should be 9943.1 mm. This is a bit longer than either of our two measurements (9942.5mm, 9940.1mm). After witnessing the measurement process at the Observatory level I think that the accuracy of the overall process is only a couple of mm, due primarily to the difficulty in aligning the NET2 to the gauge block surface, a difficult process. The measurements taken show agreement within what I believe to be the measurement error, and we should easily find best focus on-orbit, well within the adjustment capability of the SIM focus mechanism. The most probable variance between the initial focus position (our estimate of the BEST focus) and the actual on-orbit best focus will be on the order of one mm. This is based both on XRCF focus measurements and FACT test results as compared with optical alignment results. -- =============================================================================== William A. Podgorski Harvard-Smithsonian Center for Astrophysics, Smithsonian Astrophysical Observatory