Radio Astronomy Antennas by the Thousands Roger Schultz - PowerPoint PPT Presentation

Radio Astronomy Antennas by the Thousands Roger Schultz 650-964-5899 schultz_assoc@pipeline.com July 21, 2004 SKA2004, Pentiction B.C. p1 of 23

Cost Effective Radio Telescope Development • Existing cost expectations, established by ATA et. al. - 1,000$US/meter 2 • Achieve this by extending the existing evolution of steerable microwave antennas. • Careful, thorough and realistic selection from myriad innovations. • High quantity required allows development of mass production • Utilize effective analytical techniques and prototyping • Utilize new and more efficient industrial processes • Goal: Lowest cost concepts to meet SKA performance • Conceptual examples follow that relate to current USSKA strawman antenna development July 21, 2004 SKA2004, Pentiction B.C. p2 of 23

Reflector surface evolution -the old complex way • 15.5 meter S band antenna • Small man handleable panels, ~80 pie panels • Panel size drives very complex backstructure to support each panel corner • Structural frame work point (node) required at each panel corner • This complexity is unnecessarily costly July 21, 2004 SKA2004, Pentiction B.C. p3 of 23

Pie panel / reflector backstructure interaction • Dark members hold panel surface shape • White members near panels form front members of reflector back frame structure • Panels are dead weight which make no contribution to reflector back structure strength or rigidity • Connection, panel to frame, is a flexure perpendicular to reflective surface designed to relieve differential thermal expansion of steel back structure and aluminum panel July 21, 2004 SKA2004, Pentiction B.C. p4 of 23

Shop reflector surface sweep • 14 meter S/K band reflector surface adjusted in shop. • Structure is dowel pinned, disassembled, shipped, reassembled at site (next slide) • Early attempt to cut costs • Poor surface accuracy results at site • Too many pinned Photo shows outer end of sweep boom. Center pivot is out interfaces, surface had of picture to right. Dial indicators on templates below to be reset in field boom structure show reflector surface error as boom • COSTLY rotates. Dial indicators zeroed to template profile. July 21, 2004 SKA2004, Pentiction B.C. p5 of 23

Field Assembly and Alignment 14 meter S&K band antenna assembled and aligned on ground July 21, 2004 SKA2004, Pentiction B.C. p6 of 23

Panel assembly up on antenna mount • Again, panel small enough to be handled by workmen in light winds • Again, structural work point (node) required at each panel corner • Reflector surface adjusted by extending or retracting clip at each panel corner • Adjustment performed by workmen climbing through back structure. • COSTLY July 21, 2004 SKA2004, Pentiction B.C. p7 of 23

Pie panel reflector alignment in air • Rough adjustment by radial tape from theodolite mount and elevation angle (lower picture) • Targets placed on reflective surface at a fixed radius by chordal linkages resting on rough set surface (left picture) • Panel clip height fine adjusted to theodolite angle • Technique produced about 1 mm RMS accuracy for 4-6 GHz • New instruments now producing much better accuracy but process still takes too much Upper: chordal linkages guide time and too much drilling of target mounting holes. money Right: Schultz instructs height adjustment to setting angle. 30 meter • COSTLY domsat antenna, circa 1970’s July 21, 2004 SKA2004, Pentiction B.C. p8 of 23

Early Pie Panel Tooling • Flat sheet metal shaped only by tooling bows at panel bracing • Bonding (adhesive) fills gap between flat hat or Z section flanges • Rivets used • COSTLY F ig u r e 1 reflective sheet is h e ld d o w n o n i gure 1 riv e t F July 21, 2004 SKA2004, Pentiction B.C. p9 of 23

Higher accuracy “full face” pie panel tooling •Whole sweep boom (upper picture) •Female CNC cut template sweeps out surface of “full face” pie panel fabrication fixture (lower picture) •Full face tool controls panel surface more accurately than bows shown in previous slide •COSTLY July 21, 2004 SKA2004, Pentiction B.C. p10 of 23

Current panel production • “Bonded” pie panel Bonded photos courtesy – Panel back structure “bonded” Antedo, Inc. i.e. glued to reflective aluminum sheet while on shaped tooling – reflective sheet not curved until pulled to tooling – savings from deletion of drilling many holes and installing rivets – Widespread successful use confirms concept – .01 mm RMS surface accuracy • Stretch formed Panels – Aluminum sheet stretched over shaped tool to generate reflective surface – Stretch formed ribs bonded to reflective surface • BOTH COSTLY • MUST EVOLVE TO LESS COSTLY CONCEPTS Photo courtesy Patriot Antenna Systems July 21, 2004 SKA2004, Pentiction B.C. p11 of 23

First Hydroformed Allen Telescope Array dish • Dave DeBoer, SETI, examines 1st ATA 6 meter dish • John Anderson, Anderson Mfg, behind Dave • Process now making 0.2 mm RMS surface accuracy on 6 meter July 21, 2004 SKA2004, Pentiction B.C. p12 of 23

6 meter ATA antenna at Hat Creek • ATA 6 meter offset Gregorian • Extremely low weight / LOW COST antennas, $1,500/sq. meter • “Throated” reflector, gimbal mechanism within reflector structure • Elevation axis close to dish center • Less wind torque allows use of less costly mount to achieve pointing July 21, 2004 SKA2004, Pentiction B.C. p13 of 23

Challenge : 12 meter surface accuracy • 12 meter hydroforming development – Springback and thinning in the hydroforming process are repeatable and can be modeled with nonlinear FEA. Figures show calculated spring back for 12 meter dish • figures courtesy Dimitri Antos, JPL – Springback and thinning calculations are used to adjust 0.02 mold to produce desired dish shape 0.018 – process proprietary, Anderson 0.016 0.014 Mfg. Springback (m) 0.012 – Ohio State U will continue spring 0.01 back & thinning simulations 0.008 – all previous size increases at 0.006 0.004 Anderson (and there have been 0.002 many) have been successful! 0 0 1 2 3 4 5 6 Radial Distance (m) July 21, 2004 SKA2004, Pentiction B.C. p14 of 23

Splitting hydroformed dish C flange • 12 meter too large to truck with permit to remote sites • Splitting of dish with precision splice • 2 concepts being studied – C flanges (proprietary) – strip over zigzag cut Zigzag (proprietary) cut July 21, 2004 SKA2004, Pentiction B.C. p15 of 23

USSKA 12-16 meter Strawman antenna • 12 m reflector splits into 2-pieces – red members define split plane; center hub and red members are split and flanged for shipping (see next slide) – reflective 12 meter shell IS VERY STIFF and forms the front half of the “backstructure” • shell replaces all front radials and hoops and eliminates most diagonals • decreases number of radial trusses – sub reflector, tripod legs and 2 meter skirt are removed for shipping – “Throated” for low cost / performance • Mass produced components – large dish shells, castings, stampings – CNC machinings – mature industrial components (turntable bearings, gear boxes, etc.) • Minimized site assembly and alignment July 21, 2004 SKA2004, Pentiction B.C. p16 of 23

Split 12 meter reflector down the road • 6 meter wide load with permit • Passes under – 4.6 meter (15 ft) high underpass with standard trailer – 4.0 meter (13 ft) using “low boy” trailer July 21, 2004 SKA2004, Pentiction B.C. p17 of 23

12-16 meter Convertible optics • 16 meter prime focus operation • 12 meter Gregorian 1.2 to 34 down to 100 MHz GHz plus flipped feed, wide field, 0.1 to1.5 GHz operations July 21, 2004 SKA2004, Pentiction B.C. p18 of 23

“Throated” vs. Conventional antennas • Shell reflector with throat 30% to 50% lighter than for typical pie panel reflector July 21, 2004 SKA2004, Pentiction B.C. p19 of 23

Performance/ cost comparisons • Axis wind and gravity drive torque – Conventional antenna azimuth peak wind torque 50% higher than for “Throated” antenna • Thrust term moment arm much shorter – Conventional antenna elevation peak torque 100% higher than for “Throated” antenna • Conventional antenna has the elevation axis unbalance all to one side at a long moment arm • “Throated” antenna has an evenly “split” unbalance to both sides • Thrust term moment arm much shorter • Wind “gust gain” lower in proportion to steady wind comparisons above – Lowest Resonant Frequency (LRF) requirements go down in proportion to about the square root of the decrease in wind gust gain – Reduces LRF requirement to meet absolute pointing requirement • Antenna costs significantly reduced by reductions of reflector weight, drive torque and LRF requirements . July 21, 2004 SKA2004, Pentiction B.C. p20 of 23