The Flower and Cook Observatory (FCO) of the Uni-versity of Pennsylvania (Penn) housed a 0.72 m (28 inch) Fecker Cassegrain reflector in a traditional observatory dome (Koch 2010, Fig. 1). The late Robert H. Koch (1929-2010) was a Professor in Penn’s Department of Astronomy and Astrophysics, and an avid user of the reflector. The authors of this paper worked with Koch at FCO for many years in the roles of an electro-optician and observatory caretaker (RJM) and graduate student and later colleague (BDH). Many times Koch expressed to the authors his interest in obtaining a much larger research instrument to reduce photon shot noise in program object measures and noted that the FCO observatory dome was oversized and could accommodate a 1.0 m or larger aperture reflec-tor. As a result, Koch was interested in medium-aperture mirrors for a larger terrestrial telescope and for research programs, such as balloon flight experiments, where the optical tube assembly and mirror weight must be mini-mized. Koch (2010) and Holenstein et al. (2010a,b) de-scribe some of Koch’s work on medium-aperture mirrors.

This report summarizes those previously described ef-forts and adds details on Koch’s other projects related to lightweight medium-aperture mirrors.

Professor Peter Waddell from the University of Strathclyde, Glasgow, UK visited the Laboratory for Re-search on the Structure of Matter at Penn in 1991 to give a lecture on the history of television. Waddell carried with him a small 0.20 m (8 inch) aperture mirror with a polyester Mylar reflecting surface that he demonstrated

to Koch, and other Department of Astronomy and Astro-physics members, could be varied in focal length by suck-ing on a small tube. Manly (1991) provides an overview of Waddell’s mirrors. Koch approached the department chairman, Professor Kenneth Lande, and funds were allo-cated to build some prototypes, initially for balloon flight observations where weight is at a premium.

Between 1991 and 1996, mirror cells were fabricated from sizes 0.18 m (7 inches) to 1.77 m (70 inches), and experiments were conducted to characterize the mirror figure and stability. Several people participated in the fabrication and testing of the mirrors: Robert Hee served as the professional machinist, the late Samuel Seeleman and Koch as amateur ones and opticians, Richard Mitch-ell as the hardware and electronics specialist, and several graduate students as technicians. Three pneumatic imag-ing systems were designed, built, and tested, and a few department demos and talks were presented. Koch vis-ited Waddell’s lab in the UK in 1994 and studied his 0.61 m (24 inch) pneumatic mirror that had been kept under a substantial pressure difference for more than a year. Koch’s departmental efforts on the mirrors continued past his retirement in 1996 until about 2000.

Modern pictures of early Penn pneumatic mirror pro-totypes appear in Figs. 2 and 3. A thin (0.2 to 7.2 mil) aluminized reflective Mylar membrane was clamped be-tween machined surfaces. The entire assembly was then rotated or pressed down onto a machined Plexiglas ring to pre-tension the membrane and eliminate wrinkles. A variable amount of air was removed to produce the de-sired focal ratio (f/0.5 to f/10). The aluminum and Plexi-glas cells would hold a vacuum for months and never leaked any appreciable air even with thin films.

An interesting application of medium-aperture mir-rors Koch and Lande explored in the mid-1990’s was to make a lightweight telescope assembly suitable for visible and near infrared observing of point photometric sources from a high-altitude balloon. Funding for an initial proj-ect was secured by Lande from a National Aeronautics and Space Administration student grant. Koch designed the optics, Lande and Mitchell designed the electronic payload, and six undergraduate students from Penn par-ticipated with department staff in the construction, test-ing, and launch. The electro-optics used differential glob-al positioning system to calculate the payload rotation direction and rate (spin). Two oppositely oriented charge-coupled device (CCD) cameras were used at prime focus in kinetics mode to clock pixels at a rate to counteract the spin. The pneumatic mirrors and related control system were as yet not producing an image of sufficient optical quality so a 0.5 m (20 inch) Zerodur mirror was purchased for the initial flight. Figs. 4 and 5 show details of the tele-scope payload.

In 1995, the team and parts were relocated to the Bal-loon Flight Facility of the Goddard Space Flight Center on Wallops Island in Virginia for final assembly and testing. The launch day was clear with little wind for the morn-ing flight. The balloon and payload were laid out on the tarmac with the prevailing wind direction taken into ac-count. However, as the balloon was being filled with gas, the wind changed direction and the payload was shaken. The rest of the flight was a great success, but the comput-er hard drive had apparently failed at an altitude of only about 20 m and little data was recorded. Additional bal-loon flights were conducted before funding ran out.

A videotape of Koch describing the balloon flight

equipment and observing goals exists and was played, in part, by Lande at the Stars, Companions, and their Inter-actions: A Memorial to Robert H. Koch Conference held at Villanova University in August, 2011.

In 2006, Bruce Holenstein discussed with Koch his in-terest in building a pair of large-aperture light buckets for a revival of stellar intensity interferometry (Genet & Holenstein 2010). Some plans and designs were pursued, and work on the pneumatic mirror project was re-en-gaged in 2008 when Holenstein, Mitchell, and Koch start-ed a new collaboration, and brought Kevin Iott onboard as designer and machinist and Dylan Holenstein on as a lab technician.

The program was hosted at Holenstein’s employer, Gravic, Inc. in Malvern, PA, and in 2009 the fabrication of a fully operational 1.07 m (42 inch) membrane mirror telescope was completed. The mirror assembly and truss weighed under 45 Kg (100 lbs.). Figs. 6-8 detail the con-struction and deployment of the telescope.

A limit of 1 to 2 arc minute point spread function (PSF) Full width at half maximum (FWHM) was reached with the 1.07 m telescope. Analytical work described in Holen-stein et al. (2010b) details the expected photometric per-formance of the aberrated image from the telescope. The signal-to-noise-ratio (SNR) was deemed to be insufficient

except for very bright stars and so remedial work was con-ducted on the pneumatic mirror cell design (Fig. 9).

Further residual wavefront aberrations remediation resulted in the team investigating membrane tension-ing techniques with different cell designs that counteract the approximately 4% difference in the Young’s modulus between the transverse and longitudinal manufacturing directions of Mylar. Koch and the authors also designed and started building an active secondary mirror using piezoelectric actuators for conjugating the primary mir-ror wavefront aberrations. Koch outlined some photo-metric algorithms and strategies for processing highly astigmatic images. Additionally, lightweight mirror sub-strates from the Alt-Az Initiative Group were tested. One such alternate substrate mirror is made from fused plate glass over foamed glass by OTF Designs LLC (Oak Hill, FL, USA). An image of one experimental mirror cell is shown in Fig. 10.

Traditional metrics for assessing mirror optical quality failed for the highly aberrated lightweight mirrors built and tested by the team. For example, two pneumatic mir-rors with the same peak-to-valley and root mean square (RMS) surface height measures would give vastly differ-ent optical performance as characterized in Fig. 11.

A Bath interferometer was built and used to make in-terferograms of sub-regions of the mirrors under test. These interferograms were stitched and analyzed to produce Zernike wavefront coefficients characterizing the aberrations. An analytical technique was devised to estimate from the Zernike coefficients the rms gradient norm of the wavefront and the rms surface slope, and from that the FWHM of the PSF. The interferograms of the pneumatic mirrors indicated that astigmatism domi-nates. The 1.07 m pneumatic mirror had astigmatism as the dominant aberration at a magnitude of between 50 and 100 waves P-V of visible light. This corresponds to a 100 to 200 microradian rms gradient norm at the surface. The best foamed glass mirrors tested had an rms gradient norm at the surface ten or more times better, at between 5 to 10 microradians. These latter mirrors produce a PSF spot size several arc seconds in diameter, a value which is near the lowland atmospheric seeing level.

A major risk to a large mirror exists whenever it needs to be recoated. Most large professional observatories install vacuum deposition equipment in their domes so that the primary mirror does not have to travel a signifi-cant distance to get a new coat of reflective metal. Koch and Holenstein met in 2010 with Sagar Venkateswaran, the president of Peacock Labs in Philadelphia, PA to dis-cuss an alternate solution for medium-aperture mirrors. Peacock manufactures a product for cold silvering of hard surfaces. They also manufacture a clear overcoating agent called Permalac for preventing tarnish on silvered surfaces. Fig. 12 shows some small test mirrors undergo-ing tests. It was found that the Strehl ratios of the mirrors were preserved for the silvered mirrors, but the silvered and Permalac overcoated mirrors only remained useful for astronomy purposes when the protective overcoat was applied “extra thin” (under 5 ㎛). For the overcoated mirrors, the rms wavefront error measured with a Bath interferometer increased by 0.12(± 0.02) waves at 550 nm.

The optical quality of some medium-aperture mirrors developed is sufficient for high cadence aperture pho-tometry, spectroscopy, high time resolution astronomy, and other observing projects that only require very mod-est spot sizes or on-axis light gathering. The best mirrors that Koch and the authors tested encircled 94% of the en-ergy (i.e., two times the FWHM) with PSF spots that are viable for photometry of 13th magnitude and brighter objects so that scintillation, and not program object shot nor the background (sky), is the limiting noise source for fast cadence and high time resolution photometry.

The team designed a 1.6 m cast aluminum cell alt-az telescope for prime focus use (Fig. 13). The estimated construction cost of USD $65k is relatively low vs. a tra-ditional substrate mirror telescope because the mass of the optical tube assembly is only 230 Kg (500 lbs). The de-sign is flexible enough to accommodate alternate mirror substrates such as those being researched by the Alt-Az Initiative Group including spun epoxy, foamed glass, and slumped meniscus plate glass. An array of these light-weight telescopes will offer a substantial reduction in scintillation noise and improvement in the overall SNR of program object measures compared with a monolithic mirror of the same effective aperture. In fact, Genet & Ho-

lenstein (2010) show that a seven-element array of 1.5 m aperture telescopes all observing the same 10th magni-tude program object at a dark, 3,000 m elevation observ-ing site, and combining output can produce 60% of the photometric performance of a traditional single-mirror 8 m aperture telescope. This feat can be done for under 1% of the cost of the 8 m telescope.

Robert H. Koch was a polymath who was able to oper-ate on many levels. He had a keen interest in all aspects of astronomy including building novel telescopes to solve specific observing goals. During the last two decades of his life he was able to lead and participate in numerous lightweight medium-aperture telescope building proj-ects. Work started by Koch on mirrors continues today at Gravic, Inc. with plans to eventually build an East coast

USA array of seven or more 0.75 to 1.5 m telescopes with high speed Photometrics Cascade 512B emCCD cameras at prime focus. Koch mentored numerous undergradu-ate and graduate students along the way and passed on his love of astronomy to many. Fig. 14. is characteristic of how many of us remember him.