Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.stsci.edu/institute/atlast/documents/stahl_8mDesign_SPIE2008.pdf Дата изменения: Wed Jan 14 16:17:56 2009 Дата индексирования: Thu Apr 8 13:53:58 2010 Кодировка: Поисковые слова: р п р п р п р п р п р п р п р п р п р п п р п п р п п р п п р п п р п п р п п р п п р п п р п

Design study of 8 meter monolithic mirror UV/optical space telescope
H. Philip Stahl NASA Marshall Space Flight Center, Huntsville, AL 35812
ABSTRACT
The planned Ares V launch vehicle with its 10 meter fair in g shroud and 55,000 kg capacity to th e Sun Earth L2 point enables en tir ely n ew classes of sp ace telescopes. NASA MSFC has conducted a pr eliminary study that demonstrates the feasibility of launching a 6 to 8 meter class monolithic primary mirror telescope to Sun-Ear th L2 using an Ares V. Specific techn ical areas studied in cluded optical d esign ; structur al d esign /an alysis including primary mirror support structure, sun shade and secondary mirror support structur e; thermal analysis; launch vehicle performance and trajectory; spacecraf t in clud ing structure, propulsion, GN &C, av ionics, power systems and reaction wheels; operations & servicing; mass and pow er budgets; and system cost. Keywords: Large Space Telescopes, UV /Optical Space Telescopes, Ares V Laun ch Veh icle, Astronomy

1. INTRODUCTI ON
An 8-meter class space telescope offer s the opportunity to answer some of the most compelling science qu estions. How did the pr esen t Univ erse come into existence and of what is it made? What ar e th e fundamental components that govern the formation of today 's galaxies? How does the So lar System work? What ar e the conditions for planet formation and the emergence of lif e? And maybe most importantly, are we alone? (Postman, 2008; Stah l, 2007) A recen t d esign study conducted at Marsh all Space Flight Center has shown that it is possib le to pack age a 6 to 8 meter class monolithic observatory in to a 10 meter Ar es V fairing (Figure 1); have it surviv e launch; and place it in to a halo orbit about the Sun-Earth L2 poin t. (Hopkins, 2007)

Figure 1 Ares V can launch 6 to 8 m eter class monolithic mirror telescope. (Im age courtesy of Jack Frassan ito & Associates and H arley Thronson)

Specific techn ical areas studied in cluded optical d esign ; structur al d esign /an alysis including primary mirror support structure, sun shade and secondary mirror support structur e; thermal analysis; launch vehicle performance and trajectory; spacecraf t in clud ing structure, propulsion, GN &C, av ionics, power systems and reaction wheels; operations & servicing; mass and pow er budgets; and system cost.


2. THE ARES-V LAUNCH CAPABI LITY ENABLES NEW DESIGN CONCEPTS
The study star ted w ith th e uniqu e capab ilities of the Ares V vehicle (Figure 2) and then considered how those capab ilities might en able en tir ely new mission arch itectures.

OD-2 ID-2 H-2

S hroud Outer D iameter Shroud Mass OD -1 ID -1 H -1 OD -2 ID -2 H -2 Total H eight V olume

10-m 7.8 10 8.8 9.7 5.6 4.4 7.5 17.2 8 60 55.8

mT m m m m m m m m3 mT

ID-1

H-1

OD-1

P ayload to L2TO

Figure 2 A res V B aseline Sh roud Dimen sions and P ayload Mass C apability. (Please note th at the Ares V is an evolving veh icle and these are prelimin ary v alu es and m ay not m atch the latest Ares V shroud dim ensions and weights.)

First, the baseline 10 meter fairing h as an 8.8 meter in ternal dynamic env elope diameter. This is sufficien t to acco mmodate an 8- meter class monolith ic circu lar pr imary mirror w ithout the need for segmen tation. A monolith ic mirror provides super ior scien ce return because, as compar ed to a segmented mirror, it has a more un iform, symmetr ic and stable Po int Spread Function. And, it avoids the risk of deploymen t and comp lex alignment and phasing control. The 10 meter shroud also allow s an 8-meter monolith ic mirror to be launched in a f ace up configuration which provides the most benign vibration and acoustic exposure. Looking fu rther into the future, th e 10 meter fairing also allows for even larger aper ture segmented designs. Concepts ar e under consider ation for 16 to 24 meter segmented telescop es. Second, the pay load mass of 55,800 kg to an L2 Transfer Or bit enab les an entirely new par adigm ­ design simp licity. Given the available extr a mass, designers can use mor e mature technolog ies and h igher design ru le saf ety f actors to eliminate comp lexity, to lower cost and to low er risk. By using higher design margins it is possib le to min imize the march ing ar my size which also reduces the management burden ­ ev ery $100M in component cost savings reduces total program cost from $300 M to $500M. The cost sav ings of eliminating mass constr ain t is d ifficu lt to quantify, bu t anecdo tal eviden ce suggests that early in a mass constrained mission, it may cost $100K of design effort to elimin ate 1 kg of mass and that on ce the design is mature, it can cost as much as $1 M to eliminate 1 kg of mass. These two unprecedented enabling capab ilities of the Ares V formed the b asis for the foundation al question of the MSFC design study. Is it possible to launch an 8-meter class space observatory using a convention al massive monolithic ground based telescope mirror? Instead of using lightweight (very expensive and high r isk) mirror technology, is it possible to use a conven tional massive (low cost and low risk) ground telescop e mirror? And the answ er is Y ES.


3. OBSERVATORY DESIGN
3.1 Design Concept Figure 3 shows th e MSFC design concept for an 8-meter monolithic primary mirror ultravio let/op tical sp ace observatory pack aged insid e th e Ares V 10-m fairing's dynamic env elope. Th e concept h as three main sub systems: telescope, support structure and spacecraf t. Th e telescope consists of an 8-meter primary mirror, secondary mirror and forward structure/baffle tub e. The sp acecr aft provides all normal spacecr aft functions (su ch as pr opulsion; guid ance, navigation and contro l; commun ication; etc.) and houses the science instruments. The support structur e supports the primary mirror. And, it carr ies the observatory mass (of the primary mirror, telescop e forward structure and spacecraf t) providing the in terf ace of this mass to the Ares V for laun ch. 3.2 Opt ica l D esign

Telescope & Baffle Tube

Support Structure Spacecraft & Science In struments Figure 3 MSFC 8-m eter observatory concep t in Ares V dynam ic envelope

The feasibility study consid ered two differ ent telescope optical systems. An F/15 Ritchey-Chr etiИn (RC) design (Figure 4a) was ex amined for its ex cellent on- and off-ax is image q uality, comp act size, and u ltr a-violet throughput. A lso RC designs are the optical system mainly used by today's large telescopes. Therefore, it might be possible to reuse ex isting scientific instrument designs. Unfortunately, th is optical d esign has only a r elatively n arrow 1-arc minute field of view (NFOV) that is d iffraction limited at 500 nm. One metho d to achieve the desired w ide f ield performance is to use a refractiv e corrector - although with a limited spectr al r ange ­ in the scien tif ic instrumen t suite. Ano ther approach to achieving mu lti-spectr al wid e field performance is to use a three mirror anastigmatic telescope w ith f ine steering mirror design (Figure 4b). Th is configuration has a w ide 100 arc- minute (8.4 by 12 arc minutes) field of view (WFOV) th at is diffraction limit at 500 nm. But, it also has lower u ltr a-vio let throughput because of its two additional reflections. A potential solu tion is to imp lement a dual pupil configuration where UV and NFOV instrumen ts operate at the Cassegrain focus and WFOV instrumen ts operate off-axis providing th eir own tertiary mirror. The study assu med that the optical coatings will be the same aluminum with MgF overco at used on Hubble to provide good spectral transmission from 120 nm to b eyond 1 micrometers.

Figure 4 Telescope Optical Path. The or iginal op tical configuration ( a) was used to size the telescope subsystems and develop th e mass budget. The revised op tical configuration (b) resu lts in an 80X larger f ield of view 3.3 Primary M irror For either telescop e design, th e mono lithic mirror w ill b e manufactured using ex isting ground based mirror technology. This approach has two specif ic advan tag es: technical maturity and cost risk. First, it has been demonstr ated that one can actually polish an 8 meter class ground based telescope mir ror to a surface figure of better th an 8 nm rms (Geyl, 1999) (which is close to th e desir ed 5 nm r ms surf ace figure for the 8 meter Terrestrial Planet Finder program) . This is important because as shown in Table 1, while Hubble's 2.4 meter 180 kg/m2 mirror was polished to 6.4 nm rms, th e AMSD program on ly achieved 20 nm r ms on its 1 .4 meter segment 18 kg/m2 mirror. The higher th e mirror's areal density (or in actu ality its specif ic stiffness), the easier it is to ach iev e a very good surface f igure. Second, the cost for an 8 meter ground mirror is $20 M to $40M or $0.4 M/m2 to $0.8M/m2 wh ile the cost of a 50 squ are meter space technology mirror will be $200 to $500M ($4M to $10M/m2). Wh ile th is architectural choice adds approx 20,000 kg to the mass of the payload , the estimated $200M to $500M savings in mirr or hardware costs tr anslates in to total program cost savings


of from $700M to $2B (engineering design, system integration & test, managemen t and fees/program r eserv es add to th e total cost of any program by a f actor of 2.5X to 3X of th e har dware costs).
Parameter Material Diam eter Area Temperature Surface Figure Areal Den sity Areal Cost Year Table 1. Comparison of Space and G round Mirro rs HS T Spitzer A MS D JWST Ground ULE Beryllium ULE & Be Be Various G lass 2.4 0.85 1.4 1.5 (6.5) 8.2 4.5 0.5 ~1 25 50 300 4 300/30 30 300 6.4 75 20/77 25 7.5 to 15 180 28 18 26 300 to 500 10 10 4 6 0.5 1984 1999 2005 2008 Various

m m2 K nm rm s kg/m2 $M/m2

The reason for both advan tages is that ground based mirrors are v ery massiv e and hence very stiff. Thus, th ey ar e mu ch easier to fabricate than space mirrors. H istor ically, sp ace mirrors are v ery low mass (Table 1) and thus not very stiff . They have larg e gravity sags and are difficult to h andle, mo unt and fixtur e. And, they ar e diff icu lt to fabricate to very high precision. Thus, they ar e exp ensiv e. Three ground based mirror technologies have b een considered . The sp are V LT (Very Large Telescope) mirror manufactured by Schott is an 8.2 meter diameter, 200 mm thick , Zerodur solid th in meniscus blank with a mass of 23,000 kg. If edg ed to 6.2 meters d iameter and 175 mm thick, it wou ld have a mass of 11,000 kg. The Un iversity of Arizona manufactures 6 to 8 meter class borosilicate mirr ors using their honey comb spin cast technique. Recently, Arizona manufacture two mirrors for the LBT (Large Binocular Telescope) that are 8.4 meter diameter, 900 mm thick and with a mass of 16,000 kg. Fin ally, the Subaru Telescope ULE th in meniscus 8.3 meter pr imary mirror (blank manufactured by Corning and f inished by Brashear) had a mass of 21 ,000 kg. 3.4 Structural Design A fundamen tal question of th e d esign study w as wh eth er an 8-meter class ground based telescope mirror could even survive launch . The Ares V launch environment was an alyzed by Table 2 Maximum Launch Loads of an A res V the MSFC Advanced Concepts Office u sing PO ST3D. Th e via POST3D Analysis maximum laun ch lo ads (Table 2) are similar to those for ex isting Maximum Launch Lo ad Ares V launch v ehicles. Please note, these lo ads ar e not con current. They Axial (Z ) 4 g's are the max imu m load experienced at so me time during laun ch. Lateral (Y ) 7 x 10-6 g's And, the lateral loads do not include wind load ing or vibratio n. Lateral (Z) (Z is down range direction) 6 x 10-4 g's A structural analysis deter mined th at 66 ax ial support po ints keep the str ess lev el on an 8.2 meter d iameter 175 mm th ick meniscus pr imary mirror below 1000 psi (Figure 5). Thus, the mirror will survive launch.

4 g lateral 467 psi

6 g axial 710 psi

Figure 5 An 8.2 m eter 175 mm thick m irror can su rvive launch lo ads. 66 axial supports k eep bulk stress b elow 1000 psi.


The observatory structure is divid ed betw een the forward and back stru cture. The forward structure is similar to that of the Hubble Sp ace Telescope. It provides the meter ing structure between th e primary and secondary mirrors and holds the stray light b affle tube. Because of fairing length limitations, the forward structure is split in to an upper and lower part. The low er structure is load carry ing. I t ho lds the secondary mirror assembly tr ipod structure. The upper part contains th e upper baffle tube and the cover doors. The upper part slides forward on orbit to the full length of the stray ligh t baffle. Th e cover doors open and close on-orbit as r equired. A secondary tripod structure ex tend ing from the primary mir ror was considered but determined to be unable to achieve the desired system stiffness levels for an u ltra stab le telescope. The back structure has mu ltiple functions. First, it supports the primary mirror with 66 axial supports. Second, th e forward structure is attached to th e back structure as is the sp acecr aft. A key design element of th e MSFC concept is that all observatory mass is carried through the Figure 6 Ob servatory back support structure to an interface ring which attach es via other supports to the Ar es V Support Structure launch vehicle. Th is design concept allows the use of a completely conventional spacecr aft, i.e. it does not n eed extr a mass because it does no t provide th e in terf ace betw een the observatory and the launch veh icle. Structural design and analysis was performed for th e sp acecraf t using standard NA SA guid elines. problems were id entified. The primary product of this effort was a mass budget for th e sp acecraf t 3.5 Thermal D esign Standard th ermal design and analysis was performed for 4 different so lar ang les: 0, 45, 90 and 120 degrees where 0 degrees is the observatory back facing the sun and 90 degrees is th e observatory broadsid e to the sun. It was modeled that th e science instruments produce 750 W of h eat and the avion ic systems produce another 850 W of heat. The analysis assu med that th e observ atory is wrapp ed with f ive 1 0 layer MLI b lankets and th at the spacecraft has 16 .0 m2 of thermal radiators. Ther mal grad ien ts w ere calculated for both the spacecraft and the 8 meter primary mirror. (Figure 7) Withou t an activ e th ermal manag emen t system, the primary mirror temperature v aries as a function of sun angle from 160 K to 300K with approximately a 1K v ariation at each temp erature.
Table 3 P rima ry Mirror Temperature Sun Angle Temperature 0 deg 200K 45 deg 190K 90 deg 160K 120 deg 300K
-70.51 -84.31

No technical

-71.31

-85.24

Sun = 0°
-109.7

45°
28.41

Therefore, an active thermal manag emen t via 14 heat p ipes is requ ired to hold the primary mirror temp erature at a constant 300K for all sun angles with less than 1K of ther mal gradient. On-going thermal analysis will d etermine ex actly how small of a thermal grad ien t can be ach ieved . This is important because long exposure observations (such as ex tra-so lar terrestr ial planet finding and characterization) r equire a v ery stab le observatory -111 wavefront. And, th e primary mirror surface figure 90° varies as a function of temperature based on the Figure 7 P rim ary mirror temperature (°C) substrate material coeff icien t of thermal expansion (CTE) valu e and uniformity . At 300K, Corning sp ecifies U LE glass to have a mean CTE value of 5C to 35 C and Scho tt specifies that Zerodur ceramic glass has a CTE value of 0 +/- 50 ppb/K. specifies that the CTE un iformity d istr ibution within a Zerod ur mirror blank is +/- 10 ppb/K.

24.36

120°

v s sun angle

0 +/- 30 ppb/K from Additionally, Schott


3.6 Spa cecraft The observ atory actu ally has two separ ate spacecraf t: a telescope bus which is part of the optical telescope elemen t (OTE) , and a replaceable spacecraf t/instrumen t bus (SIB) (Figure 8). The SI B houses science instruments and subsystems to commun icate with and contr ol the telescope. Each sp acecraf t produces its own power. Th e telescop e has 18 m2 of body mounted solar arrays around the light tube. Th e SIB h as 9 m2 of deployable solar array wings w ith pointing ability. The SIB pow er system in cludes 800W for primary mirror thermal contro l and 750W for science instruments. The OTE performs its own on-board health diagnostics and communication to th e SI B. The SIB provides the primary communication down-link.

Figure 8 Sp acecraft/Instrument Bus (SIB ). Top contain s scien ce in strum ents, Solar panels are on side.

The spacecraft propulsion system is sized to get th e observatory from roughly a 185 x 300,000 km parking orbit (energy, or C3, of -2.60 k m2/s2) in to a h alo orb it about the Sun-Ear th L2 point and perform all station keeping operations. The spacecraf t has a dual mod e hydrazine-NTP bi-prop / hydrazine mono-prop propulsion system w ith 5 years of propellan t and redundant thrusters. Th e propellant load is b ased on an estimated station keeping V expend iture for five years of 20 m/s, plus the V requir ed to place th e telescope on to the L2 tr ansfer trajectory. Propulsion during th e tr ip from the parking orbit to L2 is provid ed by hydrazin e-BTP b i-prop 125 lbf thrusters (Northrop). Station k eep ing at L2 is provided by hydrazine mono-prop RCS 20/5 lbf thrusters ( Aerojet). The telescope h as an independen t con trol system with mono-propellan t hydrazine using 350/100 psi blowdown Aerojet thrusters. The telescope propulsion system h as 30 kg of propellant for 30 year mission. Guidance Navig ation and Point Control is provided by the spacecraf t reaction wheels. A tr ade study was performed to obtain th e optimum science p erformance as a function of wheel torque and momentum storage sp ecifications. (Figure 9) Two performance parameters wer e analy zed. Th e number of hours that th e telescope can star e at a fix ed point in space (remain at an iner tial hold) befor e needing to p erform a mo mentu m dump due to solar rad iation pressure torque. And, how fast in minu tes the telescop e can perform a 60 degree slew. Th e analysis w as done for a sun angle of 90 degrees, which is the worst condition for solar rad iation pressure torque. A t any other sun angle, the available science time in creases. And it was assumed that momentum buildup occurs in only one ax is (y-ax is). Six wheel and four wheel conf igurations were analyzed along with the worst case single wh eel failure for each configuration. Each configuration was analy zed for three differ ent TELDIX Figure 9 S cience T ime vs Slew Tim e An aly sis for v ariou s R eaction Wheels reaction wheel versions (TorqueMomentum Storage). 3.7 Mass Budget The entire mass budget for the 6-meter obs space craf t, avionics, etc. is less th an 35,000 launch capability. The mass budget for an margin, of which the pr imary mirror is diameter/vo lume, and not the paylo ad mass, ervatory in cluding primary mirror, structur e, light b affle tub e, instrumen ts, kg (Table 4) ­ a 38% mass margin on the Ar es V's 55,600 kg Sun-Earth L2 8 meter observatory is approximately 45,000 kg, with almost a 20% mass th e largest con tributor. Th ese mass budgets clear ly show that p ayload is the limiting factor in the telescope size.


Table 4 Mass Budget for a 6 m eter Telescope OTE and Spacecraft/Instrum ent Bus
Total mass = OTE W / Bus + Spacecraft and Science Inst OTE W / Bus mass Primary mirror assembly Secondary mirror assembly Telescope enclosure Avionics Subsystems Power Subsystems Thermal Management System Structures Propulsion Propellant Docking station Mass (Kg) 33,849 25,6 177 6 3,6 1 3 1,0 9 1 5 7 0 5 8 9 1 1 4 1,00 9 0 1 0 3 1 1 7 6 0 0

Spacecraft and Science Instrument Science Instrument Package Avionics Subsystems Power Subsystems Thermal Management System Structures Propulsion Propellant Docking station Launch Adapter

6,230 1500 3 3 4 7 2 1,5 1,0 3 7 8 5 4 3 0 4 7 1 5 8 6 0

2,000

Please note that several elements of this mass budget are allocations, including the science instrumen t package, launch adapter and docking stations. All mass elements will be subject to ref inement as th e design matures.

4. EXTENDED MI SSI ON LI FE BY I N-SPACE SERVI CI NG
To extend the life of th e observatory beyond its in itial design life of 5 years to a target life of 30 years or more, th e science instrumen ts and as many subsystem componen ts as possible are d esigned to be rep laced at per iodic intervals. These are all in the SIB (Figure 8) which can b e replaced as a sing le unit ev ery 3 to 5 years u sing au tonomous rendezvous and dock ing (AR&D) technology (as d emonstrated on Orbital Express). Beyond the obvious technical advantages of upgrading detectors, electronics and computer s periodically, it is anticipated th at designing subsystems for 5 years of operation instead of 10 years will produce sufficient cost sav ings to fund the per iodic servicing missions. (This n eeds to be the subject of a d esign trad e.) Th e SI B d iameter is set at 4.5 meters such that these servicing missions can b e launched via a conven tion al EELV. Using this approach, the fir st 5 years of mission lif e co uld be dedicated to UV scien ce with a narrow FOV UV spectrometer and a WFOV UV imager. Th en, the next 5 years of mission life cou ld be d edicated to visib le science such as terrestr ial planet f inding with either an ex tern al o ccu lter or an in ternal coronagraph. Even tually, it might be possible to have two differ ent SIBs on station with the ab ility to sw itch between su ites of science instruments. At th e observatory end of life, as thermal con trol d egrades, the telescope can be allowed to cool to <200K for an infrared scien ce campaign. During the period of SI B exch ange, when the SI B is undocked from th e observ atory, the telescop e spacecraf t provides basic guid ance and n avigation for station k eeping . The telescope h as 18 m2 of body mounted so lar array around light tube, used for station keep ing, and batter ies for up to 0.5 hour of attitude control contingency . The telescope av ionics systems ar e 3-fault to leran t for a 30 year life. As prev iously discussed, the telescope has a mono-propellant b low-down thrust system. The telescop e also has a low gain antenn a for communicating w ith th e servicing spacecraft. All telescope health and status d ata is sent d irectly to the spacecr aft avionics system. Also , power for th e telescope thermal manag emen t sy stem is provided by the SIB. Thus, th ere is no activ e thermal contro l during spacecr aft exch ange.


The primary subsystems for pointing, co mmunications, power, guidan ce, propulsion, as well as the science instru men t package and fine guid ance sensor, are lo cated on the SIB. One notable ex cep tion is the thermal control for th e primary mirror, which must be placed on the telescope bus. The SIB avionics and power systems are 1-fault toleran t for 5 year lif e. Power is gen erated from two 9 m2 deployab le solar array wings with pointing ability. Batter ies are sized for 2 hours of power during midcourse and rendezvous operatio ns (when the power arrays are retracted). The SI B power system includes 800W for mirror thermal control and 750W for the telescop e instru ment package. The guidance an d navigation system includ es star trackers, sun sensors and in ertial measurement units. AR&D w ill be facilitated w ith a LIDAR long range system and an optical short range system. Computers handle all normal station keeping , maneuv ers, data man agement, and ground communications. And, th e communication systems consist of K a-band HGA for ground, and s-band for local communication and backup cap ability

5. COST DISCUSSION
The proposed 8 meter monolith ic telescope concep t seeks to disprove the o ld adage that th e primary predictor of mission cost is mass. It is the au thor's assertion th at th e Ares V pay load mass capab ility is a d isruptive technology th at creates a new parad igm - by trad ing mass for simplicity it is possible to build a telescope w ith lower cost and low er risk . By eliminating complexity, it should be possib le to d esign and b uild an 8-meter monolith ic telescope with 2X the co llecting area of the 6.5 meter JWST for less cost. Consid er for example the complexity differen ce between pack aging a 6.5 meter segmen ted primary mirror into a 4 .5 meter dynamic launch envelope as comp ared to the simp licity of packag ing an 8 meter monolith ic mirror into an 8.8 meter dynamic lau nch envelope. The curren t cost for the JWST telescop e and spacecraf t (excluding science instrumen ts and operation) is approximately $3B. The total cost for an 8-meter observatory (excluding scien ce in struments and operations is estimated to be $1B to $1.5 B. To illustrate this point further, consider th e telescope pr imary mirror and its support structure. Because of launch vehicle payload mass constraints, all prev ious space based telescopes have requir ed low areal density primary mirrors - the bigger the telescope, the low er its required areal den sity. For examp le, the Hubble primary mirror has an ar eal density of 180 kg/m2 for a to tal mass of 81 0 kg or 7.4% of Hubble' s total mass of 11,000 kg. By comparison, th e JWST primary mirror has an areal d ensity of 25 kg/m2 for a total mass of 625 kg or 9.6% of the total JWST observatory mass of 6,500 kg. The explanation is simple, a primary mirror mass is limited to about 10% of a space observatory and the to tal mass of th e observ atory is limited by th e launch v ehicle. And, b ecause space mirrors have low ar eal densities, th ey are d ifficu lt to manuf acture and thus expensive. Sp ace mirrors are inh erently less stiff than ground mirrors. Thus, th ey have larger grav ity sags; exhibit a fabrication effect called quilting; and are diff icu lt to handle, mount and fix ture. Because of this mass and stiffness difference, the cost of a sp ace telescop e mirror is typically 10X higher than for a ground telescope mirror (Table 1). The Ares V elimin ates th is constrain t. Th e 8-meter monolith ic concept proposes to use existing ground based mirror technology rather than the u ltr a light-w eight mirror technology required for a large space telescope via an EELV. While this arch itectural cho ice adds approx 20,000 kg to the mass of the paylo ad, it is estimated to save $200M to $500M in mirror hardware costs and $700M to $2B in total program costs. The precursor JWST mirror technology development program A MSD demonstr ated that 1.4 meter ligh t-weight b eryllium and glass mirrors both cost approx $4M p er squar e meter. Currently, th e total cost for the 6.5 meter JWST pr imary mirror is in ex cess of $140 M or close to $6 M/m2. Thus, a 50 square meter mirror will cost $200M to $300M. Fur th ermore, I t is likely that a UV /V isib le quality version of the JWST pr imary mirror wou ld be even mor e exp ensive, maybe $500M (using th e HST $10 M/m2 areal cost). By comparison, 8-meter class (50 square meter) UV/V isible quality ground based telescope mirrors typically cost $20M to $40M or $0.4M/m2 to $0.8M/m2 . Giv en that eng ineering design, system integration & test, man agement and fees/program reserves add to the total cost of any program by a factor of 2.5X to 3X of the hardwar e costs, a $200M savings in the cost of a primary mirror translates into a $700M to $800 M total program cost sav ings. Risk is also signif icantly lower for a ground based mirror simply because th ey have been demonstrated. Curren tly th ere are nin e 8-meter class monolithic telescopes in operation. Some of these mirrors have a surface f igure better th an 10 nm rms ­ close to the r equiremen t for the Terrestr ial Plan et Find er primary mirror. Similar cost savings and risk reductions are anticipated fo r the telescope structure (ATK private conversation). On JWST, the cost of the telescope structure is approximately 2/3rd engineer ing labor and 1/3rd fabrication & tooling. The primary driver for the engineering labor is th e need to design a very low mass structure at the limit of performance safety


factors. I t has been estimated th at if the mass could be cut substantially. For an 8 meter engineering labor will account for 1/3rd of th e meter telescope should not exceed th e total co telescope op erates at ambient temper atur e).

could be incr eased by sever al factors (3X to 5X) that th e eng ineering time class structur e with a mass allocation of 10,000 kg, it is estimated th at total cost with the b alance for fabrication. And, that the to tal cost for an 8 st of JWST (note also that JWST is cryogenic and the proposed UV/optical

6. CONCLUSI ON
The unprecedented mass and volume capab ilities of NASA 's planned Ares V cargo laun ch vehicle en able entire n ew mission concep ts. Its 10 meter fairing and ability to place 55,600 kg of payload into Sun-Earth L2 comp letely changes the parad igm for future space telescopes ­ simp licity. Simple high TRL technology offers lower cost and risk. The Ar es V capacities allow one to use mass to buy down performance, cost and schedule risk. Instead of expensiv e lightweight space mirrors, one can use low-cost low-risk proven ground based mirror technology. And, instead of expending excessiv e amounts of engineering design and analysis labo r, the Ares V payload cap acity allows for mission designs with larger than normal stru ctural safety margins and f ewer complex d eployment mechan isms. NASA Marsh all Space Flight Cen ter has conducted a preliminary design study wh ich indicates that it is feasib le to launch a 6 to 8 meter class monolithic pr imary mirror ultr aviolet/visib le observatory. An 8-meter class UV /optical space observatory w ith its v ery high angular reso lution, very h igh sensitiv ity, broad sp ectral coverage, and high p erformance stab ility offers the opportunity to answer some of the most compelling science qu estions. How d id the pr esen t Un iverse come in to ex istence and of what is it made? What are the fu ndamental co mponents th at govern th e formation of tod ay's galax ies? How does the So lar System work? What are the conditions for plan et formation and the emergence of lif e? And maybe most importantly, are we alone? Finally, there is no inherent reason that an 8-meter space telescope using robust design concepts should have only a 5 to 10 year mission life. In f act, th ere is no reason that the telescope migh t not last 20 to 30 years. This ex tended mission lif e can be obtain ed via per iodic robotic servicing of the spacecraf t and scien ce instruments using autonomous rendezvous and docking technology (as demonstrated on Orb ital Express).

ACKNOWLEDGEMENTS
The author would like to thank Randall Hopkins and Alan Ph ilips of the Marsh all Sp ace Fligh t Cen ter Advanced Concep ts Office who provid ed information, analysis, data an d valuab le guidan ce regard ing the Ares V cap abilities and prelimin ary shroud designs and John Hraba, Bill Arnold and Ken Pitalo of the MSFC Optics Office.

REFERENCES
Postman, Marc, et. al., "Science with an 8-meter to 16-meter optical/UV space telescope", SPIE Proc.7010, (2008)

Stahl, H. Philip, "Ares V l aunch capability enables futu re space teles copes", SPIE Pro c.6687, 66870L (2007)
Hopkins, Randall and H. Philip Stahl, "A large monolith ic telescope p laced at the second Sun-Ear th Lagrange point", AIAA Space 2007, AIAA-2007-6166, 2007 Geyl, R and M. Cayrel. "REO SC Contribu tion to VLT and Gemini", EU ROPTO Confer ence on Optical Fabr ication and Testing, Ber lin, Ger many, May 1999, SPIE Vol. 3739, (1999) Orbital Express Mission Updates, http://ww w.darpa.mil/orbitalexpress/mission_updates.html