Future Optical Space Telescopes - i.e. Hubble 2.0 ect

[KelvinZero]

I would prefer to hear how to build a 100m telescope or 1000 meter telescope, whatever lets us blow past this SLS discussion entirely.

Using SLS to launch a large telescope would be better suited to the HLV forum IMO, but I personally accept Jim's arguments that launching a larger monolithic mirror on a larger rocket is a dead end.

[Burninate]

I would prefer to hear how to build a 100m telescope or 1000 meter telescope, whatever lets us blow past this SLS discussion entirely.

Using SLS to launch a large telescope would be better suited to the HLV forum IMO, but I personally accept Jim's arguments that launching a larger monolithic mirror on a larger rocket is a dead end.

Hubble 2.0 isn't a 100 meter space telescope.  It's a 6.5 meter space telescope (JWST), or a 2.4 meter space telescope (WFIRST-AFTA or shorter wavelength counterpart) with a wider field.  Ideally, Hubble 3.0 would be a fleet of ~4m wide angle telescopes, but I don't discount the possibility that it will be a ~15-25m class analog of JWST.

Building a 100m telescope would involve tiling a bunch of active-optics mirror cells on an in-space-constructed truss system, and it would require a highly streamlined approach to fab, launch, and assemble for less than 100 billion dollars - and I would expect it to take ~20 years or so from inception to completion at the pace we've been working at.  Building a 1000m optical telescope is probably beyond the scope of realistic investment, unless a number of things drop in cost by an order of magnitude.

There's also diffractive telescope concepts that make some sense up to absurd scales using thin film optics, but I'd venture to guess that a 1000m Aragoscope has nowhere near the light-gathering capabilities of a 1000m segmented-glass-mirror telescope (or a 10m segmented-glass-mirror telescope, for that matter).

Another idea is to go beyond the CCD - an optical MKID on the horizon might give phase, R=100 spectral output, and effectively zero-noise (photon-counting) sensitivity over a limited band.  Mix MKIDs with diffraction gratings and you can do other interesting things like tomographic integral field spectrometers in crowded fields.

[Burninate]

If you go far beyond the optical, things look rather different.  We apparently already possess ~20m aperture commercial radio dish antennas, and might even have ~100m aperture ones (probably only to low frequency though) based on rumors of the Orion / Mentor NRO satellites.  Free-flying big-dish radio telescopes with large constellation size are useful for precision interferometry, and there are also proposals to put them in a Lunar Farside Radio Quiet Zone.

We can do quite a bit from the ground with radio.  As you descend from 1mm down to 10um, though, most of that band is not feasible to do through Earth's atmosphere, and detector technologies lag way behind the very effective techniques we have for ~200nm-2000nm (CCD/CMOS variants), the moderately effective thermal infrared bolometers from 2um to 10um, and all the radio technology from 1mm to 1km (and out to 1000km and beyond with progressively lower gain).  Submillimeter orbital interferometer constellations are probably too big to be practicably synchronized, but erecting a ~100m-1km submillimeter megatelescope from an in-orbit-constructed-truss backing is entirely possible - a lot easier than the optical equivalent due to the 1-3 orders of magnitude lower surface accuracy required.

[KelvinZero]

Um.. you understand that 100 and 1000 were just numbers pulled out of my bottom right? The point was whatever is big enough that we can get past that silly conversation of launching on SLS.

You made an interesting assertion about trusses. The earlier link on the glitter telescope specifically mentions no trusses. Will it work? I dunno, but this is advanced topics so this is the right place for it.

[Paul451]

I wonder if the mass-economics of spacecraft design has shifted with the larger launchers on the horizon, whether cheap small-diameter ones like FH or expensive large-diameter ones like SLS. There's no advantage in launching light. That's doubly so with cheap multiple launches on FH. So you need to completely rethink how you design the spacecraft, always ignoring mass and instead focusing (indeed fixating) solely on cost.

So "spending" 5+ tonnes of payload mass on dumb-bulk shielding for off-the-shelf electronics is preferable to spending $100's of millions on specialised rad-hardened electronics. Adding many redundant groups of cheaper systems is better than fewer more-reliable (but more expensive) space-rated systems.

--

But as for design: Judging by JWST, designing around foldable, self-deployable segmented mirrors is expensive. So perhaps a group of 2-3m monolithic mirrors in a simple framed interferometer, launched as separate connecting segments, assembled and tested at the ISS then SEP'd to the final orbit. Think three or more el cheapo versions of HSTs on a triangular or Y-arm truss.

That might get KelvinZero his hundred meter telescope.

[Jim]

The launch costs won't justify Five tons of shielding.  And assembly at the iss is a bad idea for telescopes

[Hanelyp]

And assembly at the iss is a bad idea for telescopes

I agree with the "at ISS" part.  But why is assembly from modules a bad idea for a larger space telescope?  Automated precision near field positioning for docking could be an enabling technology.

[Paul451]

The launch costs won't justify Five tons of shielding.

If a 50t-LEO FH launch costs $150m (with a NASA tax), then every $3m you save in development is worth 1 tonne of wasted payload mass. If adding 5 tonnes of bulk shielding saves more than $15m in development costs, it's worth it.

And assembly at the iss is a bad idea for telescopes

The only negative is orbital inclination, and that's a non-issue if you are using a SEP to reach HEO or ESL.

[Sohl]

Building a 100m telescope would involve tiling a bunch of active-optics mirror cells on an in-space-constructed truss system, and it would require a highly streamlined approach to fab, launch, and assemble for less than 100 billion dollars ...

For example, the "tiles" could be hexagonal modules that can free-fly from a release point to the assembly point, orient and dock with the growing assembly, perhaps using magnetic induction forces (like mlinder experimented with, IIRC) to align before locking into the assembly mechanically.  The interlocking mechanism could be actuated to allow the body of each tile to adjust angles slightly to correct thermal expansion/contraction or other distortion effects, and the primary mirror surface could also be actuated separately to fine-tune the mirror figure to achieve the proper focus. 

If each hexagonal tile was about 1 meter across, several could be launched in a light/medium booster, or as secondary payloads on larger missions.  The main mirror assembly could grow over time as launchers and/or  payload space allow.  Maybe $1 million each, including launch?  Tile 10 thousand of them together for a 100m class mirror, for about $10 billion.  Maybe with mass production line and launcher re-usability, the cost per mirror could be closer to $50K to $100K?  Of course, I'm neglecting the equipment needed to perform main mirror pointing and station-keeping (I like the geosynchronous orbit idea though) and the sensor array module at the focal plane.

edit: wording

[A_M_Swallow]

If we are worried about the exhaust gasses from incoming rockets contaminating the mirror's surface they could dock with a spacestation several miles away. A cable could then connect the flying control room to the partially built telescope. Ordinary electric motors can carry the parts to the telescope for assembly. A small spacestation can now be purchased/long term leased for less than a billion dollars.

[hop]

The only negative is orbital inclination, and that's a non-issue if you are using a SEP to reach HEO or ESL.

Contamination would likely be a concern at ISS. SEP through the Van Allen belts is not really a great thing for the detectors either.

Realistically though, the astronomy community is going to be very lucky to get a telescope that maxes out the payload of a single FH launch (let alone SLS), so on-orbit assembly is unlikely to be relevant.

But as for design: Judging by JWST, designing around foldable, self-deployable segmented mirrors is expensive.

It's reasonable to expect that designing these things will get easier with more experience.

[Coastal Ron]

I would prefer to hear how to build a 100m telescope or 1000 meter telescope, whatever lets us blow past this SLS discussion entirely.

I'll expand on what I suggested earlier about assembling new telescopes in space.

I anticipated that there would be a lot of concern about operating in LEO, and though I can't speak to the challenges of sending telescope modules through the Van Allen Belt, what I think would be doable is to final assemble the telescopes in the region of EML.

If a space station is eventually placed at an EML point in the future, besides being a scientific outpost it could also host temporary construction crews.  If we're ever going to expand humanity out into space we need to find dual use solutions for assets we keep pushing out to the boundaries of where we can operate, so such an arrangement makes sense.

Hopefully this also means we can commoditize the construction of telescopes and other types of observatories, since once you lift the weight and size restrictions, prices for components should drop to some degree.  So instead of spending $8B for building a unique telescope like the JWST, which can not be serviced or upgraded, we could use a generic platform that could easily have components swapped out or upgraded.

Unfortunately I think this is not doable near term, since there doesn't seem to be any serious interest in an EML station, but when we eventually get to that point it could provide a lot of benefits.

[Jim]

The only negative is orbital inclination, and that's a non-issue if you are using a SEP to reach HEO or ESL.

There are many negatives WRT the ISS for telescope assembly.  Among them, it is a dirty environment for optics.  It is basically a nonstarter.

[Jim]

[
If a 50t-LEO FH launch costs $150m (with a NASA tax), then every $3m you save in development is worth 1 tonne of wasted payload mass. If adding 5 tonnes of bulk shielding saves more than $15m in development costs, it's worth it.

LEO is not a destination for telescopes.  Try GEO or L2 or L1 and your numbers don't work for those.

[AncientU]

3.8m is largest that can fit in Delta-IV 5m fairing.  4ish meters is largest practical monolithic mirror size today due to processing facilities, etc., though that could be remedied (some day, but not soon). Even monolithic mirror telescopes are rather 'fluffy' payloads, so generally will be volume limited; mirror areal density of even glass mirrors is around 100kg/m² or less.  DIV-H and FH are or soon will be able to handle this size observatory, but the cost differential favors FH and may shift even more in that direction as D-IV is phased out.  No one is currently building a 4m, though one is being analyzed.

GSO is my choice for survey science (extremely high data rate required) and L-2 seems optimum for point and stare telescopes.  Extreme cooled optics (<50K) and formation flying (such as star shade/telescope pairs) are restricted to Sun-Earth L-2 or drift-away orbits with the latter limited to shorter duration/lower data rate missions.

[Stan Black]

Instead of a circular mirror, is it possible to have a spinning strip or spoke?

[Space Ghost 1962]

I would prefer to hear how to build a 100m telescope or 1000 meter telescope, whatever lets us blow past this SLS discussion entirely.

...
Unfortunately I think this is not doable near term, since there doesn't seem to be any serious interest in an EML station, but when we eventually get to that point it could provide a lot of benefits.

Would have agreed in past. Lately much more interest in EML 1/2. Both in terms of commercial, institutional, and exploration. Examples: Sower's "distributed launch" example of refueling deliveries, resurrecting DTAL, LockMart's Jupiter as arranging Exoliners to pass cargo, numerous international interests in EML 1/2 as a "jumping off" point for post ISS HSF "exploration".

The global space "toolkit" works for EML 1/2, thus international partners see it as a next step. Stands in contrast to the dopey SLS EM-1/2 missions on the board. Good enough for a reliable science product return.

Has interesting possibilities. Could history roll out like:
  * EML 1/2 Gateway
  * Commercial LEO station with nation "tenants in common"
  * ARM(s)
  * Commercial NG space telescope(s) with national "prepay" / "long term" "rent" "service" "upgrade"
  * Mars Phobos mission/station
  * Commercial lunar/asteroid missions
  * Commercial logistics to support national exploration
  * Mars surface expeditions

IP's break the path first. Commercial follows on the path, reducing cost/risk looking for long term means to expansion beyond exploration. Reliable, scalable, economic science product is first return.

Earth telescope observatories are run by consortium's. Why not space based telescopes too?

[Asteroza]

Don't forget hammerhead payload shroud configurations may enable larger (unusual?) monolithic types as well.

Though if some on orbit assembly (as in non-self-deploying requiring full assembly, or externally assisted deploying) is allowed, launching a stack of mirror hexagons, on the face on of it, seems much easier to do.

[catdlr]

A possible future use for the two scopes given to NASA from the defense department?

NASA Mulls Spy Agency's Telescopes for Dark-Energy Mission

http://m.space.com/29192-nasa-spy-satellite-telescopes-space-mission.html

[Burninate]

Don't forget hammerhead payload shroud configurations may enable larger (unusual?) monolithic types as well.

Though if some on orbit assembly (as in non-self-deploying requiring full assembly, or externally assisted deploying) is allowed, launching a stack of mirror hexagons, on the face on of it, seems much easier to do.

It would be a lot of development work to perfect the technique, but it seems workable eventually at the cost of even more active optics supplemental feedback loops.  The JWST folded approach is much easier to do, but it has hard limits on how large it can scale, relative to the payload fairing, and it's much more expensive (at n=1 anyway) than monolithic mirrors.  Building a telescope out of individual mirror cells and constructed trusses would be the next step up (or maybe two steps up) in complexity above that, but could grant you a mirror of almost arbitrary size.