NASA calls off attempts to fix Kepler space telescope

http://www.reuters.com/article/2013/08/15/us-space-kepler-idUSBRE97E16B20130815

by anigbrowl • 13 years ago

38 44 13 years ago

reuters.com

kfury• 13 years ago

The Kepler analysis team has only gone through the first two years worth of data.

There are still two more years worth of exoplanets for us.

andrewflnr2• 13 years ago

In the future, will we be able to fix it if we can get a person up there?

InclinedPlane4• 13 years ago

It will be vastly cheaper to launch a new one.

batbomb• 13 years ago

It's wayyyyyyy too high in orbit to get up there, past the moon. Not even remotely worth it.

They are doing the same thing with the Webb.

andrewflnr2• 13 years ago

Oh, wow, wikipedia says its actually in a heliocentric orbit. Yeah, not happening.

jamesaguilar• 13 years ago

Whenever I see an article like this and something at nasa has broken, it seems like they always are soliciting a new project. It makes me wonder, why not just build the exact same thing again (with the broken parts fixed, of course)?

jerrya• 13 years ago

Building a new something that was designed, built, tested and launched quite a while ago is difficult because a lot of knowledge of the processes have been lost, along with tooling, supplies, etc.

However, I do wonder why they build 1 Kepler space telescope and not 2, or 3, and then store the already built ones until they are wanted.

mturmon• 13 years ago

Given what the previous instrument taught us, the science questions change, so the instruments you need also change. For instance, it surprised everyone when it turned out that most stars are more variable than the Sun is, and this extra variability has complicated the Kepler data analysis.

Additionally, you can incorporate lessons learned (from an engineering perspective, not science) into the next instrument. Some things work well and are preserved, and others don't.

Here's a link to some tech awards relevant to future exoplanet instruments: http://exep.jpl.nasa.gov/technology/

sigstoat6• 13 years ago

kepler wasn't built that long ago, as these things go. and ball aerospace which designed and built it is already designing and building an almost identical spacecraft for the b612 foundation. so of all the spacecraft you might want to duplicate, kepler would be a really easy one at the moment.

ak217• 13 years ago

As sigstoat said, Kepler is a fairly recent spacecraft. And yeah, it's interesting that they don't treat spacecraft like this as small-batch production runs. The incremental cost of producing 5 Kepler replicas and keeping equipment and processes documented at the time of initial production seems very small compared to doing the same project all over again a decade later. Even if they have to do some major changes, I imagine having a ready to go spacecraft helps.

On the flipside, I guess the arguments for componentization apply. If off-the-shelf components are improving all the time, integrating them into a full spacecraft should be getting cheaper and easier over time.

What I really wonder about is, given how troublesome and relatively lightweight gyros are, why didn't they have more spares?

MichaelMoser1232• 13 years ago

Can they possibly fix it? They don't have a Shuttle to take the thing in so it can be brought back and fixed. Whatever they say about the Shuttle, it was a powerful vehicle.

Can a robot fix the thing?

sigstoat6• 13 years ago

i don't think he was talking about fixing kepler. just building a new, less-shitty one.

and however "powerful" the shuttle was, it couldn't get to something 3 light-minutes away (which was kepler's distance, the last time i looked)

Moto7451• 13 years ago

Kepler actually orbits the Sun, not the Earth[1]. Its orbit is slower than the Earth's and as a result is getting farther away from the Earth. Assuming it needs an EVA to fix, as Hubble did, it's impossible to fix as we lack a manned spacecraft capable of getting there.

[1]http://en.wikipedia.org/wiki/Kepler_(spacecraft)#Spacecraft_...

InclinedPlane4• 13 years ago

Kepler is millions of miles away from Earth, even with a Saturn V and an Apollo spacecraft we couldn't fix the thing and come back.

Also, it was never economical to fix Hubble with the Shuttle. It was just fortuitous that the Shuttle program was desperate for things to do and ended up subsidizing the launch costs for the Hubble servicing missions. In terms of the ultimate bottom line the US spent nearly $8 billion to fix a $2.5 billion telescope. Granted, I think that even that expenditure was worthwhile but from a strict cost accounting it would have made more sense to simply build and launch additional telescopes.

InclinedPlane4• 13 years ago

Budgeting and launch costs are probably the biggest reasons.

Kepler, for example, has already used up its budget to build, launch, and operate the telescope for several years. Even if free replacement hardware was available there may not be enough money to launch it. Also, the way most spacecraft are built is as one off prototypes, so building another copy, even if all the spare parts were available, is a very non-trivial endeavor. Even if it didn't take as long as building the original it would still be fairly expensive just to go through all of the necessary testing and integration steps.

The end result of all of this is that it rarely makes sense to spend the time and money to refly a spacecraft and instead it usually makes more sense to simply forge ahead and seek funding for a follow-on mission of a new design.

Here's a few case studies. The inaugural launch of the Ariane 5 rocket failed, which resulted in the loss of 4 "Cluster" spacecraft which were designed to study Earth's magnetosphere (a loss of about $370 million in hardware). However, ESA decided to fund replacements and they built 4 new satellites, making use of flight spares and extra hardware to save on costs, and then launched the new satellites on 2 separate Soyuz-Fregat flights. However, it still cost almost as much money to launch Cluster II as for the originals, so this was a rather heroic effort on ESA's part.

In 1999 the Wide Field Infrared Explorer (WIRE) spacecraft was launched, shortly after launch there was a malfunction that caused a loss of the main science instrument on the orbital telescope. The mission was not replaced. However, the loss of WIRE caused other IR survey mission proposals to receive greater interest, which led to the funding of the "Next Generation Sky Survey" Medium-class Explorer mission. This infrared full-sky survey mission was launched in 2009 (as WISE, the Wide-field Infrared Survey Explorer) and achieved all of the science goals of the original WIRE mission.

In 1993 all contact was lost with the Mars Observer spacecraft right at the moment of planned orbital insertion. This loss led to a re-evaluation of the unmanned Mars exploration program at NASA and eventually led to around 10 new missions to Mars over the following 2 decades, most of them successful.

jbri• 13 years ago

In general, I'd expect that you learn a lot in the process of building and flying a spacecraft, even if it ultimately fails to meet its objective.

If you're going to have to launch a new spacecraft anyway, why not take some of those lessons into account when you build the next one?

caf• 13 years ago

The impression I had from the article is that they're soliciting new projects that make use of the broken Kepler. It's not completely cactus, it just can't fulfil its original mission anymore.

This strikes me as an excellent thing to do - if they can find a project that can fund the ongoing costs of Kepler, then it makes sense to reuse it.

ngoldbaum• 13 years ago

No one else has brought up that another Kepler wouldn't necessarily be as impactful as the original. True, it would probably discover another couple thousand planets, but Kepler has already done that once. What's another factor of two in the number of known planets?

Kepler looks off the plane of the Galaxy at stars that are typically several kiloparsecs away. In other words, we are never going to visit these worlds. Now that Kepler has revealed that planets are common, it will be much more interesting to find nearby planets that we have a better chance of characterizing, seding probes to, or even visiting one day.

This is what TESS will do: http://en.wikipedia.org/wiki/Transiting_Exoplanet_Survey_Sat...

InclinedPlane4• 13 years ago

The good news is that there's still a lot of juice left in the Kepler data set. Only about half the data has been analyzed, and a lot of good stuff will come out of the rest. Also, it takes 3 observations of a transit for planetary confirmation, but not all 3 observations need to be made by Kepler. With only 2 transits having been observed it'll be easy to extrapolate future potential transit dates. Ground based observatories can then do targeted observations of the most promising candidates and fill in that data, essentially moving a lot of observations from the "candidate" planet category into confirmed territory for a comparatively minimal effort.

moocowduckquack3• 13 years ago

Bye Kepler. You were awesome. http://kepler.nasa.gov/Mission/discoveries/

ChuckMcM• 13 years ago

Ok, so trolling around I've not found specs for reaction wheels. Lots of descriptions of what they are, how they work, etc but no specifics on building one. Given that this type of product is so critical to space operations (The ISS has replaced several, so has Hubble, its like the oil-filter of space or something) Seems like either a more durable item or some sort of cartridge system where you can have like 20 spares ready to go, eject one and slot in a new one or something.

lutusp6• 13 years ago

I'm a former NASA design engineer. Reaction wheels are a real weak point in spacecraft design. They basically allow one to accurately point an observing spacecraft, but without using propellant.

If you use propellant, you eventually run out and the mission is over. With electric-motor reaction wheels (a minimum of three), you can keep them going with solar energy. So in principle, they're a big improvement over the alternatives.

The problem is that they're constantly in motion -- constantly. If you turn a spacecraft and bring it to a new stable position, that new position must be maintained by constant reaction wheel motion. So they're essentially running all the time, there's constant friction, and their parts and bearings eventually wear out.

Seems like either a more durable item or some sort of cartridge system where you can have like 20 spares ready to go, eject one and slot in a new one or something.

You're overlooking something. To turn a spacecraft like Hubble or Kepler in a reasonable time, you must either have a massive reaction wheel, or you need to make it spin fast (it's all about momentum). Typical reaction wheels represent a compromise between mass and rotational speed. If you want to reduce frictional wear, you make them bigger, more massive, which increases the mission cost. If you need to reduce the size and mass of the spacecraft, you make the wheels spin faster, and they wear out sooner.

So solving the problem by adding more reaction wheels is a non-starter -- they're too physically large and massive to just add a lot of spares.

keenerd• 13 years ago

a minimum of three

Correct me if I am wrong, but don't most systems use six? It is a whole lot easier to use three pairs of wheels. Pairs can be spun up from a dead stop and in the event of reduced power can be used to turn in either direction by braking action alone.

Or does "a reaction wheel" imply two counter-rotating flywheels by definition?

For an alternative, I've heard some good things about laser ablative thrusters. Still fairly young, but they are powered by electricity and have no moving parts. They do need reaction mass, but unlike traditional engines the mass is a solid block and can't boil off, corrode, leak, freeze or whatever problems affect liquid systems.

rurounijones• 13 years ago

I am so far out of my depth here that it is not funny but..

Why not use wheels that use magnets instead of bearings so that there is no friction?

moocowduckquack3• 13 years ago

Could you do something like spin ferrofluid inside a hollow donut with some coils round it to make a reaction wheel that would take ages to wear out? No bearings and the only friction would be the fluid against the inside of the donut.

mapt2• 13 years ago

Laminated rolls of carbon fiber spinning at high RPM on magnetic bearings have been the state of the hype for terrestrial flywheel power applications for the last 13 years: http://www.wired.com/wired/archive/8.05/flywheel.html

Magnetic bearings have been used in fans, and I'm familiar with a bearing-analogous application in Inductrack maglev trains as well.

Has this type of tech penetrated into aerospace?

elteto2• 13 years ago

Why do the wheels need to be in constant motion? Is there a constant external torque acting on the spacecraft?

ISL• 13 years ago

A huge thank-you to all the teams involved with Kepler. It's a very fine experiment that's taught us a ton. Hit the lifetime goal too.

Great job!

NASA calls off attempts to fix Kepler space telescope

Related Stories