NASA - James Webb Space Telescope - Discussion Thread 2

[LouScheffer]

Webb has been tested, retested, back tested, front tested, side tested, top tested and bottom tested so many times, I'm confident all will go well.  Just hope the scope will be working well into the bonus years.

There will likely not many "bonus" years due to propellant usage for station keeping.

There are folks seriously considering re-fueling comsats.  Given Webb's cost, refueling would make a lot of sense, if it's working well otherwise.

Was this considered at all in the design for Webb?   It would not have had to have been much - perhaps some optical fiducials, a docking plate, and an exposed refueling port.  Even if these do not match the eventual standards for comsat refueling, they could make the job of custom re-fueler much easier.

[deadman1204]

Webb has been tested, retested, back tested, front tested, side tested, top tested and bottom tested so many times, I'm confident all will go well.  Just hope the scope will be working well into the bonus years.

There will likely not many "bonus" years due to propellant usage for station keeping.

There are folks seriously considering re-fueling comsats.  Given Webb's cost, refueling would make a lot of sense, if it's working well otherwise.

Was this considered at all in the design for Webb?   It would not have had to have been much - perhaps some optical fiducials, a docking plate, and an exposed refueling port.  Even if these do not match the eventual standards for comsat refueling, they could make the job of custom re-fueler much easier.

Ignoring the fact that webb won't be in earth orbit, it's unrefuelable. The maneuvering thrusters of an approaching craft would destroy the sunshield.

[Eric Hedman]

It would not have had to have been much

I love it when I read stuff like this on the internet.

You guys have no idea what goes into designing a spacecraft.

I've discovered over the years that most people have no clue about the details and challenges in any industry they have never been directly involved in.

[LouScheffer]

It would not have had to have been much

I love it when I read stuff like this on the internet.

You guys have no idea what goes into designing a spacecraft.

With all due respect, I stand by my statement.

Would adding a port potentially capable of refueling be trivial?  Of course not - nothing about designing a spacecraft is trivial.  But compared to the challenges of folding beryllium cryogenic mirrors, or microshutter IR detectors, it's comparatively simple.

Comsats are already adding these ports.  Comsats are typically worth a few hundred million dollars at the beginning of their life (or at least that's what they are insured for).  Refueling would only be used to extend their life, at which point they would already be a decade or so old.  So assuming the comsat designers are not idiots, the cost to add the port, including all the engineering required, must be less than the value of a depreciated comsat.  The Webb is worth roughly 100 times as much, and a lot harder to replace, so the effort to add a potential refueling port cannot be a major addition to the engineering required.

Looking at this another way, the Webb is roughly a $10 billion dollar project.  How much might you reasonably spend on a change that could double the life of the observatory?  Perhaps 1% of the total?  For Webb, that's $100 million.  And if was that expensive, the comsat folks would not be considering it.

Finally, from an engineering viewpoint, adding a potential refueling port is not conceptually difficult.  It's a replacement for the fill/drain valve, which must already exist.  From a technical maturity point of view, refueling valves already exist and are space-qualified.  From a programmatic point of view, Northrop Grumman is an investor in a satellite refueling company, so the access to the details of the technology is there.

Overall, I still don't think adding potential for servicing (as did Hubble when they added a capture ring) would have been a major effort.  I can also understand the opposing view, that given all the challenges and brand new technology of Webb, there was no interest in anything that was not absolutely required.  But if in 10 years, Webb's working well but running out of gas, that decision is going to look pretty short-sighted.

[armchairfan]

Was [refueling] considered at all in the design for Webb?   It would not have had to have been much - perhaps some optical fiducials, a docking plate, and an exposed refueling port.  Even if these do not match the eventual standards for comsat refueling, they could make the job of custom re-fueler much easier.

Ignoring the fact that webb won't be in earth orbit, it's unrefuelable. The maneuvering thrusters of an approaching craft would destroy the sunshield.

A reasonable question from a senior member with (at the very least!) a demonstrated history of understanding basic engineering followed by a respectful, reasonable albeit negative reply. I happen to agree with Lou but I can understand arguments to the contrary -- especially given prevailing opinions when JWST was designed.

I love it when I read stuff like this on the internet.

You guys have no idea what goes into designing a spacecraft.

While I acknowledge that you're a space industry professional whose opinions I often find insightful, this post was unhelpful and arrogant. Please take it down a notch.

(rant)
While I'm not directly involved in the space industry, I have an engineering background and my company did various consulting gigs before we developed our own products. One of my take aways from consulting was as you get closer to "how the sausage is made" some things impress you while others don't. There were numerous reasons for the latter but a frequent one was complacency: "we've always done it that way and it's good enough". The problem was that good enough then had become not good enough now given the new competitive environment.

Certainly I can sympathize with your frustration over apparently clueless questions or suggestions from industry outsiders. Perhaps you can sympathize with others that are sometimes frustrated by what they consider overly conservative thinking or approaches in an industry that they perceive is (or should be) undergoing rapid innovation.
(/rant)

With all due respect, I stand by my statement.

Good for you!

[LouScheffer]

Ignoring the fact that webb won't be in earth orbit, it's unrefuelable. The maneuvering thrusters of an approaching craft would destroy the sunshield.

These are real problems, but not insurmountable. 

Refueling is already designed to work in GEO.  L2 is close enough that it is not considered deep space, so the same radio system can be used, though uplink power may need to be increased, and downlink will be slower.  Likewise it's close enough to Earth orbit so the solar arrays, sun sensor, and so on should be OK.  Overall, operating a GEO refueler at L2 should not be a huge task.

The subshield is indeed delicate.  But at L2, the gravity gradients are very low, and the refueler can approach very slowly.  Then the final alignment and deceleration could be done with ion thrusters or other very low force devices.  The sun facing sunshade layer is about 50 microns thick and  hence will mass about 50 grams per square meter.  The film itself can hold up under earth gravity (about 500 millinewtons per square meter) and so should be OK with the ion thruster forces, typically 100 millinewton or so.  The supporting structure for the sunshield is strong enough to survive when the station-keeping thrusters are used (about 35 newtons) and should also be OK with milli-newton forces.  Using noble gas ion thrusters should also reduce worries about contamination (although Webb itself uses chemical thrusters on the sunward side of the sunshade, so this may not be a limitation).

So approaching Webb from the sunshield side seems possible without damaging it.

[LouScheffer]

it's comparatively simple.

Back around 2009 or so there was a proposal to add a grapple fixture to JWST--the kind of thing that they used for payloads to be gripped with the shuttle's arm. I sat in on a meeting where this was discussed. Adding something like that late in the design would have rippled throughout the entire spacecraft. It affected the structural loads, the thermal loads, the control and guidance programming, everything, for adding something about the size of a doorknob to the spacecraft.

I completely agree that adding anything after the detailed design is complete is not practical.  That's why, when a science instrument is not ready, it is replaced by an inert block of the same mass, drawing the same amount of power.

On the other hand, designing it in from the start is not hard, as shown by the adaptation by the comsat community.

So the answer to the original question seems pretty clear.  Refueling was NOT considered in any way during the design of the Webb.  It WAS considered later (at the meeting you sat in on), but at that point it would have required so much redesign that it was rejected.

This answers my original question.

[Jim]

The subshield is indeed delicate.  But at L2, the gravity gradients are very low, and the refueler can approach very slowly.  Then the final alignment and deceleration could be done with ion thrusters or other very low force devices.  The sun facing sunshade layer is about 50 microns thick and  hence will mass about 50 grams per square meter.  The film itself can hold up under earth gravity (about 500 millinewtons per square meter) and so should be OK with the ion thruster forces, typically 100 millinewton or so.  The supporting structure for the sunshield is strong enough to survive when the station-keeping thrusters are used (about 35 newtons) and should also be OK with milli-newton forces.  Using noble gas ion thrusters should also reduce worries about contamination (although Webb itself uses chemical thrusters on the sunward side of the sunshade, so this may not be a limitation).

So approaching Webb from the sunshield side seems possible without damaging it.

No, throwing ions at the spacecraft is not a good idea.   
And ions thrusters are going to be insufficient for translation control and even attitude control of the approaching vehicle.  They can't react fast enough considering the mass of the servicing spacecraft.

[LouScheffer]

No, throwing ions at the spacecraft is not a good idea.   
And ions thrusters are going to be insufficient for translation control and even attitude control of the approaching vehicle.  They can't react fast enough considering the mass of the servicing spacecraft.

Attitude control is not a problem.  The combination of reaction wheels and ion thrusters (to keep the wheels unloaded) works well.  This has been proven on all-electric comsats.

Translation in the two axes perpendicular to approach can use normal chemical thrusters - Webb itself does this. This leaves slowing down as you approach the Webb as the main problem.  The Shuttle docked at about 3 cm/sec.  A 100 milli-newton thruster, acting on a 1000 kg spacecraft, can add or subtract this much velocity in about 300 seconds (5 minutes).  During this time the spacecraft will travel about 5 meters.  So a slow approach, especially given the low gravity gradient of L2, seems feasible.

As as far as the impact of the ions on the sunshade, this clearly needs to be looked at.  The sunshade is covered with a layer of silicon, and the ion thrusters are generating inert gas, so it's likely not a problem.  Here is a video of an ion thruster acting on aluminum foil at close range.  No damage is apparent, but obviously a more formal study would be needed.

[Jim]

1.  Attitude control is not a problem.  The combination of reaction wheels and ion thrusters (to keep the wheels unloaded) works well.  This has been proven on all-electric comsats.

2.  Translation in the two axes perpendicular to approach can use normal chemical thrusters - Webb itself does this. This leaves slowing down as you approach the Webb as the main problem.  The Shuttle docked at about 3 cm/sec.  A 100 milli-newton thruster, acting on a 1000 kg spacecraft, can add or subtract this much velocity in about 300 seconds (5 minutes).  During this time the spacecraft will travel about 5 meters.  So a slow approach, especially given the low gravity gradient of L2, seems feasible.

3. As as far as the impact of the ions on the sunshade, this clearly needs to be looked at.  The sunshade is covered with a layer of silicon, and the ion thrusters are generating inert gas, so it's likely not a problem.  Here is a video of an ion thruster acting on aluminum foil at close range.  No damage is apparent, but obviously a more formal study would be needed.

1.  It is a problem.  This is not a stationary comsat, but spacecraft doing prox ops.  It has to be able to rotate and react quicker.  Reaction wheels are not used, thrusters are.
3

2.  No, the plumes from the other two axis will still affect the sunshield. 
The spacecraft is going to weight much more than 1000kg. 
 cm/sec was final velocity and not approach.  The shuttle was braking.
Again, too slow to react.  JWST isn't stationary.  It is orbiting L2 and reacting to solar pressure

3.  The issue I was raising was charge and not pressure.

[rsnellenberger]

Translation in the two axes perpendicular to approach can use normal chemical thrusters - Webb itself does this. This leaves slowing down as you approach the Webb as the main problem.  The Shuttle docked at about 3 cm/sec.  A 100 milli-newton thruster, acting on a 1000 kg spacecraft, can add or subtract this much velocity in about 300 seconds (5 minutes).  During this time the spacecraft will travel about 5 meters.  So a slow approach, especially given the low gravity gradient of L2, seems feasible.

As as far as the impact of the ions on the sunshade, this clearly needs to be looked at.  The sunshade is covered with a layer of silicon, and the ion thrusters are generating inert gas, so it's likely not a problem.  Here is a video of an ion thruster acting on aluminum foil at close range.  No damage is apparent, but obviously a more formal study would be needed.

Even assuming that a tanker can approach and dock on the shade-side without physically damaging it — how do you propose to protect that great, gaudy mirror and the just-as-important secondary mirror? Hubble had a cover to protect its optical components, but Webb’s mirrors are exposed even in the launch configuration - and the tanker brings along a cloud of propulsion byproducts (hypergol, argon/krypton ions, or a combination) when it arrives regardless of its approach trajectory or strategy.

[LouScheffer]

Translation in the two axes perpendicular to approach can use normal chemical thrusters - Webb itself does this. This leaves slowing down as you approach the Webb as the main problem.  The Shuttle docked at about 3 cm/sec.  A 100 milli-newton thruster, acting on a 1000 kg spacecraft, can add or subtract this much velocity in about 300 seconds (5 minutes).  During this time the spacecraft will travel about 5 meters.  So a slow approach, especially given the low gravity gradient of L2, seems feasible.

Even assuming that a tanker can approach and dock on the shade-side without physically damaging it — how do you propose to protect that great, gaudy mirror and the just-as-important secondary mirror? Hubble had a cover to protect its optical components, but Webb’s mirrors are exposed even in the launch configuration - and the tanker brings along a cloud of propulsion byproducts (hypergol, argon/krypton ions, or a combination) when it arrives regardless of its approach trajectory or strategy.

Well, the center of the sun-shield on the sun side is where the Webb's own hypergol and monoprop thrusters are, and it uses them even after the mirror is deployed.  Since L2 has a good vacuum, reaction products will travel in straight lines. So if you approach from the direction of the sun, the mirror should be protected from reaction products for the same reason it's protected from sunlight.

[DaveS]

- and the tanker brings along a cloud of propulsion byproducts (hypergol, argon/krypton ions, or a combination) when it arrives regardless of its approach trajectory or strategy.

Exactly. After each service mission HST needed several months for the new components to offgas and all the crud from the shuttle to dissipate before any serious observation campaigns could begin. And that was with the HST Aperture Door (AD) closed the entire time.

[Paul451]

Refuelling is the alternative to end-of-mission. Even a diminished performance at some wavelengths due to vented gases depositing on the mirror is better than zero performance. Hand-wringing over possible damage is pointless.

[LouScheffer]

1.  Attitude control is not a problem.  The combination of reaction wheels and ion thrusters (to keep the wheels unloaded) works well.  This has been proven on all-electric comsats.

2.  Translation in the two axes perpendicular to approach can use normal chemical thrusters - Webb itself does this. This leaves slowing down as you approach the Webb as the main problem.  The Shuttle docked at about 3 cm/sec.  A 100 milli-newton thruster, acting on a 1000 kg spacecraft, can add or subtract this much velocity in about 300 seconds (5 minutes).  During this time the spacecraft will travel about 5 meters.  So a slow approach, especially given the low gravity gradient of L2, seems feasible.

3. As as far as the impact of the ions on the sunshade, this clearly needs to be looked at.  The sunshade is covered with a layer of silicon, and the ion thrusters are generating inert gas, so it's likely not a problem.  Here is a video of an ion thruster acting on aluminum foil at close range.  No damage is apparent, but obviously a more formal study would be needed.

[Answers re-ordered to put like problems together]
1.  It is a problem.  This is not a stationary comsat, but spacecraft doing prox ops.  It has to be able to rotate and react quicker.  Reaction wheels are not used, thrusters are.  The spacecraft is going to weight much more than 1000kg.  3 cm/sec was final velocity and not approach.  The shuttle was braking.

Again, too slow to react.  JWST isn't stationary.  It is orbiting L2 and reacting to solar pressure

2.  No, the plumes from the other two axis will still affect the sunshield.

3.  The issue I was raising was charge and not pressure.

1:  JWST will be essentially stationary.  It orbits L2 with an orbit that takes 6 months and is larger than the moon's orbit.  Complications from orbital dynamics show up when the final sequence takes more than a fraction of an orbit.  Since Webb's orbit is almost 3000 times longer than the Shuttle's, a refueler could approach much more slowly, up to about 3000 times more slowly if needed.  Reaction wheels and ion thrusters, even acting on a mass of 2-3 thousand kg (the mass of existing servicing satellites MEV-1 and MEV-2), are more than fast enough at this scale.

Solar pressure is not enough to cause a problem here.  The Webb sunshade is about 200 m^2.  The solar power incident upon it is about 3x10^5 watts, resulting in a force of about 2 milli-newtons (assuming it's perfectly reflective).  An ion thruster is about 100 milli-newtons, and so can easily compensate for any JWST motion caused by solar pressure.

2:  The Webb thrusters do exactly this, so the sunshade must already withstand it.

3:  In an ion engine, the outgoing ions are neutralized, otherwise any spacecraft with ion thrusters would accumulate an enormous charge.  And in any event, the Webb's sun-facing shield is coated with a conductive layer of doped silicon, precisely to prevent any charge from accumulating on the membrane.

[deadman1204]

I think we should stop this conversation. Its obvious that some people are considering most likely scenarios vs pie in the sky idealism. We're just talking past him.

[lrk]

3:  In an ion engine, the outgoing ions are neutralized, otherwise any spacecraft with ion thrusters would accumulate an enormous charge.  And in any event, the Webb's sun-facing shield is coated with a conductive layer of doped silicon, precisely to prevent any charge from accumulating on the membrane.

That's not how an ion thruster works, there is a separate electron gun to dissipate the accumulated negative charge, but that doesn't thoroughly neutralize all the outgoing ions. 

I agree that this is pretty off topic - mods, could this discussion on refueling Webb be moved to a seperate thread? 

[redliox]

Earlier there was some speculation on an earlier launch, but I presume the date of December 18th still scheduled as is even with the prior Ariane 5 launch going off early?

[Jim]

1: Reaction wheels and ion thrusters, even acting on a mass of 2-3 thousand kg (the mass of existing servicing satellites MEV-1 and MEV-2), are more than fast enough at this scale.

2:  The Webb thrusters do exactly this, so the sunshade must already withstand it.

3:  In an ion engine, the outgoing ions are neutralized, otherwise any spacecraft with ion thrusters would accumulate an enormous charge.  And in any event, the Webb's sun-facing shield is coated with a conductive layer of doped silicon, precisely to prevent any charge from accumulating on the membrane.

1.  no, not true. JWST IS orbiting L2 and is not stationary. 

2.  Wrong.  Webb's thrusters are not pointed at the sunshield.

3.  No, the exhaust is not completely neutralized.

[Jim]

Earlier there was some speculation on an earlier launch, but I presume the date of December 18th still scheduled as is even with the prior Ariane 5 launch going off early?

That has no bearing on the spacecraft schedule once on site.