NASA - James Webb Space Telescope - Discussion Thread 2

[etudiant]

A lovely piece, but I could weep.
No backup at all, on a complex product where the track record for similar deployments is iffy.

frankly, SpaceX's SS can't come soon enough, NASA science spends most of their scarce $$ on product engineering because of launch constraints.
Lift those and the same $$ can do 10x more and better science.

Most of the costs on the Webb had little to do with launch constraints.  Most of the costs were associated with building a telescope that has to be so precisely built and maintain its shape with the thermal environment it is operating in.  The cost of developing the cameras to operate in this environment has also been a major challenge.  Starship would not have changed that.  A telescope of that size and capability would have been incredibly expensive even if it didn't have to unfold on the way to its operational location.

Vehemently disagree with you. 
The vast bulk of the cost is paying for the marching army of non contributors that are part of the effort.
They need to get paid whether the project gets done or not.
The cost escalation of both Hubble and the JWST (or Sofia) are because of that.
Move fast and break things is good advice, even in science.

[deadman1204]

A lovely piece, but I could weep.
No backup at all, on a complex product where the track record for similar deployments is iffy.

frankly, SpaceX's SS can't come soon enough, NASA science spends most of their scarce $$ on product engineering because of launch constraints.
Lift those and the same $$ can do 10x more and better science.

What does spaceX have to do with any of this? NASA isn't even paying for the launch., so cost is irrelevant. Also, name basically ANY mission in the last 30 years that did have backups? There aren't any.

[Eric Hedman]

A lovely piece, but I could weep.
No backup at all, on a complex product where the track record for similar deployments is iffy.

frankly, SpaceX's SS can't come soon enough, NASA science spends most of their scarce $$ on product engineering because of launch constraints.
Lift those and the same $$ can do 10x more and better science.

Most of the costs on the Webb had little to do with launch constraints.  Most of the costs were associated with building a telescope that has to be so precisely built and maintain its shape with the thermal environment it is operating in.  The cost of developing the cameras to operate in this environment has also been a major challenge.  Starship would not have changed that.  A telescope of that size and capability would have been incredibly expensive even if it didn't have to unfold on the way to its operational location.

Vehemently disagree with you. 
The vast bulk of the cost is paying for the marching army of non contributors that are part of the effort.
They need to get paid whether the project gets done or not.
The cost escalation of both Hubble and the JWST (or Sofia) are because of that.
Move fast and break things is good advice, even in science.

Even if you are right on paying for an army of non-contributors, how would Starship have changed that on the Webb telescope?  They were still only going to build one.  The launch vehicle has had very little to do with the cost of it.

Before you disparage the people who worked on Webb, who exactly are the army of non-contributors?  Are the in NASA, Northrup Grumman?  And what were they twittering away their time on?  I'd like to hear about an example?

[Blackstar]

A lovely piece, but I could weep.
No backup at all, on a complex product where the track record for similar deployments is iffy.

frankly, SpaceX's SS can't come soon enough, NASA science spends most of their scarce $$ on product engineering because of launch constraints.
Lift those and the same $$ can do 10x more and better science.

Most of the costs on the Webb had little to do with launch constraints.  Most of the costs were associated with building a telescope that has to be so precisely built and maintain its shape with the thermal environment it is operating in.  The cost of developing the cameras to operate in this environment has also been a major challenge.  Starship would not have changed that.  A telescope of that size and capability would have been incredibly expensive even if it didn't have to unfold on the way to its operational location.

Vehemently disagree with you. 
The vast bulk of the cost is paying for the marching army of non contributors that are part of the effort.
They need to get paid whether the project gets done or not.
The cost escalation of both Hubble and the JWST (or Sofia) are because of that.
Move fast and break things is good advice, even in science.

Even if you are right on paying for an army of non-contributors, how would Starship have changed that on the Webb telescope?  They were still only going to build one.  The launch vehicle has had very little to do with the cost of it.

Before you disparage the people who worked on Webb, who exactly are the army of non-contributors?  Are the in NASA, Northrup Grumman?  And what were they twittering away their time on?  I'd like to hear about an example?

I've learned over time that it's not necessary to respond to certain posts.

[Jim]

Vehemently disagree with you. 
The vast bulk of the cost is paying for the marching army of non contributors that are part of the effort.
They need to get paid whether the project gets done or not.
The cost escalation of both Hubble and the JWST (or Sofia) are because of that.
Move fast and break things is good advice, even in science.

Completely clueless

[VSECOTSPE]

A lovely piece, but I could weep.
No backup at all, on a complex product where the track record for similar deployments is iffy.

The sunshield on JWST is 220m^2.  We’ve successfully deployed similar thin-membranes in space at that scale, like the NanoSail-D2 solar sail at 110m^2.

JWST’s primary mirror only folds into three sections.  And each of those three sections has independently actuated mirror subassemblies to correct inaccuracies and distortions.

The secondary mirror and solar panel deployments are old hat.

frankly, SpaceX's SS can't come soon enough, NASA science spends most of their scarce $$ on product engineering because of launch constraints.
Lift those and the same $$ can do 10x more and better science.

The diameter of SS is 9m.  Even the shorter length of JWST’s sunshield is 10m.  It just won’t fit without folding.

JWST costs ~$10 billion (with a “b”).  The Ariane V that JWST is launching on goes for about $200 million (with an “m”), and ESA is providing the launch in-kind.  Even if SS launches were available at $2 million per and NASA was paying $200 million for that Ariane V, moving to SS would only save $198 million, or less than 2% of JWST’s cost.  While lower cost and better launch is important to a lot of space missions and applications, SS won’t change the economics of a mission like JWST.

Vehemently disagree with you. 
The vast bulk of the cost is paying for the marching army of non contributors that are part of the effort.
They need to get paid whether the project gets done or not.
The cost escalation of both Hubble and the JWST (or Sofia) are because of that.

HST and JWST suffered from poor cost estimates.  HST further suffered from bad optics that were later corrected.  SOFIA suffered from poor engineering and management from concept thru operation.  None suffered from “non-contributors”.  There is no doubt a lot of underutilized workforce at NASA, but the Space Science Enterprise/Directorate has not had that problem since the late 90s/early 00s when its budget was nearly doubled.  In some places these days, there’s too much work.

Move fast and break things is good advice, even in science.

There are places to move faster, try new things, and even break instruments in NASA space science.  But they’re done in labs, suborbitally, or on smaller space missions, not flagship-scale missions.

[Oersted]

A lovely piece, but I could weep.
No backup at all, on a complex product where the track record for similar deployments is iffy.

frankly, SpaceX's SS can't come soon enough, NASA science spends most of their scarce $$ on product engineering because of launch constraints.
Lift those and the same $$ can do 10x more and better science.

Most of the costs on the Webb had little to do with launch constraints.  Most of the costs were associated with building a telescope that has to be so precisely built and maintain its shape with the thermal environment it is operating in.  The cost of developing the cameras to operate in this environment has also been a major challenge.  Starship would not have changed that.  A telescope of that size and capability would have been incredibly expensive even if it didn't have to unfold on the way to its operational location.

But a fixed main mirror and a fixed sunshield, had the fairing been sufficiently roomy, would reduce the risk factor considerably. Maybe the telescope would have cost only 1/3 less but its chance of success would be much bigger, and that surely counts for a lot.

A fixed sunshield could probably be built with smaller dimensions and SpaceX could probably also design a bigger fairing for an expendable Starship. That way there would be perhaps one or two major deployments instead of the many complex ones we see now. 

Crossing fingers and eating peanuts. (Too much) Excitement guaranteed. It will be unbearable if it doesn't work out. Not just for the loss of science but also because it will harden the resistance to big science. Many elected representatives will be swayed from funding future missions if this one fails. 

[Eric Hedman]

A lovely piece, but I could weep.
No backup at all, on a complex product where the track record for similar deployments is iffy.

frankly, SpaceX's SS can't come soon enough, NASA science spends most of their scarce $$ on product engineering because of launch constraints.
Lift those and the same $$ can do 10x more and better science.

Most of the costs on the Webb had little to do with launch constraints.  Most of the costs were associated with building a telescope that has to be so precisely built and maintain its shape with the thermal environment it is operating in.  The cost of developing the cameras to operate in this environment has also been a major challenge.  Starship would not have changed that.  A telescope of that size and capability would have been incredibly expensive even if it didn't have to unfold on the way to its operational location.

But a fixed main mirror and a fixed sunshield, had the fairing been sufficiently roomy, would reduce the risk factor considerably. Maybe the telescope would have cost only 1/3 less but its chance of success would be much bigger, and that surely counts for a lot.

A fixed sunshield could probably be built with smaller dimensions and SpaceX could probably also design a bigger fairing for an expendable Starship. That way there would be perhaps one or two major deployments instead of the many complex ones we see now. 

Crossing fingers and eating peanuts. (Too much) Excitement guaranteed. It will be unbearable if it doesn't work out. Not just for the loss of science but also because it will harden the resistance to big science. Many elected representatives will be swayed from funding future missions if this one fails.

A fixed sunshade would not have been any smaller.  It is sized for just how big it needs to be to keep the telescope cold enough.  And fully deployed with layers separated it might not survive the launch forces stretched out.  It is very fragile.  Packed up protects the thin film and support structure better during the high vibration and acceleration of launch.

[woods170]

The secondary mirror and solar panel deployments are old hat.

Emphasis mine.

I'll be very blunt here: I don't ever want to hear you say that again.
Solar array deployment failures and boom-type deployment failures constitute ~70 percent of all deployment failures on spacecraft.

Also, there is quite a few engineers at Airbus D&S that will vehemently disagree with your statement. And the recent little mishap with Lucy guarantees that NG now also has engineers disagreeing with your statement.

[RotoSequence]

The secondary mirror and solar panel deployments are old hat.

Emphasis mine.

I'll be very blunt here: I don't ever want to hear you say that again.
Solar array deployment failures and boom-type deployment failures constitute ~70 percent of all deployment failures on spacecraft.

Also, there is quite a few engineers at Airbus D&S that will vehemently disagree with your statement. And the recent little mishap with Lucy guarantees that NG now also has engineers disagreeing with your statement.

Now that you mention it, I'd rather like to live in the world where these things aren't nerve wracking moments for the mission engineers. It'd be great if there was a good way to go about that.

[VSECOTSPE]

Emphasis mine.

I'll be very blunt here: I don't ever want to hear you say that again.
Solar array deployment failures and boom-type deployment failures constitute ~70 percent of all deployment failures on spacecraft.

Around the world, literally hundreds of solar arrays and booms deploy successfully every year.  JWST’s solar array and secondary mirror are old hat from that perspective.  They’re also old hat compared to, say, JWST’s thermal blankets, where we have only a handful-ish of prior thin-membrane deployments.

That doesn’t take away from your points.  Engineers are human.  Once in a while we screw up.  And some systems are more prone to failures than others by the nature of the physics involved.  And so there needs to be effective checks and balances to minimize those human errors and doublecheck those error-prone systems.  We should never say, “Well, we don’t need to test the solar array deployment because that’s been done before.”

But to be clear, that’s not what I was saying or implying.

[Blackstar]

What an awesome Christmas present, it would be.

A lot of the people involved were hoping to launch it and then go home to their families for Christmas. So I doubt they think this is awesome.

Guess I'd feel a lot better if the whole thing was put off a couple of weeks.
Let the staff enjoy the holidays and then come back refreshed to do a great job. After a 10 year delay, it is not that the schedule really matters.

The logistics is too great

https://spacenews.com/communications-problem-delays-jwst-launch/

"He added that more than 150 people from NASA are in Kourou to support the launch. “We have given to NASA to maximum visibility into our launcher.”

So that's 150 people who are being supported at the location. Flying them all home, then back again, would be difficult.

[etudiant]

Emphasis mine.

I'll be very blunt here: I don't ever want to hear you say that again.
Solar array deployment failures and boom-type deployment failures constitute ~70 percent of all deployment failures on spacecraft.

Around the world, literally hundreds of solar arrays and booms deploy successfully every year.  JWST’s solar array and secondary mirror are old hat from that perspective.  They’re also old hat compared to, say, JWST’s thermal blankets, where we have only a handful-ish of prior thin-membrane deployments.

That doesn’t take away from your points.  Engineers are human.  Once in a while we screw up.  And some systems are more prone to failures than others by the nature of the physics involved.  And so there needs to be effective checks and balances to minimize those human errors and doublecheck those error-prone systems.  We should never say, “Well, we don’t need to test the solar array deployment because that’s been done before.”

But to be clear, that’s not what I was saying or implying.

Apologize if what I said about the 'marching army of non contributors' is deemed offensive, but it remains true that the bulk of aerospace costs are government mandated administrative overheads. That is equally true for NASA as well as for military programs and helps understand why privately administered SpaceX was able to underprice ULA's Atlas by more than half.
Those administrative costs are annually recurring, add zip to the technology effort, just balloon the cost of any long term development effort such as the JWST. Any protracted development gets dragged down by this burden, so there is an inherent conflict between NASA's innovation mandate and their budget constraints. The miracle of the JWST is that it has survived despite this, but it sure would not have cost much extra to build two instead of just one.

[Jim]

, but it remains true that the bulk of aerospace costs are government mandated administrative overheads. That is equally true for NASA as well as for military programs and helps understand why privately administered SpaceX was able to underprice ULA's Atlas by more than half.

That is wrong on many levels

[VSECOTSPE]

Apologize if what I said about the 'marching army of non contributors' is deemed offensive, but it remains true that the bulk of aerospace costs are government mandated administrative overheads.  That is equally true for NASA as well as for military programs and helps understand why privately administered SpaceX was able to underprice ULA's Atlas by more than half.

It’s not that what you’re saying is offensive.  It’s that what you’re saying is not true and has nothing to do with JWST.

It is true that SX is largely vertically integrated, which means they don’t have a lot of external suppliers and subcontractors that need to show a profit margin.  It is true that this gives SX a pricing advantage vice horizontally integrated launch providers like ULA who have to feed the profit margins at many external suppliers and subcontractors.  It is _not_ true that this is due to “government mandated administrative overheads” (otherwise SX would suffer the same as a NASA and DOD contractor).  And it is _not_ true that this represents the “bulk of aerospace costs”.  And most importantly, the profit structure of launch providers like SX and ULA has _nothing to do_ with the costs of science spacecraft like JWST, which are built by neither.

it sure would not have cost much extra to build two instead of just one.

The production learning curve is nowhere near that steep.  A second copy of JWST would have cost billions of dollars.  It would easily have eaten the budget for the Roman space telescope that is to follow JWST.

[freddo411]

it sure would not have cost much extra to build two instead of just one.

The production learning curve is nowhere near that steep.  A second copy of JWST would have cost billions of dollars.  It would easily have eaten the budget for the Roman space telescope that is to follow JWST.

I don’t believe the assertion that a copy would cost billions more.    Did the spirit and opportunity mission cost billions more than a single copy?  Compare the cost with the persy program.

[RotoSequence]

The production learning curve is nowhere near that steep.  A second copy of JWST would have cost billions of dollars.  It would easily have eaten the budget for the Roman space telescope that is to follow JWST.

Isn't Roman fundamentally, and considerably less capable than JWST?

[Blackstar]

I don’t believe the assertion that a copy would cost billions more.    Did the spirit and opportunity mission cost billions more than a single copy?  Compare the cost with the persy program.

The two rovers cost about twice as much as a single rover. But comparing MER to JWST is a lousy comparison. JWST is a considerably more complex mission.

[LouScheffer]

Solar array deployment failures and boom-type deployment failures constitute ~70 percent of all deployment failures on spacecraft.

Around the world, literally hundreds of solar arrays and booms deploy successfully every year. 

Both of these statements are true.  Every year there are hundreds of deployments, and only a few fail.  But of those that fail, most are solar arrays or booms.

However, my worry level is much closer to Woods170.   The reason most deployments succeed is that the bugs have been worked out.  New deployment designs, even those very similar to existing designs, are much more likely fail (ask Lucy, or Galileo, or ...).  Webb has lots of one-off deployment mechanisms.  Sure, they are based on existing designs that are known to work.   But that's only mildly comforting - most designs that fail were also based on designs that worked.   And Webb adds designs without even this heritage, such as the quite complex sunshield deployment.  I guessing at least something will not deploy according to plan, and hope it's recoverable.

[Jim]

it sure would not have cost much extra to build two instead of just one.

The production learning curve is nowhere near that steep.  A second copy of JWST would have cost billions of dollars.  It would easily have eaten the budget for the Roman space telescope that is to follow JWST.

I don’t believe the assertion that a copy would cost billions more.    Did the spirit and opportunity mission cost billions more than a single copy?  Compare the cost with the persy program.

MSL ans M2020