Hacker News Reader

Comment by larrydag | Hacker News Reader

I've been super impressed with how efficient SpaceX is in getting launches into orbit. I did a visual comparison to the Space Shuttle program. It looks like SpaceX will overtake the Space Shuttle launch efficiency by the end of the year.

http://snag.gy/wUTD7.jpg

sources:

http://en.wikipedia.org/wiki/List_of_space_shuttle_missions

http://en.wikipedia.org/wiki/List_of_Falcon_9_launches

View on HN

That graph might be a little more intuitive if you swapped the axes, FWIW.

Definitely. Also choose colours that aren't almost identical.

The blue and black don't look very close to me, and they're pretty color-blind safe. Odd.

I cannot tell them apart, other than length.

Be nice to have one solid, the other dashed or dotted.

Don't underestimate the damage of bad screens, bad screen settings and people blaming the source rather than their setup.

Even just laptop screen brightness or glare can play a huge role here. I know it's offtopic, but if anyone is aware of a guide or methodology for picking colors that will be less susceptible to this problem, I would love to see it.

Convert to grayscale.

colorbrewer

'Cheat' and use something else to differentiatie between the two, in addition to color (line thickness, line dash, markers)

You probably should always do that, as you won't know where your chart will be used (printed in black and white, very lossily converted to JPEG, whatever)

It's all right to blame the source. Imagine half a population having grayscale screens. Would you pick colors that map to the same shade of gray? I guess not, so why would you pick colors and thickness that can be mistaken on so many setups?

In this case it doesn't matter, as it's obvious which line is which.

It would be a lot more intuitive. Even knowing exactly what was being described and how - It still is difficult for me to understand a graph in which Time is on the Y-axis. I considered for a few moment that I was actually being trolled with that graph.

The space shuttle launches are manned missions. I don't think it's a fair comparison.

good point. I was considering the comparison as resupply missions and (future) re-usability of the vehicle.

can you compare these to the Russian unmanned supply runs to the ISS?

Shuttle was alway fairly low-frequency, far below its design goals, so possibly not a great comparison. Here's 2014 in space launches: http://en.wikipedia.org/wiki/2014_in_spaceflight

At peak, in the 70s, there were about 140 launches per year.

An expendable vehicle should be inherently more efficient in launching frequency than a reusable one. The reusable one is fundamentally limited by the number of reusable vehicles you have on hand and the time it takes to turn each around. The expendable one is only limited by the speed of your manufacturing.

So, since SpaceX is moving to a reusable vehicle, this graph could easily change slope in the future.

Telecoms and the market for satellites has only recently boomed. You cannot compare the two programs in such a way.

The limiting factor of the Shuttle wasn't low demand for its services, I don't think.

Demand was ~1/10 of projected demand which dramatically increased costs not only by increasing overhead by also by extending vehicle lifespans well beyond what they where designed for.

Put another way, scrapping a shuttle mission saved less than 90 million.

PS: It was also horribly compromised based on Airforce requirements yet never flew a single Airforce mission.

I am not sure what the rated vehicle lifespans were at the time of building, but the design life was 100 missions per shuttle, and the most-flown shuttle (Discovery) only flew 39.

The initial assumption was 40 missions a year. Based on the estimated risks you might get 100 missions from a shuttle on average. So 5 orbitors * 100 / 40 = 12.5 years or ~ 10 years with an early loss. They where in service for 30 years and averaged 4.5 missions per year.

For comparison compare the matence cycle a low mileage 0 to 10 year old car vs a low mileage 15 to 30 year old car.

I've heard this reasoning before (from NASA personnel), but I have never seen the reasoning behind it. Many of the design life requirements were impossible to meet to begin with, such as the 50-launch design life of the Space Shuttle main engines (because of the high chamber pressures, and problems with bearings). If you have some evidence of the veracity of this claim, I would be very interested to read it.

I wonder if that was "We believe they are capable of 100 missions, because we require 100 missions" or "We believe they are capable of 100 missions, because that is what our test data actually suggests".

See: Feynman's Appendix F. Specifically his criticism of the Shuttle reliability numbers quoted by NASA management and how they differed from the reliability numbers quoted by NASA engineers.

Design life and rated life (as built) are two completely different things. The Space Shuttle requirements stated that it should be capable of 100 missions, but I have never seen any document describing what they were rated for, after they were constructed, and the flight tests were completed.

> PS: It was also horribly compromised based on Airforce requirements yet never flew a single Airforce mission.

The shuttles flew a few DoD missions, including I believe at least one for the Air Force (a DSCS satellite during STS-51-J I believe? Maybe others, which DoD mission was for who isn't nicely broken down anywhere that I can find. It looks like most were for the NRO).

However none of the Shuttles ever flew a polar orbit mission; an ability that the Air Force demanded but never used. The large cross-range capabilities of the shuttle (necessitating such large wings) was for polar orbit missions, so those massive wings were dead weight caused by the Air Force.

Funny because wasn't the shuttle-c used for military missions and it has almost no wings.

What do you mean? The shuttle-c was never built.

I meant its punier newer incarnation, http://en.wikipedia.org/wiki/Boeing_X-37

I should have been more clear.

The airforce had two mission profiles outside NASA's goals. Polar orbits AND capturing Russian satilites. Neither of those missions types where ever flown. The Airforce ended up having a huge impact on the design including much larger cargo bay etc but it was those two mission profiles that really fucked up the design.

If NASA had gotten it's way the shuttle would have been more reliable and cost less than 1/3 as much per mission at the cost of significantly lowered cargo capasity. Intact both shuttle disasters can be traced back to these compromises.

Not that I disagree with you, but capturing Russian satellites may be the kind of thing you want the ability, but are not actually planning to do it.

Funnily enough, Russia 'copied' the shuttle design for the Buran.

That ship was almost exclusively designed for millitary use. It was speculated that they planned to carry nuclear weapons up in the cargo bay.

I know, wouldn't a constellation of nukes in low earth orbit be amazing. It would only take 200 to 300 seconds to deorbit from LEO.

Ask the people of Palomares Spain how much they like that idea. They had a couple of nuclear bombs accidentally drop out of the sky there after a mishap in the 60's when the US had nukes in the air at all times.

From what I've read, the idea doesn't actually work very well. Since they're in orbit, you only have a relatively small window per orbit to get it to hit any particular place. If you want to be able to hit any particular place with that kind of notice, you need like dozens of them in the same orbit. Maybe a bunch of orbits, depending on how the cross-range capability is. You could well end up with a constellation of hundreds in orbit.

You're better off putting those weapons on surface-based ICBMs. It'll take a little longer to get to target than the idea case in orbit, but you can launch all of them at a small area at once if you need to. They're a lot easier to maintain and decommission on the ground too - worth considering, since I think we'd all rather not ever have to use the things.

First strike capability, especially hitting their ICBMs before they could make a decision and launch, was considered very important, and may have been considered worth the additional expense/complication. Probably would have been massively destabilizing to reduce lead time that much, though.

Thinking about some more in strategic terms, I think you'd get the most strategic difference by having a lot of weapons in roughly the same orbit. That would give you the ability to conduct an extremely rapid first strike, but only at the time that they happened to be in the right orbit. The downside is that they'd be a terrible second strike weapon, since it would be obvious where they were and they couldn't do much in the way of armoring them or having them avoid enemy weapons. So yeah, it would be pretty damn destabilizing to build a weapon system so massively good at first strike and so bad at second strike. It would tend to tempt the builder to use it or risk losing it in a tense situation, and the opposing nation would know this and perhaps be tempted to make their own first strike or build their own system.

It's probably better/more stabilizing to do what we did, which is to deploy lots of SSBNs - just okay at first strike, great at second strike since it's almost impossible to destroy them all at once. You're less likely to want to pull the trigger on your own first strike when you know you have a nearly unstoppable second strike capability. The enemy nation is less likely to make a first strike if they believe they can't stop your second strike.

Both sides feared their opponent might obtain first strike, but did not want it themselves.

Problem is that some actions that help prevent your opponent from obtaining first strike move you closer to first strike.

For example, having nuclear missiles in submarines makes it very hard for your opponent to take out most of your nuclear weapons in a first strike, but it also gives you the capability to fire from closer range, giving the opponent less time to launch their missiles before your missiles take them out.

This would be similar, as there would be less time between launch and impact.

You say amazing, I say horrifying and a violation of the treaties against militarisation of space. Also the world's worst maintenance liability.

Grave of Fireflies is the saddest movie I have ever seen.

Space is already militarized, and out of the entire population of HN you were the only one to complain about my comment. I didn't get a single downvote.

https://www.youtube.com/watch?v=d3oSS_FP-cA&src_vid=ZWSMoE3A...

Actually, in later years the internal reasoning has been published, so we no longer need to speculate. (This following quotes are taken from the book "Energiya-Buran, the Soviet Space Shuttle" by Bart Hendrickx and Bert Vis). And the actual reason is even more hillarious.

So, first, we now know the US space shuttle was the outcome of some bureaucratic processes in the US (NASA wants a new program to replace Apollo, the Air Force accidentally ends up in a super good negotiating position, Nixon does not want to be responsible for ending manned flight but also doesn't want to pay for a trip to Mars), so in the end it didn't end up a very sensible design. But the Soviets did _not_ known this, and assumed there must be some secret rational reason for building it:

> As TsNIIMash [Central Scientific Research Institute of Machine Building] director Yuriy Mozzhorin later said: "[The Space Shuttle] was introduced as a national program, aimed at 60 launches per year ... All this was very unusual: the mass they had been putting into orbit with their expendable rockets hadn't even reached 150 tons per year, and now they were planning to launch 1,770 tons per year. Nothing was being returned from space, and now they were planning to bring down 820 tons per year. This was not simple a program to develop some space system ... to lower transportation costs (they promised they would lower those costs tenfold, but the studies done at our institute showed that in actual fact there would be no cost savings at all). It clearly had a focused military goal".

So what might this secret military goal be? Apparently the Soviet speculations was that either it might be part of some nuclear fractional-orbit bombardment system, or it might be part of a plan to launch lots of laser weapons into orbit. In any case, we can't have a space shuttle gap, so:

> "[Central Committe Secretary for Defense Matters, Dmitri] Ustinov had made the following argument: if our scientists and engineers do not see and specific use of this technology now, we should not forget that the Americans are very pragmatic and very smart. Since they have invested a tremendous amount of money in such a project, they can obviously see some useful scenario which is still unseen from Soviet eyes. The Soviet Union should therefore develop such a technology so that it won't be taken by suprise in the future".

Let's build it now, and see what it is good for later! And then the Americans started speculating about what the secret military purpose of the Soviet shuttle might be! :D

Nor is it for Falcon 9. They're launching them as fast as they can build them, and have a significant payload backlog.

Are you taking into account inflation adjusted cost?

Reply on news.ycombinator.com