Hacker News Reader

Comment by sandworm101 | Hacker News Reader

There are differences in the ULA and SpaceX launch options. There is far more to spacelaunch than cost per kilo to orbit.

One of the big issues is that SpaceX doesn't do vertical payload integration. It's upper stage options are also very different than those of ULA. For many sats that isn't an issue, for some military sats it is a very big deal. When the payload is worth many times as much as the rocket, with incalculable trickledown impacts should its launch not go perfectly, your choice of rocket has to tick absolutely every box.

View on HN

Participation in the second NSSL requires vertical payload integration. SpaceX had to promise to be able to do it to submit a bid.

They have never done it so far, but they will definitely do it in the near future.

What is vertical payload integration?

With rockets like the Atlas 5 (flown by ULA), the rocket is assembled vertically in a tall "hanger", and is then rolled out to the launch pad, remaining in the vertical position.

The Falcon 9 is assembled rolled out to the pad horizontally, then it raised into the vertical position in preparation for launch.

It can be more complicated to design satellites (esp satellites that have large optics in them) to deal with gravity in two different orientations.

Atlas 5 "roll out": https://spaceflightnow.com/2020/07/28/photos-atlas-5-rolled-...

Falcon 9: https://www.youtube.com/watch?v=TYJuHuhPxGU

Bonus video, Smarter Every Day chatting w/ the CEO of ULA while watching a Delta IV Heavy roll out to the launch pad: https://www.youtube.com/watch?v=OdPoVi_h0r0

Complicated and less weight-efficient. Designing a structure for gravity/g/force in one axis is wasteful enough. That strength is useless once in space. Horizontal integration means the structure must be strong in two axis: double waste. It might not matter when you are launching cubesats, but in a spy satellite the size/weight of a bus that extra reinforcement means sacrifices in many other areas.

But it's more economical, money-wise, as Russian launchers history shows. In general, American rocket school tends to have better mass efficiency but less economical efficiency that Russian rocket school.

"When he wants more thrust, Ivan doesn't look for a fancy propellant with a higher specific impulse. He just builds himself a bigger rocket. Maybe he's got something there." - J. Clark, Ignition! https://library.sciencemadness.org/library/books/ignition.pd...

> When he wants more thrust, Ivan doesn't look for a fancy propellant with a higher specific impulse. He just builds himself a bigger rocket

On the other hand the Soviet Union did develop syntin... (also from Ignition!)

Of course the point was that one can use simple and crude solutions, or effective and sophisticated ones. Another example would be that in Russian rocket engines pumps tend to be put on the same axis as the turbine, so frequencies of turbine and pumps are the same and some of them can be non-optimal. In American tradition one uses gears, which makes both turbine and pumps (more) optimal, for the cost of gear mechanism. Practice shows that both approaches can work, but, just as with vertical assembly, which requires a tall building or a vertical transportation of a rocket, some ways look simpler than others. Simpler not necessarily better - but in rocket assembly it's probably cheaper to do it horizontally.

It's a bit more than cubesats. Nearly every satellite is or can be fine with horizontal integration. Witness every SpaceX and Russian payload ever launched.

Only exotic (presumably optical) sats from three-letter agencies really demand it, and even then we don't know if they really need it or would rather just spend the money to make it the launcher's problem rather than do a redesign or evaluation of their system.

Today’s cheap and high cadence commercial launch systems should make the extra weight needed for horizontal assembly a non-factor.

"Cheap and high cadence commercial launch systems" are just beginning to happen right now, it'll take time for the industry practices to readjust.

SpaceX undercut the market price for space launches by nearly 70% in 2010, and has averaged a launch every two and a half weeks for the last four years.

How much tine does the industry need to adjust?

Lead times for large satellites are measured in years. The industry is absolutely adjusting, but it's also very risk averse.

How is the satellite transported to the launch site from the point of manufacture?

Usually by truck for at least part of the journey (airplanes, boats, and trains are also commonly used). They are carried in custom built frames that hold them in the correct orientation. The frame is then surrounded by a sealed box that keeps the satellite clean, and with the appropriate climate controls (temperature, humidity, etc).

Here's a good overview of the process (though in this case it's a satellite that can safely be transported in multiple orientations): https://qz.com/1346279/big-plane/

It just seems to me that it would need to be horizontal, and likely encounter greater stresses during transport than during a tightly controlled horizontal assembly process.

They are not horizontal during transport.

ULA basically built a portable clean room around the tip of the vertical rocket, to put the payload on top. Because the rover has to be free of earth organisms.

SpaceX is getting money to do vertical integration for this contract.

Ya, but they haven't done it yet. When you are organizing launch contracts for billion-dollar spy satellites you only ever go with people who have done things a few times before.

Like the Vulcan rocket which has never flown?

Has ULA ever designed a new rocket?

ULA is just a consortium of Lockheed and a Boeing, so ‘technically’ no, but in practice the actual people and facilities have, yes.

The Delta IV and Atlas V were developed twenty years ago. Given turnover and the new management structure, lots of those people are gone or not in the same roles.

ULA has designed and is building the Vulkan rocket now. Presumably they have some staff that were involved with the Atlas V and Delta IV back when they were with their parent companies.

That was 20 years ago, they’ve had lots of turnover since. So while they clearly retain some expertise, most appears to be in launch operations, not launch system design. Which is one reason they had to farm out engine design to Blue Origin.

Treating the ULA bid as if it were lower risk than SpaceX appears to be contrary to the facts. The Falcon 9 has been flying for a decade, and has a very good safety record.

The Vulkan is a paper rocket that is years from actual service using an unproven engine, being designed by an effectively new team.

Reply on news.ycombinator.com