Author Topic: Why does SpaceX attach reaction frame to TE, and Russians not?  (Read 13560 times)

Offline LouScheffer

  • Senior Member
  • *****
  • Posts: 3381
  • Liked: 6109
  • Likes Given: 836
Random question here.  SpaceX attaches their reaction frame to the transporter, but the Russian's don't.  (see below)  Any ideas about why  each of them made the choices they did? The Russian approach, to me, seems like it has one less fluid/electrical connection to worry about.



Offline abaddon

  • Senior Member
  • *****
  • Posts: 3051
  • Liked: 3900
  • Likes Given: 5273
One possibility is that FH will have a different reaction frame (which I believe is true?), so having it permanently installed is not an option.

Offline Mike_1179

  • Member
  • Full Member
  • ****
  • Posts: 670
  • New Jersey
  • Liked: 383
  • Likes Given: 87
I was under the impression that the Soyuz is not held up from the ground like the Falcon 9 but is suspended where the boosters attach to the core. This way, when the rocket is vertical, it is being loaded the saw way it is loaded in flight.

Offline pippin

  • Regular
  • Senior Member
  • *****
  • Posts: 2575
  • Liked: 312
  • Likes Given: 45

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 37440
  • Cape Canaveral Spaceport
  • Liked: 21450
  • Likes Given: 428
Random question here.  SpaceX attaches their reaction frame to the transporter, but the Russian's don't.  (see below)  Any ideas about why  each of them made the choices they did? The Russian approach, to me, seems like it has one less fluid/electrical connection to worry about.


There is no reaction frame for Soyuz.

The Russian approach doesn't save any connections.    Each stage/core/booster has at least one.

For Spacex:
The services for the upper umbilicals (which are on the mast/erector/transporter) come through connections on launcher frame.

Most of the mass of the launch vehicle is supported by the frame during transport.
« Last Edit: 04/26/2017 07:00 pm by Jim »

Offline old_sellsword

  • Full Member
  • ****
  • Posts: 632
  • Liked: 531
  • Likes Given: 470
One possibility is that FH will have a different reaction frame (which I believe is true?), so having it permanently installed is not an option.

There will only ever be one reaction frame per pad, 39A's will do double duty for F9 and FH.

Offline darkenfast

  • Member
  • Full Member
  • ****
  • Posts: 1539
  • Liked: 1829
  • Likes Given: 8739
Remember, the Soyuz, with its "Tulip" support, was the FIRST really big rocket, so the Soviets were inventing the whole concept of how to launch something of that size out in the near-desert conditions of Khazakstan, with its extremes of weather.   Since then, they have pragmatically used a variety of pad systems to suit the rocket, but the one consistent choice they prefer is horizontal integration.  Look at the similarities and differences between Sozuz, Proton, Zenit and the N-1/Buran complex.
Writer of Book and Lyrics for musicals "SCAR", "Cinderella!", and "Aladdin!". Retired Naval Security Group. "I think SCAR is a winner. Great score, [and] the writing is up there with the very best!"
-- Phil Henderson, Composer of the West End musical "The Far Pavilions".

Offline deruch

  • Senior Member
  • *****
  • Posts: 2422
  • California
  • Liked: 2006
  • Likes Given: 5634
Random question here.  SpaceX attaches their reaction frame to the transporter, but the Russian's don't.  (see below)  Any ideas about why  each of them made the choices they did? The Russian approach, to me, seems like it has one less fluid/electrical connection to worry about.

SpaceX from the very beginning has been talking about reducing the amount of time it takes to exit the integration facility, go erect, and launch.  Maybe they figured that they could launch faster this way?  i.e. If they determined that it would take them longer to mate (and checkout) the vehicle to the reaction frame than it would to mate (and checkout) the frame to the pad, then doing the vehicle/frame mating in the HIF would save them time.
Shouldn't reality posts be in "Advanced concepts"?  --Nomadd

Offline dmc6960

  • Member
  • Full Member
  • **
  • Posts: 277
  • Liked: 34
  • Likes Given: 3
Where did this name "reaction frame" come from?  Its always been called a launch mount or launch table.
-Jim

Offline Zucal

  • Member
  • Posts: 81
  • California
  • Liked: 288
  • Likes Given: 17
'Reaction frame' or 'reaction fixture' is what it's called by SpaceX.

Offline Grandpa to Two

Correct me if I'm wrong but older models of Soyuz didn't roll, so the table turned to set the azimuth. Is that the correct term? Therefore the table and it's turning mechanisms needed to be static. I can't provide a reference but it may have been Anatoly Zak.
"All truths are easy to understand once they are discovered; the point is to discover them" Galileo Galilei

Offline cppetrie

  • Full Member
  • ****
  • Posts: 792
  • Liked: 552
  • Likes Given: 3
Correct me if I'm wrong but older models of Soyuz didn't roll, so the table turned to set the azimuth. Is that the correct term? Therefore the table and it's turning mechanisms needed to be static. I can't provide a reference but it may have been Anatoly Zak.
Most versions of Soyuz rockets can't perform a roll maneuver. That is still true for some versions still flying today. Only the most recent versions can perform a roll on their own and don't require the launch platform to negate the need for it.

Offline baldusi

  • Senior Member
  • *****
  • Posts: 8356
  • Buenos Aires, Argentina
  • Liked: 2539
  • Likes Given: 8273
Correct me if I'm wrong but older models of Soyuz didn't roll, so the table turned to set the azimuth. Is that the correct term? Therefore the table and it's turning mechanisms needed to be static. I can't provide a reference but it may have been Anatoly Zak.
Most versions of Soyuz rockets can't perform a roll maneuver. That is still true for some versions still flying today. Only the most recent versions can perform a roll on their own and don't require the launch platform to negate the need for it.
More specifically, Soyuz-2 can perform roll maneouvers. Thus, of he currently launching, only the crew rated Soyuz-FG (and the Soyuz-U before) can't roll. But the pad structure an most systems in general for the modern Soyuz-2.1a/b/v have this legacy. But modern Soyuz pads like ELS in Kourou or the pads in Voistochny, do not feature the rotary table.

Offline IainMcClatchie

  • Full Member
  • ***
  • Posts: 394
  • San Francisco Bay Area
  • Liked: 279
  • Likes Given: 411
How does Soyuz achieve roll stability?

And, how do they manage to avoid being able to roll?  All that's needed is multiple independently gimballed rocket motors.  At liftoff, those five clusters are obviously independent, and if they can't be gimballed at all, how does the rocket achieve guidance?

Once the boosters are gone, the core still has the four smaller, independently gimballed steering motors, right?  So why can't it use those to roll?

Offline cppetrie

  • Full Member
  • ****
  • Posts: 792
  • Liked: 552
  • Likes Given: 3
How does Soyuz achieve roll stability?

And, how do they manage to avoid being able to roll?  All that's needed is multiple independently gimballed rocket motors.  At liftoff, those five clusters are obviously independent, and if they can't be gimballed at all, how does the rocket achieve guidance?

Once the boosters are gone, the core still has the four smaller, independently gimballed steering motors, right?  So why can't it use those to roll?
I think it has to do with the analog flight computers the older versions use(d). Unable to iterate fast enough to control the complex motion and independently gimbaled engines, perhaps? Some who is a rocket scientist might have to explain.

Offline CorvusCorax

  • Full Member
  • ****
  • Posts: 1860
  • Germany
  • Liked: 4009
  • Likes Given: 2738
How does Soyuz achieve roll stability?

And, how do they manage to avoid being able to roll?  All that's needed is multiple independently gimballed rocket motors.  At liftoff, those five clusters are obviously independent, and if they can't be gimballed at all, how does the rocket achieve guidance?

Once the boosters are gone, the core still has the four smaller, independently gimballed steering motors, right?  So why can't it use those to roll?

I think it has to do with the analog flight computers the older versions use(d). Unable to iterate fast enough to control the complex motion and independently gimbaled engines, perhaps? Some who is a rocket scientist might have to explain.

That actually makes sense. The rocket is *physically* able to roll. But they weren't confident about mixing the controls on non-independent physical axis together and having them still work reliably.

Meaning:
There is a simple roll control loop, but all it does is zero out any existing roll tendency and bring the rocket back to a zero roll attitude.
There's also a pitch control loop that only does the correct thing when the rocket is at zero roll, but wouldn't work when the roll is non-zero.
(Remember, if the roll is nonzero, the engines need to gimbal in two directions instead of just one -- sideways)
There likely is also a yaw control loop, which is also simple and makes sure the rocket doesn't deviate sideways. It would have the same limitations as the pitch loop and stop working if it suddenly had to give mixed output to two different control axis.

Now mixing pitch and yaw together based on roll is technically dead easy - IF you have a really good roll estimate, and if the two axis (pitch and yaw, aka horizontal and vertical gimbaling) share the same parameters - control response, maximum throw, speed/delay, ...) but in practice both isn't the case, your roll estimate can have errors, and on top of it you'd be trying to do this with 1960's Soviet era analog computers.

So you apply kiss principle: The less error prone solution is to keep roll zeroed out in a straightforward 1 axis controller, yaw perfectly straight in another 1 axis controller, and roll the launch table as needed (especially since back then they likely had rotatable launchtables anyway, so that was likely a nobrainer - aything after that falls under category "never touch a running system" (spoken with Russian accent in this case))
It's a much less complex solution. Granted it involves giant machinery to rotate the platform but that only needs to be built once, and then can be measured very exactly if it points in the right direction before launch.

Versus a dynamic system that can produce errors in flight and get the rocket of course.

nowadays in the day and age of digital controls that's an antiquated constraint, but it's pretty obvious where it's coming from. I'd do the same thing in that situation.


Offline drzerg

  • Full Member
  • *
  • Posts: 117
  • Kyiv
  • Liked: 56
  • Likes Given: 24
I think it has to do with the analog flight computers the older versions use(d). Unable to iterate fast enough to control the complex motion and independently gimbaled engines, perhaps? Some who is a rocket scientist might have to explain.

Soyuz-FG still has it. and if i am not mistaken it was(still?) manufactured in Kharkiv (Ukraine) only Soyuz-2 has new all digital russian manufactured flight computers.

Offline cppetrie

  • Full Member
  • ****
  • Posts: 792
  • Liked: 552
  • Likes Given: 3
I think it has to do with the analog flight computers the older versions use(d). Unable to iterate fast enough to control the complex motion and independently gimbaled engines, perhaps? Some who is a rocket scientist might have to explain.

Soyuz-FG still has it. and if i am not mistaken it was(still?) manufactured in Kharkiv (Ukraine) only Soyuz-2 has new all digital russian manufactured flight computers.
Yes. With the recent retirement of the Soyuz-U, the only remaining version with analog computers is the Soyuz-FG. Oddly enough, it is also the only currently man-rated rocket in service. Seems kinda crazy that the only rocket we'll send people to space in is using analog computers.

Offline CorvusCorax

  • Full Member
  • ****
  • Posts: 1860
  • Germany
  • Liked: 4009
  • Likes Given: 2738
I think it has to do with the analog flight computers the older versions use(d). Unable to iterate fast enough to control the complex motion and independently gimbaled engines, perhaps? Some who is a rocket scientist might have to explain.

Soyuz-FG still has it. and if i am not mistaken it was(still?) manufactured in Kharkiv (Ukraine) only Soyuz-2 has new all digital russian manufactured flight computers.
Yes. With the recent retirement of the Soyuz-U, the only remaining version with analog computers is the Soyuz-FG. Oddly enough, it is also the only currently man-rated rocket in service. Seems kinda crazy that the only rocket we'll send people to space in is using analog computers.

Actually not that crazy. If you look at the resiliency of these analog systems - especially in respect to cosmic radiation events - and compare that to modern CMOS technology,  the difference is fundamental.

In an analog computer the rad event causes a momentary spike in a voltage level, but not significant enough to affect the accumulated voltage in capacitors. It's basically just static noise, which won't significantly affect the computation result.

In a digital system, a rad event can cause a gate to flip and turns a 1 into a 0 or the other way around - effectively altering memory in a persistent way - and depending where that happens (worst case the most significant bit in an accumulator) it can have drastic results on the control output. The system is no longer inherently tolerant, as such you need much more complex external redundancy (like SpaceX multiple computers checking each other) or use shielding and hardening of the component to make radiation less likely to flip bits.

The advantage of digital systems, they are cheaper and much more capable. Digital system are also much easier to test and certify in simulation.

Because of that, effectively no one in their right mind would design a new rocket with an analog system these days. But if you have a flight proven* rocket with all the required functionality already in it, with hundreds of launches flight-heritage, that would actually be a very safe system to fly. They might be extremely antiquated, but they also have unmatched robustness.

A bit similar to the question: How do you certify your space rated checklist-tablet to operate in an emergency situation?
a) extra batteries under the seat
b) solar cells and an emergency hand-crank on a dynamo to recharge
c) the tablet uses paper and comes with a pencil on a string


Edit: *Due to recent events and to prevent confusion, I think the attribute "flight-proven" should only be used in the context of individual rockets, not rocket families. The term "flight-heritage" fits better.
« Last Edit: 05/18/2017 09:15 am by CorvusCorax »

Offline cppetrie

  • Full Member
  • ****
  • Posts: 792
  • Liked: 552
  • Likes Given: 3
I think it has to do with the analog flight computers the older versions use(d). Unable to iterate fast enough to control the complex motion and independently gimbaled engines, perhaps? Some who is a rocket scientist might have to explain.

Soyuz-FG still has it. and if i am not mistaken it was(still?) manufactured in Kharkiv (Ukraine) only Soyuz-2 has new all digital russian manufactured flight computers.
Yes. With the recent retirement of the Soyuz-U, the only remaining version with analog computers is the Soyuz-FG. Oddly enough, it is also the only currently man-rated rocket in service. Seems kinda crazy that the only rocket we'll send people to space in is using analog computers.

Actually not that crazy. If you look at the resiliency of these analog systems - especially in respect to cosmic radiation events - and compare that to modern CMOS technology,  the difference is fundamental.

In an analog computer the rad event causes a momentary spike in a voltage level, but not significant enough to affect the accumulated voltage in capacitors. It's basically just static noise, which won't significantly affect the computation result.

In a digital system, a rad event can cause a gate to flip and turns a 1 into a 0 or the other way around - effectively altering memory in a persistent way - and depending where that happens (worst case the most significant bit in an accumulator) it can have drastic results on the control output. The system is no longer inherently tolerant, as such you need much more complex external redundancy (like SpaceX multiple computers checking each other) or use shielding and hardening of the component to make radiation less likely to flip bits.

The advantage of digital systems, they are cheaper and much more capable. Digital system are also much easier to test and certify in simulation.

Because of that, effectively no one in their right mind would design a new rocket with an analog system these days. But if you have a flight proven* rocket with all the required functionality already in it, with hundreds of launches flight-heritage, that would actually be a very safe system to fly. They might be extremely antiquated, but they also have unmatched robustness.

A bit similar to the question: How do you certify your space rated checklist-tablet to operate in an emergency situation?
a) extra batteries under the seat
b) solar cells and an emergency hand-crank on a dynamo to recharge
c) the tablet uses paper and comes with a pencil on a string


Edit: *Due to recent events and to prevent confusion, I think the attribute "flight-proven" should only be used in the context of individual rockets, not rocket families. The term "flight-heritage" fits better.
Wasn't trying to suggest it was a safety issue, just that it seemed a bit odd that in a virtually 100% digital computing environment of 2017, the entire world's only man-rated rocket is using analog computers. I guess to me it is a testament to how badly we (the US specifically) have lapsed in maintaining/advancing HSF capabilities.

Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
1