Author Topic: BFS Design Requirements  (Read 18726 times)

Online Norm38

  • Full Member
  • ****
  • Posts: 1112
  • Liked: 473
  • Likes Given: 666
Re: BFS Design Requirements
« Reply #20 on: 02/07/2018 05:39 PM »
I posted this question in a thread that's now locked, so I'll ask again here.

They will need to test landing and takeoff from unimproved terrain.  Nothing beats a real test.  So couldn't they use Spaceport America for that?  Brownsville may work too, but the soil in the New Mexico desert is probably a better match to Mars.  I would expect that long before the BFS ever lands on Mars, it will have made many desert landings on many different soil types. Demonstrating resistance to debris, object ingestion and the ability of the legs to handle large rocks, etc.

Offline vaporcobra

  • Full Member
  • ****
  • Posts: 1050
  • Tacoma, WA
  • Liked: 1841
  • Likes Given: 2230
Re: BFS Design Requirements
« Reply #21 on: 02/07/2018 11:24 PM »
Something I haven't seen posted WRT to 6 and 7. Elon is CEO and Director of a certain company that almost certainly has the most mature and capable machine learning algorithms yet developed for navigation, terrain/object recognition, and minute trajectory management. Add in SpaceX's literally unprecedented experience with autonomously landing rockets, and it shouldn't be hugely difficult to implement AI, radar, and imaging that can prepare BFS' legs or safely choose a landing site during the late stages of reentry.

HiRISE imagery has basically blanketed Mars with 0.5m resolution coverage and will continue to do so until it dies, and that global imagery (stereoscopic in many cases) is almost without a doubt enough to roughly point to a landing spot small enough for onboard sensors to be able to autonomously choose a landing site in the last 10-30 seconds before touchdown.

The real issue with 6 and 7 is FOD avoidance and tolerance - that's gonna be hugely difficult to successfully integrate with goals of reliable and total reuse, as well as economically viable lifespans between serious refurb.

Offline the_other_Doug

  • Senior Member
  • *****
  • Posts: 2726
  • Minneapolis, MN
  • Liked: 1782
  • Likes Given: 3387
Re: BFS Design Requirements
« Reply #22 on: 02/08/2018 12:19 AM »
Something I haven't seen posted WRT to 6 and 7. Elon is CEO and Director of a certain company that almost certainly has the most mature and capable machine learning algorithms yet developed for navigation, terrain/object recognition, and minute trajectory management. Add in SpaceX's literally unprecedented experience with autonomously landing rockets, and it shouldn't be hugely difficult to implement AI, radar, and imaging that can prepare BFS' legs or safely choose a landing site during the late stages of reentry.

HiRISE imagery has basically blanketed Mars with 0.5m resolution coverage and will continue to do so until it dies, and that global imagery (stereoscopic in many cases) is almost without a doubt enough to roughly point to a landing spot small enough for onboard sensors to be able to autonomously choose a landing site in the last 10-30 seconds before touchdown.

The real issue with 6 and 7 is FOD avoidance and tolerance - that's gonna be hugely difficult to successfully integrate with goals of reliable and total reuse, as well as economically viable lifespans between serious refurb.

I will be honest -- I don't think nearly as much of Mars has been imaged at 50cm resolution as you seem to think.  HiRISE is a great instrument, but its highest resolution strips are fairly narrow, and Mars is still an entire planet.  With land area equal to Earth's -- just without the intervening oceans/seas.

There is likely 3- to 20-meter resolution photography covering 80% of the planet, and down to .5-meter resolution coverage on another 20%.  Enough to offer a lot of very high resolution looks at candidate landing sites, but not covering every portion of the planet.

Not even the Earth is mapped at .5-meter resolution to that large of an extent...
-Doug  (With my shield, not yet upon it)

Offline vaporcobra

  • Full Member
  • ****
  • Posts: 1050
  • Tacoma, WA
  • Liked: 1841
  • Likes Given: 2230
Re: BFS Design Requirements
« Reply #23 on: 02/08/2018 01:15 AM »
Something I haven't seen posted WRT to 6 and 7. Elon is CEO and Director of a certain company that almost certainly has the most mature and capable machine learning algorithms yet developed for navigation, terrain/object recognition, and minute trajectory management. Add in SpaceX's literally unprecedented experience with autonomously landing rockets, and it shouldn't be hugely difficult to implement AI, radar, and imaging that can prepare BFS' legs or safely choose a landing site during the late stages of reentry.

HiRISE imagery has basically blanketed Mars with 0.5m resolution coverage and will continue to do so until it dies, and that global imagery (stereoscopic in many cases) is almost without a doubt enough to roughly point to a landing spot small enough for onboard sensors to be able to autonomously choose a landing site in the last 10-30 seconds before touchdown.

The real issue with 6 and 7 is FOD avoidance and tolerance - that's gonna be hugely difficult to successfully integrate with goals of reliable and total reuse, as well as economically viable lifespans between serious refurb.

I will be honest -- I don't think nearly as much of Mars has been imaged at 50cm resolution as you seem to think.  HiRISE is a great instrument, but its highest resolution strips are fairly narrow, and Mars is still an entire planet.  With land area equal to Earth's -- just without the intervening oceans/seas.

There is likely 3- to 20-meter resolution photography covering 80% of the planet, and down to .5-meter resolution coverage on another 20%.  Enough to offer a lot of very high resolution looks at candidate landing sites, but not covering every portion of the planet.

Not even the Earth is mapped at .5-meter resolution to that large of an extent...

OOOF, I was very over-optimistic ;D It's actually probably closer to 3% of the surface imaged by HiRISE itself, with 85%+ imaged by the context camera (about 20m/px). Still, the point is that HiRISE can almost without a doubt be tasked by SpaceX, so long as it is alive. That is probably the first cooperative component NASA would likely be able to offer ahead of actual funding, and that is only necessary if that ~3% somehow doesn't include most of the viable first-base locations SpaceX would consider.

Once we get into the 2020s, I have little doubt that SpaceX will have a fair amount of experience with designing, building, launching, and operating their own satellites, and orbital imaging isn't a huge stretch beyond that point. At worst, SpaceX contracts out to the several highly experienced satellite manufacturers that exist today, and options like Planet's Terra Bella branch are also a possibility as they gain their own experience.

Getting a bit too deep in the weeds here, my main point is more simply that imagery and safe, autonomous landing and last-mile touchdown decisions are almost certainly not the biggest hurdles ahead of BFS.

Online speedevil

  • Senior Member
  • *****
  • Posts: 2278
  • Fife
  • Liked: 1129
  • Likes Given: 1255
Re: BFS Design Requirements
« Reply #24 on: 02/08/2018 11:25 AM »
Getting a bit too deep in the weeds here, my main point is more simply that imagery and safe, autonomous landing and last-mile touchdown decisions are almost certainly not the biggest hurdles ahead of BFS.

Some of the remaining issues might be greatly helped by a FH ~2020 launch of various probes.
Beagle 2 mostly landed safely (it was not debris at least, and it seems likely if you had many of them some would have worked) with 60kg aeroshell from Mars.

Impactor probes which shed much or all of their final velocity in impact at 150-50m/s or so have been planned or flown, with masses in the single digit kilos. ( https://forum.nasaspaceflight.com/index.php?topic=44853 )

A 'low risk' mission involving a simple bus which does a mars flyby and picks up entry data from a few dozen dispersed probes that may or may not work would provide a _lot_ of data on a wide area.

A day before encounter, the bus manoevers by a few meters a second and drops out impactors/very simple landers designed to have half of them fail.
Some small relay sats that aerobrake into orbit, for atmospheric studies.

Almost all of it can fail and work to some degree independently, even if you're left paying for time on the various mars relay sats.

Online guckyfan

  • Senior Member
  • *****
  • Posts: 6710
  • Germany
  • Liked: 1757
  • Likes Given: 1727
Re: BFS Design Requirements
« Reply #25 on: 02/08/2018 12:18 PM »
The first 2 scout BFS can have rovers that search for the best landing sites and remove some lose rubble, if needed. Then mark the site with radio beacons and radar reflectors for final descent. They would have to sit on Mars for 4 years before return is possible. I don't think they will ever return, so some damage to them on landing can be accepted.

Offline DecoLV

  • Full Member
  • *
  • Posts: 137
  • Boston, MA
  • Liked: 80
  • Likes Given: 23
Re: BFS Design Requirements
« Reply #26 on: 02/17/2018 09:58 PM »
I am a little surprised we have not heard about plans for a concept vehicle, especially for Mars EDL  This is what I thought Red Dragon would do this year. It could be an itty-bitty balsa wood thing (well, not literally, just making a point), but just if your little spam-covered vehicle can actually get there and land. That's been a hard enough challenge for NASA spacecraft as it is.


Online speedevil

  • Senior Member
  • *****
  • Posts: 2278
  • Fife
  • Liked: 1129
  • Likes Given: 1255
Re: BFS Design Requirements
« Reply #27 on: 02/17/2018 10:02 PM »
I am a little surprised we have not heard about plans for a concept vehicle, especially for Mars EDL  This is what I thought Red Dragon would do this year.

What does your concept vehicle do that BFS does not, and why?
Red Dragon was cancelled due to the realisation that BFS could in principle be developed in 5 years from internal revenues, and Red Dragon wasn't going to help at all with that.

Offline vaporcobra

  • Full Member
  • ****
  • Posts: 1050
  • Tacoma, WA
  • Liked: 1841
  • Likes Given: 2230
Re: BFS Design Requirements
« Reply #28 on: 02/17/2018 10:30 PM »
I am a little surprised we have not heard about plans for a concept vehicle, especially for Mars EDL  This is what I thought Red Dragon would do this year. It could be an itty-bitty balsa wood thing (well, not literally, just making a point), but just if your little spam-covered vehicle can actually get there and land. That's been a hard enough challenge for NASA spacecraft as it is.

As I understand it, Crew Dragon will give SpaceX the knowledge necessary to develop orbital, crew-rated spacecraft with long in-orbit or deep space longevity. By using Falcon 9 to test supersonic retropropulsion at points in Earth's atmosphere that are Mars atmosphere facsimiles, as well as countless experience recovering and reusing orbital Dragons, I think SpaceX believes they properly understand the flight regimes BFS will need to survive.

SpaceX likely judged Red Dragon to be too disimilar from BFS to be worth the high cost of R&D and nonexistent ROI. Instead, they'll begin serious testing of BFS prototypes as soon as early 2019, using Earth to flesh out its capabilities and fix any bugs. Red Dragon would be a distraction from the vehicle SpaceX actually needs and wants, although it would be an incredible symbolic achievement :)
« Last Edit: 02/17/2018 10:30 PM by vaporcobra »

Offline Dante2121

  • Member
  • Posts: 48
  • United States
  • Liked: 17
  • Likes Given: 57
Re: BFS Design Requirements
« Reply #29 on: 02/18/2018 12:26 PM »
Something I've been curious about what others think - I think some amount of the Merlin's iterative improvements and reliability can be traced to the fact they built so many of them. If the booster is successfully reused from the getgo - they just won't have as many chances to iteratively test changes.

With the Merlin every launch likely had a small incremental improvement to the engine. You won't have that opportunity with Raptor - which means changes will get grouped together in large sets - increasing risk. 

Offline livingjw

  • Full Member
  • ***
  • Posts: 356
  • Liked: 385
  • Likes Given: 167
Re: BFS Design Requirements
« Reply #30 on: 02/18/2018 12:37 PM »
Something I've been curious about what others think - I think some amount of the Merlin's iterative improvements and reliability can be traced to the fact they built so many of them. If the booster is successfully reused from the getgo - they just won't have as many chances to iteratively test changes.

With the Merlin every launch likely had a small incremental improvement to the engine. You won't have that opportunity with Raptor - which means changes will get grouped together in large sets - increasing risk.

But they have 38 engines per BFR/BFS, all of which will be tested. During development, testing will probably include individual engine FOD and RUD containment tests as well. I think it will be done similar to turbine engine development.

John

Online speedevil

  • Senior Member
  • *****
  • Posts: 2278
  • Fife
  • Liked: 1129
  • Likes Given: 1255
Re: BFS Design Requirements
« Reply #31 on: 02/18/2018 01:53 PM »
But they have 38 engines per BFR/BFS, all of which will be tested. During development, testing will probably include individual engine FOD and RUD containment tests as well. I think it will be done similar to turbine engine development.

A heavy emphasis on 'mass' production from the start, rather than carefully hand-building each engine might also be valuable to them in terms of consistency, and something they probably have a better handle than pretty much anyone in the industry with.

They 'know' they are going to need quite a lot of them by 2024.
At least 6 BFS, a couple of tanker BFS, a couple for earthly operations, three BFRs say, one BFR worth in testing.

That's a couple of hundred engines. Even if you discount the P2P ambitions entirely.

The cheapness of LNG also doesn't hurt if you want to do 'ridiculous' lengths of testing, with lots of restarts per engine.

Throwing bolts into the turbopump and seeing how violently you can get them to explode would be an entertaining series of blooper reels.
Throwing rocks and gravel at the operating engine similarly - what happens if you fire up the bell 6" off sand, gravel, or very large rocks, if the ambitions of rough landing are to be achieved.

Offline Dante2121

  • Member
  • Posts: 48
  • United States
  • Liked: 17
  • Likes Given: 57
Re: BFS Design Requirements
« Reply #32 on: 02/18/2018 02:02 PM »
True it's 38 engines for one complete set - but that's going to be 38 engines at once (likely with exact same design) whereas 38 Merlin engines represented 4+ sets of engines and thus four opportunities to iterate.

I think building 15 cores a year over several years has allowed Spacex to iterate a lot more than building a couple BFS per year will allow.  This allowed pretty rapid iteration - in true Silicon Valley style.

Maybe they have learned a lot and won't need to iterate as much - but that remains to be seen. If they need to iterate - it's going to be a lot more expensive to iterate BFS than Falcon.

Online speedevil

  • Senior Member
  • *****
  • Posts: 2278
  • Fife
  • Liked: 1129
  • Likes Given: 1255
Re: BFS Design Requirements
« Reply #33 on: 02/18/2018 02:25 PM »
Maybe they have learned a lot and won't need to iterate as much - but that remains to be seen. If they need to iterate - it's going to be a lot more expensive to iterate BFS than Falcon.

I'm not sure that follows, at least for many aspects.

It may be more expensive to iterate BFS/R in the sense that if you want to replace the whole fleet, yes, 10*$200M or whatever = ow.

But, other than re-usability or safety bugs, I'm not sure it largely is, for most things you may want to do with any BFS/R.

Given the rapid reusability and stackability, and more importantly, the vast overperformance of the stock BFS in the face of any plausible paying customer in the near term, there may be little incentive to replace first generation BFS that 'just' can launch 50 tons without refuelling in orbit to get down, because they're 50 tons overweight and the thermal protection doesn't work as well as later ones, and the legs are three times heavier.

Or, indeed, upgrading from ones that can 'just' launch 150 tons and are at baseline IAC performance.

About the only thing where you'd care lots about replacement is if you have a design that explodes every few launches, or if you are getting so many launches that the marginal costs of keeping a different inventory becomes important.

Otherwise, if they are safe, it just becomes a case of selecting a high performance pair if you've got a high performance launch requirement. (>100 tons to orbit, or Mars)

Online acsawdey

  • Full Member
  • ***
  • Posts: 376
  • Minnesota
  • Liked: 220
  • Likes Given: 398
Re: BFS Design Requirements
« Reply #34 on: 02/18/2018 02:31 PM »
But they have 38 engines per BFR/BFS, all of which will be tested. During development, testing will probably include individual engine FOD and RUD containment tests as well. I think it will be done similar to turbine engine development.

John

Do you expect that they will do tests analogous to the blade-out test that is done on commercial turbofans, i.e. intentionally cause a turbopump failure to test the engine controller shutdown response and the containment system? I seem to recall hearing that spacex has already done a "nut ingestion" test.

Offline QuantumG

  • Senior Member
  • *****
  • Posts: 8582
  • Australia
  • Liked: 3474
  • Likes Given: 820
Re: BFS Design Requirements
« Reply #35 on: 02/18/2018 10:07 PM »
Not a viable solution. SpaceX has already stated complete unwillingness to use FH for any Moon or Mars missions.

huh? I'm pretty sure they said "we can do it, just send money" like they do for everything.


Jeff Bezos has billions to spend on rockets and can go at whatever pace he likes! Wow! What pace is he going at? Well... have you heard of Zeno's paradox?

Online speedevil

  • Senior Member
  • *****
  • Posts: 2278
  • Fife
  • Liked: 1129
  • Likes Given: 1255
Re: BFS Design Requirements
« Reply #36 on: 02/18/2018 10:18 PM »
Not a viable solution. SpaceX has already stated complete unwillingness to use FH for any Moon or Mars missions.

They really haven't.
They have said that they believe BFS is the future, and they recommend others wait for it.

If, in mid 2019, it looks like 2022 might actually happen for first BFS landing on Mars, things get serious.

If BFS is not able to launch anything to Mars for the 2020 window, and they need more information, they either have to launch in 2022 and risk it, or they can try and get that information by using a relatively cheap launch (for them) and repurposed Starlink sats and similar hardware.

Might this happen?
I would be shocked if they've decided in a concrete manner either way yet.

Offline MaxTeranous

  • Member
  • Posts: 59
  • Liked: 46
  • Likes Given: 10
Re: BFS Design Requirements
« Reply #37 on: 02/19/2018 12:13 PM »
Not a viable solution. SpaceX has already stated complete unwillingness to use FH for any Moon or Mars missions.

huh? I'm pretty sure they said "we can do it, just send money" like they do for everything.

Indeed, they'll not pay for it off their own back but pretty sure if NASA waves a $5 billion commercial contract out there to go play on the Moon in 2020 then they'd submit FH as a proposal.

Offline livingjw

  • Full Member
  • ***
  • Posts: 356
  • Liked: 385
  • Likes Given: 167
Re: BFS Design Requirements
« Reply #38 on: 02/19/2018 03:08 PM »
But they have 38 engines per BFR/BFS, all of which will be tested. During development, testing will probably include individual engine FOD and RUD containment tests as well. I think it will be done similar to turbine engine development.

John

Do you expect that they will do tests analogous to the blade-out test that is done on commercial turbofans, i.e. intentionally cause a turbopump failure to test the engine controller shutdown response and the containment system? I seem to recall hearing that spacex has already done a "nut ingestion" test.

Yes. They already have shown a containment cover for the engine. Most likely Kevlar or similar. Safe containment and shutdown will probably be part of development. Probability of a malfunctioning engine goes up with the number of engines, and they have a lot of them.

John

Online envy887

  • Senior Member
  • *****
  • Posts: 4311
  • Liked: 2278
  • Likes Given: 1308
Re: BFS Design Requirements
« Reply #39 on: 02/20/2018 01:35 PM »
But they have 38 engines per BFR/BFS, all of which will be tested. During development, testing will probably include individual engine FOD and RUD containment tests as well. I think it will be done similar to turbine engine development.

John

Do you expect that they will do tests analogous to the blade-out test that is done on commercial turbofans, i.e. intentionally cause a turbopump failure to test the engine controller shutdown response and the containment system? I seem to recall hearing that spacex has already done a "nut ingestion" test.

They did a FOD ingestion test during Merlin dev that consisted of dropping a #10 nut into the turbopump inlet, supposedly it ate the nut and kept running fine. Haven't hear of a similar test during Raptor dev, but it would probably be good to have some FOD tolerance.

Tags: