From OpenLuna
Jump to: navigation, search

I attempted the design of a rolled steel pressure-fed booster called Greenstar at this scale, which didn't work out because of total thrust pressure problems and taught me that "big dumb booster" is a pick-any-two situation as a result (btw, the same total thrust pressure issue explains the gracefully tapered shape of the Soyuz booster, derided by some as a result of "crude" engine design.) The next generation booster design (very incomplete) is Lilmax, a "big practical booster" with modular configuration similar to Falcon 9H, but bigger. The features planned are something like this:

- The turbopump engine is borrowed from the next smaller booster in the Ascent Roadmap series, Kilder, and itself has these features:
-- oxykerosene, approximately 2.40 mixture ratio and 320sec vacuum Isp, 290sec sea level (I don't have my AFAL numbers on me atm)
-- turbopump impellers are dual-open-faced Barske impellers based on the Sundyne AnySpeed pump ("AnySpeed" is a Sundyne registered trademark) with on-shaft eye inducers
-- turbine will probably be an 100% impulse turbine so we can have thick blades for active cooling and long engine life.
-- cycle undecided, but open cycle gas generator is most likely
-- "flow liner" style regenerative nozzle with non-stress-bearing channel liner relieves nozzle hot wall of thermally induced stresses, reduces stress cracking and leads to much longer engine life
-- requires a relatively high inlet NPSHR due to high speed of turbopump and on-shaft impeller (i.e.: lack of boost pumps such as on SSME and RD-171.)

- The first stage module has four engines, the upper stage has one with an extended nozzle (it is hoped that with a guided vernier-powered separation, the nozzle extension can be designed to catch on the core stage forward skirt for rapid and simple extension, and reduce the staging coast to 8 seconds.)

- The tanks of the first stage are simple, cold rolled steel (like a submarine), tang-and-clevis segmented construction, of the widest diameter that is logistically practical (probably 4.0m. using a specially designed highway trailer) with no integral stiffening. If the tanks are unstable without pressure, they will be "jigged" when unpressurized, internally (using an insert) for transport, and externally for assembly (find a picture of an Atlas assembly line to see how this works.)

- The tang-and-clevis joint envisioned is similar to the one that famously failed in SRB BI-26R's aft field joint, destroying the Shuttle Challenger and her crew on STS-33. Even with an identical joint design on Lilmax (unlikely, since we have lessons learned from STS-33), the burn-through type failure is impossible because:
-- No hot gas involved, just liquid fuel and oxidizer; a leak will not lead to ongoing erosion as in a flame leak
-- No ignition transient. Challenger's joints failed (BI-26R's forward field joint burned through after the stack broke up) because the booster casings were pressurized so rapidly (<600ms) that the O-rings did not have time to respond and did not seat properly, leaving the sealing job to an unreliable insulation putty, the reason why Challenger didn't blow up on the launch pad. Lilmax' tanks will be pressurized over a period of several seconds.
-- Obviously Viton O-rings are not an option, since they will be brittle at LOX temperatures and flammable in the presence of high pressure GOX likely to be present at pressurization. Soft metal gaskets, compressed during assembly, placed in joint areas assured to be compressed as a result of joint rotation, will be used instead. These gaskets are in different places than the Shuttle's O-rings due to the lessons learned from Challenger
-- The "capture feature" on the RSRM joint is not required

- The tank pressure is about 200psia, very high for a turbopump booster, but well below that of pressure-feds and solid boosters (i.e.: SRB is 950psia.) MEOP is 200psig, test pressure is 300psig, and burst pressure is 400psig. These are standard aerospace safety margins (although lower ones have been used.) It may be possible to sell non-flight tanks as ASME Sec XII or Sec VII/2 pressure vessels with a MEOP of 130psig. This could provide additional revenue and production experience.

- The high tank pressure provides the strength needed for recovery impact...a landing on desert rock is assumed.

- The booster's recovery parachutes are carried in a piggyback pod (find a picture of Titan III or Titan IV and look for the strap-on booster's TVC nitrogen tetroxide tank to visualize how this will be installed.) This allows the booster to land nose first to save the engines.

- Airbags? Probably. This will need to be traded with the tank's impact stress life and the size of the parachutes. It may be that the booster will not require nose airbags, but will require airbags on the aft skirt to protect the engines from the post-rotation impact. With the parachutes on one side of the booster, the initial impact will not be perfectly vertical, so rotation will be predictable.

- Reuse is really easy with land recovery, and somewhat harder with ocean recovery. Describing land recovery reuse:
-- Drive a purge truck up to the booster (could well be a Big Eagle truck from Calgary, to illustrate the off-the-shelf possibilities), vent the booster to the minimum transport pressure (probably about 20psig), and then while maintaining minimum transport pressure, purge the tanks with dry nitrogen to safety them. Venting the tanks to transport pressure will probably be done remotely or automatically. Gas monitors will ensure the booster is safe enough to approach prior to the purge. If it isn't, one of those big fans used to create fake storms on movie sets should be enough to make it so. (NASA has these for the Shuttle Orbiters on the NASATV long enough after a landing to see them. Also, oxykerosene propellants are orders of magnitude less likely to create a dangerous situation than the Shuttle's toxic hypergols.)
-- Pyros will be studiously avoided during Lilmax design for the precise reason that they are more dangerous during ground processing and recovery than are SADs (Spring Actuated Devices.) Safety switches and long poles can be used to discharge these safely, and they can be reused. If a faulty redundant SAD is really jammed that the safety switch doesn't work, a few well aimed rifle shots should be able to discharge it safely, although the resulting damage to the booster could get expensive. The danger to avoid in this situation is providing a spark for a flammable atmosphere. Of course, a dead non-redundant SAD would, at the very least, render the booster non-reusable.
-- The booster's recovery trailer will pick it up like a roll-on bin and take it back to the launch site.
-- The booster is pressure-tested, checked for leaks at MEOP after a few minutes at test pressure (this test uses water.)
-- If it passes, it is simply refuelled and launched again
-- If it fails at a joint, it is disassembled and reassembled, as most likely the gaskets are worn out. The proof test is repeated
-- If it fails away from a joint, repeatedly fails at a joint, has reached the end of its service life, or is buckled, fractured, etc., it is disassembled, the gaskets are scraped off (if necessary), and it is cut up and sold as standard issue no. 2 ferrous scrap.

- The upper stage is more sophisticated:
-- It is expendable, high strength steel, welded construction, with a single engine (as the program matures, ones flown in first stages will be used instead of new ones), a set of pressure-fed verniers for separation, roll control during powered flight, LEO circularization and deorbit.
-- Late during its operation, the undersized pressurization system will run out. The trottleable engine's thrust will be reduced so that its NPSHR will remain below the tanks NPSHA. This also eliminates burnout accelleration problems with light payloads (a big deal with solid fuelled upper stages like those on Delta II, Taurus, and Pegasus.)
-- It will use a hydrogen peroxide RCS system...Soyuz still uses one that lasts six months, so it's obvious that one will be viable for a few hours. This will be a heck of a lot safer and cheaper than standard issue hydrazine.
-- Pressurant and RCS tanks can be had, probably off the shelf or modification of existing designs, from Pressure Systems Incorporated. The first stage will probably use commercial grade pressurant tanks like Dynecell (from Dynetek), and do not require RCS.
-- The guidance system can probably be knocked together from movie motion capture technology and Z80s (processors common in embedded systems and old video games)...if the Ascent Roadmap is followed in order, it'll date from the much smaller booster Symtex. Triple redundancy should be enough. The bugaboo of space computers is latchup from solar and cosmic radiation. The Lilmax guidance system will only be exposed to these levels of radiation for a few hours. Some get to Earth's surface, which means consumer electronic devices should occasionally take hits that cause latchups. I use consumer electronics quite a bit (my cell phone is on constantly), and encounter crashes that might be caused by radiation-induced latch up every couple of weeks (my cell phone has never crashed!) As with Shuttle (which has the smartest rocket guidance system flying, except maybe Falcon's, which I don't have the details of...made of IBM AP-101S computers much older than the Z80, which, despite its age, is still in production...or at least more recent versions of it are.)
-- The payload adapter will use a motor driven Marmon clampband opener. These are common on all launch vehicles except the Delta II (which seems to be behind in a great many areas! It also has the dumbest guidance system on any booster currently in service...except the Delta IV, with which it is tied.)
-- The upper stage can be stretched for bigger versions of Lilmax.

- The booster fairing is expected to be relatively expensive, since it must be sealed prior to the rollout of the payload, and then must separate without generating any crap that might gum up cameras or other sensors on the payload (this happened to Mars Pathfinder's sun sensors...until the program was recalibrated, it was so lost only the solar cell currents assured mission controllers she wasn't going to die. The booster was a Delta II 7920 with the 2.95m aluminum biconic fairing.) It will also be wider than the booster segments, meaning it will need to be shipped by special aircrafts and other vehicles.

- separation of the stages will be accomplished by SADs

- Using Lilmax modules only, three first stage configurations are possible (1, 3, and 5 first stages), giving payload options from 20 to 60 tonnes LEO or LEO equivalent.

- For 3 and 5 first stage configurations, crossfeed of booster propellants will keep the core first stage fully loaded until the first set of strap-on modules deplete. On 3 module configurations, there is only one set; on 5 module configurations, there are two. The core stage will throttle down after the first set on a 5 module configuration depletes. The core stage of a five module configuration might have a tough time during recovery, as it will wind up thousands of miles downrange, probably in deep water. Carbon steel hates salt, and chartering a blue water ship big enough to recover it won't be might be better to sell it to a salvage company prior to launch!

- The dry mass fraction is expected to be about 12% for the first stage...really heavy by orbital booster standards, but adequate for lower stages. This results from cost/mass tradeoffs on systems and materials: pyros vs. SADS, integral stiffening vs. high tank pressure, turbopump design vs. high tank pressure, expendable vs reusable, low density of oxykerosene relative to solids, segmented construction to ease logistics of factory-to-launch-site transport, crossfeed plumbing, serial vs. parallel arrangement, aluminum vs. steel, fancy heat treatments and integral stiffening vs. cold rolled steel and high tank pressure, etc..

- The dry mass fraction is expected to be about 8% for the upper stage...about the same as the Soyuz core.

- The crossfeed plumbing will have a substantial impact on the tank design, so there wind up being three versions of the first stage: one that has no crossfeed plumbing (used as core for 1 module, and second set strap-on for 5 module), one that has crossfeed out (used as strap-on for 3 module, and first set strap-on for 5 module), one that has crossfeed in on two opposite sides (used as the core for 3 module and 5 module.) Except for the tank nozzles and crossfeed manifolds and pipes, these version will be substantially identical.

- The forward skirt of any version first stage can accept either an interstage or a nose cone. These will be made as simple as possible since they will probably be written off by impact. (The nose cone is likely to be simply fibreglass...technically glass reinforced plastic (GRP)..."fibreglass" is derived from the "Fibrglas" trademarked insulation.)

- stay frosty: the LOX tanks are uninsulated...probably (trade LOX boiloff during longest stay-tanked scrub turnaround...and its likelihood vs. the expense of adding some spray-foam. In the spray-foam scenario, it is fine if it comes off since there is nothing to hit that is weak enough to be damaged by it.) The spray foam application is unlikely, since it will be relatively easy to unload the booster propellants and reload them during a scrub turnaround lasting more than an hour.

That's about all there is on Lilmax at the moment. The pressurization system is a major open item.

Personal tools