Quote:
|
Originally Posted by frognbuff
Astra is finally dumping the "Rocket 3.3." It's too small and unreliable. Next up - "Rocket 4.0!"
|
I have a question about this, but first: I found an article that provides some information about the new vehicle:
Here (see:
https://spacenews.com/astra-reveals...-larger-rocket/ ) is a
Space News article--with an illustration--about Astra's Rocket 4 (which is part of what they call "Launch System 2.0"--a higher flight-rate, lower crew-size, easily-transportable [in shipping containers] total system), which is significantly different from Rocket 3.3. (I didn't see any more information on Astra's
https://astra.com/ "Newsroom" page, but I might have missed something from a few months back; the
Space News article itself is from May 12.) In addition:
Unlike the plans for Rocket 3.3's first stage, Astra doesn't plan for Rocket 4 to be reusable. (They're going "to MCD rather than to reusability"--MCD is engineer Arthur Schnitt's [of the Aerospace Corporation, a think tank] Minimum-Cost Design philosophy, where cost is traded along with performance when designing rockets and spacecraft.) Russian and Chinese ones are similar to MCD designs, see: "LEO on the Cheap: Methods for Achieving Drastic Reductions in Space Launch Costs," by John R. London (here's a free PDF copy of it:
http://foyle.quarkweb.com/lc93/leocheap_book.pdf ). Also:
Rocket 4 will have two larger, turbopump-fed (rather than electric, battery powered pump-fed) kerolox first stage engines that produce a total of 70,000 pounds of thrust (35,000 lbf each; no information on the upper stage engine/engines was available). The 35,000 lbf engine *could* possibly be Firefly Aerospace's Reaver engine, or a similar derivative (it was reported last year--a link is in the
Space News article--that Astra discussed licensing Firefly's Reaver engine, whose parameters are similar to Rocket 4's first stage engine. The vehicle is designed to inject a 300 kilogram payload into low Earth orbit (or 200 kilograms into Sun-synchronous orbit) for $3.95 million. Now, the question that occurred to me is this:
With the orbit-capable Rocket 3.X vehicles Astra flew, second stage separation--as it appeared in the onboard camera views--never looked "solid" and "positive" to me (as the Saturn, Thor-Delta, Atlas-Centaur, Delta II, Delta IV, Titan III & IV, Commercial Titan, Atlas V, Electron, Ariane, and Falcon staging did/do), even on the successful Astra Rocket 3.X missions that did reach orbit. It was always a slightly delayed, "cliff-hanger" moment before the fairing halves separated, even after they fixed the delay (implemented the proper delay time, that is, so that the second stage/payload wouldn't slam up into the still-"un-split" & un-separated payload fairing. The delay times, even after having been fixed, also seemed slightly variable from flight to flight, which seems to have "fit in" with something else (which is unwise to do with rockets):
From everything I've read about Astra (including comments from CEO Chris Kemp, and other company personnel), no two Rocket 3.X or Rocket 3.3 rounds were identical; they were always "tinkering" with the designs (Rocket 3.3's most noticeable change was its "stretched" first stage tankage), making endless iterations--
even after successful orbital launches had been conducted. Was this the case? If so, that sounds like an unwise way to develop a reliable launch vehicle, for this reason:
If one tests a prototype of anything, especially if it's intended for mass production (be it a rocket, car, airplane, weather balloon radiosonde, etc.), and it works as intended, tinkering with it seems foolish--*unless* telemetry and/or physical inspection (if possible) reveal(s) that one or more parts or systems very nearly failed during the test run or flight. Now:
(A famous example of NOT doing this--which resulted in high reliability--was the development of the Thor-Delta launch vehicle. Douglas used already well-proved components--the Thor first stage, the Vanguard second the third stages [Able and Altair, respectively], the Thor-Able bulbous nose fairing [a cone-cylinder fairing, for narrower payloads, replaced the hemisphere-cylinder Thor-Able fairing], and the AC Spark Plug Titan I guidance system--in the Thor-Delta, only making changes--and usually only one at a time--when they gradually upgraded the vehicle for higher performance.) But they flew each uprated variant a number of times, to be certain it was fundamentally sound, before incorporating the next upgrade,
BUT:
Astra's vehicles, having precious few previous flights to draw upon for experience (and for making meaningful reliability computations; their "sample size" was essentially *one*...for each iteration, making such reliability calculations nearly meaningless), were the least "well-placed" for doing such iterations on. If they continue this way of doing things with Rocket 4, they will likely have the same problem; when each round is a "one-off," figuring out--even with telemetry--what "failure chains" are likely, and why--will be much more difficult than "freezing" (at least for a while) a design that works, and flying it numerous times to prove its reliability. (Douglas [and later, McDonnell Douglas] even investigated why *successful* Thor-Delta and Delta II flights worked, to find and fix [or beef-up, as appropriate] any marginal parts or systems that hadn't failed, but could easily do so, if they weren't lucky.) Plus:
I remember Astra CEO Chris Kemp comparing their launch vehicles to computer software (he came from a Silicon Valley background, if memory serves), which is behind their philosophy of "continuous iterations" of their rockets' designs (in total, they've built--and static-tested, and launched--several), from their original Ventions design to Rocket 1, Rocket 2, and Rocket 3 (in all of their iterations); the launches took place from our Pacific Spaceport Complex - Alaska (PSCA) on Kodiak island. Tinkering with computer programs is one thing (even the most spectacular Kerbal Space Program launch pad explosions and in-flight failures cost nothing), but doing the same with even the smallest orbit-capable *physical* rockets involves expending money, time, and materials, and creating some risk to company--and even flight range--GSE (Ground Support Equipment) and--in the latter case--the range's buildings, cranes, tracking facilities, etc.! :-)