[AR] Re: Congratulations to USC for Traveler IV launch
- From: Henry Spencer <hspencer@xxxxxxxxxxxxx>
- To: Arocket List <arocket@xxxxxxxxxxxxx>
- Date: Sat, 25 May 2019 13:57:33 -0400 (EDT)
On Sat, 25 May 2019, Bruno Berger SPL wrote:
Starting with a solid motor (eg a commercial one) is IMHO not wrong. So
you get your experience with the airframe, avionics, telemetry, recovery
etc before you maybe switch to liquids. The chance is great that you
will never fly something if you start with the development of a liquid
engine first...
The flip side of this, though, is projects that decide to fly first with
solids and then switch to liquids, but die when they try to make that
transition.
Whether starting with solids makes sense depends on your goals. For sure,
getting something liquid working is a big hurdle. But if you *must* jump
that hurdle to achieve your goals, flying with solids first might not be
the best use of effort. For example, if your goal is a system that does
DC-X-style powered landings, then learning how to do reliable parachute
recovery (which is harder than it looks, especially given that almost
every failure means building a new rocket) might be a costly distraction.
Part of the problem is that there's a hidden mistake here, lurking in
Bruno's own wording. Quite likely it *is* a mistake to start with
development of a liquid *engine*, if your goal requires development of a
liquid *propulsion system*. With solids, the engine pretty much is the
whole propulsion system. With liquids, not so -- not even close. People
seldom do post-mortems on failed projects, but I think a lot of them would
show that the project died between "engine drawings complete" and "first
successful firing", because getting the *rest* of the propulsion system
working was planned to take a weekend, and after two 14-hour days they
realized they were only getting started.
People who want a liquid propulsion system should not start, as so many of
them do, by trying to design their ideal liquid engine. Better to start
with the simplest and crudest engine possible, with no intention that it
will ever fly, to get experience with plumbing, fluid handling, controls,
ignition, plumbing, data acquisition, test operations, plumbing, etc. :-),
before you maybe switch to a more sophisticated engine design. Not least,
because that experience is going to change your ideas about what the more
sophisticated engine should look like.
for student projects it's important that you see results soon to keep
them motivated.
Agreed, and in fact I would omit "for student projects" -- seeing results
soon is important for any volunteer effort, and it sure doesn't hurt for
investor relations either.
For low-budget projects with an inexperienced team and nervous sponsors,
the dinospace dogma of all-up testing -- build and fly the complete final
system the first time, and of course it will work -- is a snare and a
delusion(*). Think incremental development and testing instead.
(* In fact, if you look at how the all-up concept developed -- most
notably, its use for Project Apollo -- the people involved didn't believe
for an instant that the first test was certain to succeed. They tried for
it, they hoped for it, and they were prepared to exploit it if it
happened, but they were far from sure of it. The "and of course it will
work" part got added later, by the ignorant and overconfident. )
Henry
Other related posts:
