An important point that I wrote about a few years ago with Falcon 9 is
that, while more engines and engine-out capability can improve mission
reliability, it can hurt schedule reliability, because you have much
greater chance of scrubbing a launch due to an anomaly with a single
engine (or however many you decide it's OK to launch with even if
showing an issue).
On 8/24/21 10:21 AM, William Claybaugh wrote:
Ian:
The trade here is cost vs. reliability.
Making a lot of smaller engines enables both production learning and rate effects to lower the average cost per engine; smaller engines also cost less to develop.
However, clustering engines requires higher reliability in order to avoid mission ending failure(s). *Proving* higher reliability means more ground testing on the nominal flight thrust profile and on certain excursions. (*Designing out* complexity is a big cost savings here, but proving the engine is sufficiently reliable requires testing…lots of testing).
The required reliability depends on how many engines out the stage can tolerate and on how much confidence one wants in the reliability (do you prefer 99% reliability at 90% confidence or 90% reliability at 99% confidence? The difference could be hundreds of static tests.)
My personal rule of thumb is to build two of everything…I have found having a spare part in inventory saves both money and time such that the cost of building two is justified. But I also found clustering to be complicated, even when using simple solid rockets. In the specific example I built, a single solid motor equal in performance to the cluster of four would clearly have been lower cost.
Bill
On Tue, Aug 24, 2021 at 9:45 AM Ian M. Garcia <ian@xxxxxxxxxxxxx <mailto:ian@xxxxxxxxxxxxx>> wrote:
From an amateur perspective, does the same logic (engineering,
financial, other?) that drives SpaceX to use many smaller
engines rather than fewer big ones, scale down?
If a simple-ish design for a small liquid engine was
developed, using modern additive or subtractive machining,
could we see a larger number of home builders put together a
bunch of engines on an Open Source platform to dial in a wide
range of vehicles without needing to design everything from
first principles?
Leaving aside the regulatory and security issues, does the
tech scale down?
That's an excellent question that I ponder myself often. As a
GNC/avionics-ish guy I always tend to push to put lots of high
reliability, cheap engines, rather than develop a big one
(assuming efficiency and thrust to weight is good enough). Just
more of the same, right? However historically this seems to be a
dead end.
(Let's say, in the context of amateur liquid rockets.)
Why is a having a single or a few large engines better than lots
of small ones? What doesn't scale down? The valves? The feed
system? Reliability should not be the main issue if engine out
capability is possible.
ian