>>10843853One interesting comment from an old Ars article
"
SpaceX got tremendously lucky - the design compromises they were forced to adopt ended up benefiting them in the long term. Allow me to expand.
Rockets are, in general, over 90% propellant by mass. When you launch, you need enough engine thrust to counter all that weight and impart some acceleration on your vehicle. However, if you try to land your rocket after boosting the payload, you have expended most of your mass, and your launch thrust is wildly excessive - if applied fully, you will produce tens of g's of acceleration, destroying your precious vehicle. You need to somehow reduce it to a small fraction of design maximum - but rocket engines are, in general, really bad at deep throttling, as it can cause combustion instability, followed by flow separation, followed by a destroyed engine.
This is where SpaceX got lucky - initially, their plan was to launch small satellites on the Falcon 1, followed by parachute recovery of the first stage. To that end, they developed the Merlin 1 engine for the first stage, and Kestrel engine for the second stage. This plan failed, and nearly bankrupted the company. They were saved by NASA awarding them a contract to develop the Falcon 9 and Dragon. With only Merlin and Kestrel in their toolbox, the latter being far too small for the class of rockets they got contracted to work on, and limited funds, they opted to make the entire rocket work using their relatively small Merlin 1C engine, which, at that point, was rated for 42 tons of thrust at sea level and 48 tons of thrust in vacuum. For comparison, the RD-180 engine used on Atlas V is rated for 383 tons of sea level thrust.
"
