>>12168003The interesting thing about the Raptor is that unlike the Merlin which went from 1B all the way to 1D, that allowed it to massively improve thrust performance, this new engine has most of the optimization built into it. That said, as they're doing testing with hopper hops, bopper pops, static fires, and continuous improvement to material designs, sciences, and just engineering the engine over and over again; there's still enough optimization potential available to make a Raptor 1A and take it all the way to the upper ceiling of 1B or lower floor of 1C.
If they do this, then they don't need 37 engines. If R1A = 1.00, and R1B = 1.25 and R1C = 1.5, then the floor of 1C would be ~1.3 or 1.35. Let's pretend that they manage to go from 1A (start of development) to 1B by time Super Heavy flies and 1C (floor) by the time Mars missions are ready to go (2024-2025). Then, Gen1 SH MAY need only 29 engines. Gen2 SH (floor of 1C say 1.3 to be safe) means that they MAY only need 28 engines. The central 7 engines are a must. On top of that I'd imagine they'd have two vertical pairs per legs (pairs for redundancy on engine out) for another 12 engines (which brings us to 19). Then they can go around the inner most ring and do every other engine for a set of 6 engines, which brings us to 25 engines. and then you can for the upper and lower forks of the leg configuration, add 2 engines in the skipped areas, to get to 29 instead of 28.
I think that SpaceX will always maintain an odd set of engines in the configuration to allow for the possibility of an engine out on the booster. F9's configuration allowed for up to 2 engine outs with successful payload delivery (but 2 engines failing meant that the probability of booster recovery was nil). Raptor is designed for massive reuse. We're talking dozens if not hundreds of ignition, shutdown, and re-ignition. So the stresses on it are much greater. Nonetheless, redundancy is important especially for the booster ferrying people.
