Landing engines for Lunar Starship - powered by Tesla?

Major problem for lunar Starship design is actual lunar landing. Last mile problem of landing with bottom positioned engine causes huge debris problem. There are three possible solutions (that can be combined):

1. powered landing using Raptors and hope for the best
2. terminate thrust above the ground and use landing legs for impact absortion (high stress risk)
3. have separate landing engines (additional dead mass)

Separate landing engines are missing component of Starship architecture that could have significant impact on the timeline. Empty lunar Starship will have 100 tons. Plus 100 tons of cargo. Plus 100 tons of fuel for ascent to LLO. That is 300 tons that needs to land gently. Which equals to 500kN downforce. More than a single Raptor can throttle. For redundancy, we can expect more than one engine on different axis. On the above image, there are nine engines put into three separate groups, which brings single engine around 60kN thrust. 

SuperDraco engines at Hawthorne

They will probably require some angle, which provides thrust authority in all dimensions. Lets say 30 deg angle, that gives just 14% cosine loss. Not bad. And lets assume they will have similar dimensions like SuperDraco. SuperDraco consumes 30kg/s of NTO/MMH. And it has 71kN of thrust at sea level. Close enough. It is already developed, fully tested and can be directly applied to Starship. Providing 30 seconds of full thrust requires 8100kg. Plus tanks. Plus engines. At least 10000kg for lunar landing only. And the system cannot be used for Earth landing. But it does add a lot of dead weight, with completely separate high pressure tanks, feed lines and engine components. That is how old space companies would approach the problem. 

Is there a better way? Elon Musk indicated it would use existing methane and oxygen as propellant. Reusing the same tanks. That saves one or two tonnes. Pressure-fed SuperDraco version using metholox would due to lower density of methalox use just 20kg/s. Thus the thrust would be (with the same expansion ratio) around 53kN with Isp of 270s at the sea level. So around 60kN in vacuum. Add nine of them and you can easily hover and land on the lunar surface. And if you can land with cargo....you can take off without cargo. So it is a simple solution for addressing landing and ascent issues. 

I am aware that such engine would have to have super reliable igniter, since metholox is not hypergolic. Luckily, Raptor has the same problem, just on larger scale. So such engine would be a scaled down Raptor without nozzle cooling and turbopumps. It would be pressure fed. These engines are definitely not RCS engines, but they can still have many additional uses. During Earth landing they can provide useful additional thrust providing better control authority against wind. They could also be used for landing on Mars (since the increased gravity would be compensated with lower mass without ascent propellant). During actual touchdown they can reduce stress on landing legs during Raptor shutdown. During launch they could address wind shear forces. During atmospheric entry, they could affect hypersonic and supersonic shockwave formation. Or augment or even substitute airflow control surfaces. So such as system has variety of uses beyond lunar landing support by itself. But it is not the same as RCS. 

Real issue with pressure-fed engines like SuperDraco is really usage of high pressure tanks. Main tanks are designed to operate between 3-6 bar. SuperDraco expects 150 bar pressure. And experience with Falcon 9 has shown that cryogenic high pressure vessels are very risky. So what could be an alternative? Smaller turbopumps? It requires completely new engine design and testing, which represents major schedule risk. Plus turbopumps have issues with slow spinup and throttling response, making them less than ideal for RCS like applications. 

RocketLab Rutherford has shown that electric pump engines can be highly reliable and provide much better startup and shutdown control. And we know for a fact that Starship will utilize Tesla batteries (at least 200kWh) and have solar panels. Plus Tesla has expertize in designing high performance electric motors. For example, existing Tesla Model S battery can power 515kW engines. And battery day has shown tabless design of new 4680 cell that will probably have much higher power output. So 200kWh battery could easily be adapted to output 2MW or more. It would not be additional mass, as the same 2500kg mass would have other uses (to power electronics during ascent, occultation events, main propulsion, propellant transfer, controlling fin actuators during reentry and landing etc.). So the thrust-to-weight ratio of such engines would probably be better than Raptors. Especially because they would not have large expansion ratios or high pressures like Raptor.

If we use Rutherford engine for comparison, 2MW could power 18 engines, giving combined 468kN of thrust in vacuum. Quite close to our target of 500kN of thrust. Of course there are many variables that influence the numbers (metholox has higher 4% higher specific impulse, but 20% less density than kerolox). But it should be much easier to design electrically powered pump than a turbopump. SpaceX can easily scale number of engines and battery power density to address desired thrust level. 

How will these rocket engines be sized? Starship will definitely use Tesla derived batteries and electric motors, so it makes a lot of sense to use the same technology to pressure-fed metholox engines. Metholox ideal oxidizer to fuel mass ratio is approximately 3.5. But their density ratio is 2.7. Since pump power roughly depends on liquid flow, they will probably use the same electric motors as pumps. 

Electric motors used to drive fin actuators could have second purpose to provide pump power for pressurizing fuel and oxidizer to the pressure-fed metholox engines. Only drawback might be location of the pumps and engines, so it might be a design challenge to use the same motors due to different locations on the starship. It is much easier to provide electrical wiring and have separate electric powertrain for fin actuators and landing engines. Since 300kW model S engine has less than 35kg, it might be easier to simply put additional engines. To provide redundancy and easier cooling, multiple low power engines (two per each rocket engine) could be used.

This approach would greatly leverage existing Tesla expertise. And we have witnessed that Elon Musk really likes to crosspollinate expertise from these two companies (Roadster SpaceX package, Cybertruck stainless steel alloy). And SpaceX buys standard components from Tesla. So it is highly likely that the lunar landing system will be powered by Tesla components. Once implemented, it can be used for Mars and Earth landings too, as standard component of Starship.