Why use a Mars orbital Earth return vehicle for sample return?

Question

NASA's plan is to use two separate missions to bring home the drill core samples that the Perseverance rover will take. The samples will in plan be picked up by a landed mission and launched to Mars orbit. There the samples will dock with another spacecraft that brings them to Earth.

Why not instead dock two launches in Earth orbit, so that enough mass can be landed on Mars to both pick up the samples AND bring them home? For example, first launch a fully fueled upper stage that can bring the thing to Mars, then launch the pick up, ascent and return vehicle to dock the two. If there's a failure in that process, a new attempt can be made without losing the samples that Perseverance will spend several years to collect.

Docking in LEO has become routine with hundreds of successes and I think only two failures in history. But an autonomous docking in Mars orbit has never been tried. Why try to do it the hard way?

Answer 1

Earth Orbit Rendezvous is a method for applying brute force. Mars Orbit Rendezvous actually improves efficiency, potentially by a lot.

A Mars sample return (or, for that matter, a straight-up crewed mission to Mars) needs to do the following in order:

0. Launch from Earth (*)

1. Get on transfer orbit to Mars (*)

2. Land on Mars with an ascent vehicle ready

3. Take off from Mars and get into, at an absolute minimum, Mars orbit (*)

4. Get on transfer orbit to Earth (*)

5. Either land on Earth or be recovered in Earth orbit.

The starred items absolutely require a significant amount of delta-V -- any vehicle that does them, with present or near future technology, will be mostly fuel and will have a mass several times as large as their payload, at a minimum. This is the tyranny of the Tsiolkovsky Rocket Equation.

Items 0 and 3 absolutely need a high-thrust rocket, which regrettably today means chemical fuel, which means abysmal specific impulse and enormous boosters in relation to the payloads.

Earth Orbit Rendezvous provides a method to assemble a very, very large spacecraft in Earth orbit when you don't have a booster large enough to launch it in one piece. This implies at least two expensive heavy-lift booster launches, and we have some pretty heavy lift boosters today, and will have some even heavier ones (Starship, SLS) in the future.

A slight complication is that when sending unmanned spacecraft to other planets, we often use the booster to provide the initial impulse into their transfer orbit (the payload is a lot less than the booster's max payload to LEO) rather than having them fly from LEO to their destination under their own power. However, this is mostly just an operations detail.

But lets look at that list of maneuvers. The rocket for getting on a transfer to Earth (#4) is something fairly substantial (at an absolute minimum, several times the mass of the payload delivered to Earth), and it's not needed on tasks 2 and 3.

If we launch into Mars orbit, and then jettison the Mars ascent vehicle and dock with an Earth return vehicle, then both the Earth return vehicle and the Mars ascent vehicle may be roughly equal weight: several times that of the payload going to Earth.

If we make a staged rocket to launch into Mars orbit carrying the Earth return vehicle, and then light off the Earth return vehicle to go back to Earth, the Mars ascent vehicle needs to be several-squared times that of the payload going to earth. That's a LOT more -- especially if "several" is more like 10 than, say, 3.

You can bet that making this rocket larger and heavier will also complicate landing a great deal, and will cost for getting it all to Mars in the first place.

Answer 2

In this question I suggested, that, on the face of it, the SS520-5 sounding rocket could get about 5kg from the surface of Mars onto a trajectory to hit Earth. That rocket (a three stage solid fuel rocket) masses 2.6 tons at liftoff. Let's make the rather large assumption that I haven't missed any problems (someone has already suggested the cold as an issue). Then you would also need to land some kind of lightweight version of the launch rail and associated systems that are normally used with an SS520. On Earth that's a truckload of stuff, but let's be very optimistic and say we can get it down to 1.4 tons, making 4 tons that we need to land on the Mars surface.

Working backwards, we need to wrap that in a heatshield and landing system. For Curiousity, that system massed about 3 times the mass of the mass of the lander, so let's, again optimistically -- landing large things on Mars is hard -- go for the same ratio, so we need 16 tons launched into Mars Transfer orbit. This is actually just within the capacity of a Falcon Heavy, a rocket that was not available when these missions were designed, but there's not much to spare.

Working forwards, our 5kg payload has to manage any trajectory corrections on its way to Earth, communications and power for the cruise, a reentry at escape velocity, and some kind of landing, as well as payload, all in a mass budget about 1/3 of what the MarCo cubesats had.

So the mission profile you designed isn't actually against the laws of physics, and doesn't even need an Earth Orbit Rendezvous, although that might buy a few more kg, but it would need new and untried technology to land a multi-ton spacecraft on Mars and then to build a 5kg payload capable of getting anything useful back to Earth intact. Mars orbit rendezvous looks a lot less risky to me.