This week NASA announced evidence for life on Mars, for the sixth or seventh time. This time it was an incredibly cool rock, and would you believe the Mars Perseverance rover already has a sample, taken July last year, banked and ready for return to Earth?
In addition to this, NASA has already been working on a Mars Sample Return mission for many years!
Unfortunately, that mission has been “paused” for almost two years after costs grew from $7b to a projected $11b, and after $3b had already been spent. Now, I love Mars rocks as much as anyone, but $11b seems like a lot. I wrote a post at the time unpacking where NASA had gone wrong. Since then, the space community has known that NASA finding a “please uncancel MSR” rock was only a matter of time.
The purpose of this post is to discuss the general problem and point a way to a better approach.
Mars Sample Return sounds simple. Collect some rocks, fly them back. Mars lacks rocket launch pads, so we should expect this process to be insanely expensive. Given that we’re going to spend billions of dollars on it, why not make sure that expense also helps to buy down risk for an ambitious program of future exploration, including sending people? The alternative would be to drop $11b on a bunch of technology we get to use only once and which doesn’t buy down any future program risk at all. But that’s all.
Here’s a diagram of the most recent reference design for MSR.
We have the rover, a rocket flying back to Earth, and some kind of lander. We also have a helicopter, and a gigantic satellite in Mars orbit. This seems pretty complicated. Why is there a giant lander and a helicopter and a satellite? Did lobbyists for Big Satellite get into this program? Check out my previous post to see how this happened.
It turns out that something like 90% of the complexity and cost of MSR is driven by a requirement that the samples “break the chain” on the way home, which is to say, perform some kind of COVID-era sterile hand off on an autonomous orbiting platform we’ll make the Europeans provide because why not?
As Elon Musk says, there is nothing more dangerous than a requirement from a smart person. OSIRIS-REx collected so much sample its container couldn’t even close, it came back to Earth, and no-one has died of asteroid dust exposure. Mars rains so many asteroids down on Earth that Steve Jurvetson owns several – any of which could contain a dormant bacteria that survived the trip in space.
I’m fully willing to concede that if it’s possible to add safeguards at a calibrated price, we should do it, but if the requirement has already killed the default mission plan, maybe we should double check the sanity of the requirement. For example, it seems reasonable that we close the airspace when the samples fly back to land in Utah. You’d have to be a very unlucky plane to get nailed by a capsule smaller than the world’s largest pumpkins, but it costs NASA nothing to call up the FAA and ask. On the other hand, it has cost NASA so much to gratify the “planetary protection” constituency that the entire mission has already failed. Doh!
So if we oriented MSR around getting rocks back expeditiously, instead of around gratifying every constituency in the known universe, what would it look like? Rocketlab has taken a crack at this question, here’s my answer.
We need only the bits that help get rocks back to Earth.
No orbiter. No solar electric propulsion demonstrator. No fleet of Mars helicopters on MSR’s budget. No Europeans (sorry!). No gigantic new lander that costs over a billion dollars to engineer and doesn’t expand downmass capacity much beyond the MSR skycrane. Yes, it’s time to do lifting bodies.
Let’s start with the return vehicle.
If we delete the orbiter, we need a rocket that can fly back to Earth. It will require at least 5.71 km/s of delta V, and more like 6 km/s for some freedom on trajectory choice. The mid 2030s are truly dismal for return trips so let’s get on with it!
Mars has a thin atmosphere so there’s no need to worry about the rocket being skinny, or needing high thrust to punch through the atmosphere on the way back. On the other hand, the samples are very durable so a high G launch back to Earth will minimize gravity losses.
Now comes the fun part. Assemble your sample Earth Return Vehicle from the Northrop Grumman Propulsion Products Catalog. With the appropriately chosen motors, a two stage solid fueled direct return vehicle can send a payload of about 10% of its launch mass back to Earth. Or, if we’d prefer the simplicity of a single stage, about 6% of the launch mass can come back, though this also places a constraint on trajectory selection.
This may seem surprising, since the default MSR architecture calls for landers, helicopters, and European orbiters, but the Mars Ascent Vehicle alone delivers nearly 4 km/s out of the total 5.7 km/s – getting from orbit to Mars escape and a trans-Earth injection requires relatively little additional fuel.
Use the upper two stages of the Minotaur V (a Star 48BV and a Star 37FMV), we can deliver a payload weighing up to 342 kg back to Earth, with a total launch mass of 3630 kg. If Starship is able to refuel in 2026 but has not yet completed its mass diet, such that it can carry only 25 T to orbit, refill, and then attempt a landing on Mars, those 25 T are adequate to carry not one but six Star 48-based ERVs. Each would be packaged within individual launch tubes, plus as many helicopters and rovers as any space lab could build by the launch window, which opens in November 2026, and then January 2029.
For reference, the Varda return capsule weighs about 90 kg, and the total satellite weight is about 300 kg, of which 75 kg is fuel and 30 kg is the return payload that sits in the capsule, which is plenty to bring back many samples. By comparison, MSR’s sample canister weighs 11 kg and contains 0.5 kg of samples.
If we insist on landing the ERV within the large rover heat shield, back shell, parachute, and sky crane, we have about 1100 kg to play with. If we put the sample can on a mass diet, the 46 kg OSIRIS-REx/Stardust return capsule should be able to fit behind a Star 24 + Star 37 ERV.
But this “conservative” approach would likely cost more, due to the demands of miniaturization, and it would have no upside or future value unlock, due to abysmal mass constraints. If we load up a few Starships with mostly COTS ERVs, then we only need one to succeed for total mission success, and even if they all fail, we will at least have run multiple ambitious Mars Entry/Descent/Landing (EDL) experiments with a gigantic lifting body spacecraft that, when the bugs are worked out of the system, will be able to transport hundreds, then millions of tonnes of stuff to and ultimately from Mars.
How about, instead of arguing over whether it’s reasonable to spend $8b, or $11b, or more on an MSR project that has zero chance of hitting the relatively easy Mars return window in early 2033 and which will teach us literally nothing about what is needed to land more than one single ton on Mars, we cut some checks to the private sector and let them cook.