Deep Space Questions Episode 6 - Radiation Hardened Computers, Propellent Depots, Scary Rocket Fuels 

"launch profiles, that's a boring subject" Scott, we're watching video about rocket science, no part of it is boring. Especially the parts that may lead to "you won't go to space today" scenarios.

"Chemistry can do neat stuff" - Scott Manley 2021

The fuel topic made me laugh. When I was a kid in the late 70's, my Titan II ran on Aerozine 50 and unsymmetrical-dimethyl-hydrazine. Standing under the engines of fully loaded Titan II knowing there were thousands of gallons of liquid over your head that would melt you in seconds was uber sobering. Also, the knowledge that it was hypergolic and only separated by thin aluminum was another thought to ponder. Never mind the 9MT warhead on the upper part. It was a fun job while it lasted and I was only 25 when I left after three years.

For the "greatest" liquid propellants one should get hold of a copy of "Ignite!" by John D. Clark. Some of those mixtures would make even Scott's hair curl

I see FOOF in the thumbnail and know this is going to be an energetic video. It’d be a very good rocket oxidizer if it didn’t threaten to oxidize the entire rocket.

Regarding Lagrangian points, you're forgetting that they orbit along with the smaller body. L1 is NOT where Sun's and Earth's gravity balances out. It's point where Earth's gravity reduces Sun's gravity just enough so you can orbit Sun at orbital speed with angular velocity equal to Earth's. Similar thing with L2.

Also on Titan you could have a camping stove with a O2 bottle as your “fuel” source.

Lol. I never thought I’d get poop jokes on this page. Gotta say. I’m not disappointed. 🤣

Nitpick: The 14/10/7/5nm designation does _not_ refer to a specific feature size. It is used to express a ~2x increase in transistor density (sqrt(2) reduction in side length => 2 reduction in density), but there are different types of transistors and non-transistor features on a chip, so the terms end up being mostly marketing for the different companies.

15:40 The great book Ignition! includes a story of him getting a request from higher-up to try firing dimethylmercury in an engine (with RFNA), but when he called his supplier to ask if they could make him *a hundred pounds* of the stuff he heard a "horrified gasp" from the other end of the phone line and the supplier politely declined and hung up. Disturbingly, the idea of using mercury in an engine (to achieve high impulse density, for situations where physical space is more at a premium than weight) did not end there. To satisfy the request he wrote up a concept using liquid mercury injected into a liquid monopropellant engine, expecting to "see it bounce back in a week with a "Who do you think you're kidding?" letter attached". It didn't. They were directed to proceed with experimental verification and they now had to figure out how to test-fire this horrifying thing while collecting and scrubbing the exhaust to avoid giving everyone in the county mercury poisoning. In the end, before they got around to test-firing it the Naval Air Rocket Test Station (where this was taking place) was shut down and they were ordered to ship the setup to the Naval Ordnance Test Station (NOTS). "With a sigh of relief, we complied, and handed them the wet baby. Saved by the bell!" In the end NOTS did in fact fire it (in the middle of the desert, without bothering with exhaust scrubbing) with a bipropellent (UDMH+RFNA) rather than monopropellent engine, and it performed close to theoretical calculations. The project was declared a technical success, and never touched again with a ten foot pole.

Scott Manley: I don’t see Jupiter as a place humans can colonize Issac Arthur: Hold my drink and snack

8:15 Sound like a perfect idea for a time capsule. Take some things, put it in a can, seal it and send it on a long, elliptical orbit through deep space/outer planets. Presumably on a safe orbit that would leave it linking up with Earth in like 500 years time. If we really wanted to, we could give our descendants quite a shock by making it look like some kind of space ship from a lost civilization. Of course keeping it simple will probably work just as well.

This guy says more interesting & sensible stuff about rockets / space than I've cumulatively thought in my (over 50 year) lifetime. Thanks Scott. Keep it up.

You do this Q&A stuff well. Various topics with lots of info. Thank you.

A surprisingly large amount of military/aerospace equipment still relies on old-fashioned 54 series (military version of 74 series) individual logic gates and analog components in their control circuitry, without using microprocessors at all. Part of this is carryover designs from decades ago, but also because a lot of those components are available 'easily' in rad-hard versions. It's a fun topic to dive into, and it's interesting how far they will go to avoid microprocessors when dealing with high radiation environments.

You can tell Scott is having so much fun recording these episodes and talking about this. I love it. Nerdiness rules!

"The Clouds of Titan". Sounds like a Niven or Heinlein story.

I would love a deep dive on JWST orbit mechanics. I learned this week that it is actually orbiting the Lagrange point rather than resting in it and I would love to understand how that works.

Fyi the 5nm size is a bit of a misnomer, most modern architectures are based on a 14nm process that is stacked into multiple layers to give what is effectively a 5nm transistor size when all of the transistors are placed directly next to one another on the silicon die. The actual smallest features are thus still 14 nm. I'm not an expert on this but Dr. Ian Cutress (TechTechPotato on RU-vid) does a great job explaining this.

Hi Scott I am always really excited for when you upload

Accelerando by Charles Stross is one of my favorite books. I'm glad to hear him name dropped.

Wow, I was going to suggest “A Tall Tail” to you - thanks for covering it!

if it's 200 vs 5 nm feature size and it does work like you said, then it's 40x not 400x, and you do get 1600x Scott basically did two math mistakes that cancelled out LIVE, now that's the kind of things i like

Great Scott, I wanted to quickly thank you for your work on YT. You reignited my interest in space when KSP was gaining traction. You also inspired me to push my research career (what a dead end that is!) to a more programming-oriented professions. Thank you.

More likes than seconds since release. That's good to see!

Fantastic episode Scott! Love this occasional format. Is flammability of the environment on Titan a concern? Lakes of liquid ethane or methane and their vapors seem like risky objects to fuel and ignite propulsion systems near, but then again between the cold temps and lack of oxygen perhaps it's not a concern? Thanks so much!

This was a particularly good one, Scott! Love your work.

I really like these viewer questions videos Scott, more please!

I'd love an episode with more in depth info on computer hardware in space!

Jupiter Balloons: You could use balloons on Jupiter, since the atmosphere has a good fraction of helium in it, pure hydrogen would float. Though it would be a gigantic balloon. Having read Ignition like most people here, FOOF/acetylene seems a good backup for a crazy combination. High performance, easy ignition, destroys your rocket from normal use so no one can steal it...

I'm really enjoying these videos, educational and entertaining. Keep it up!

I don't know why but I really like the way you explain everything, even if I don't have enough science background to understand it all, it's interesting and I think I get a little more from each video.

Dioxygen difluoride - One of the few chemical composition onomatopoeias.

You misspoke when you said the atmosphere on Titan is so thin. 1.45 atm. Gravity 0.138 G. So flapping muscle powered wings would work :)

One of the very first microprocessors - the RCA COSMAC CPD1802 was produced in a radiation hardened version (silicon on sapphire) and is still being produced today by Renesas. First shipped in 1976 so 45 years in production!

It’s always nice when you looking at RU-vid videos going they all look ugh! Then you see Scott Manley has dropped a video and then its Oh hell yeah!!

I'd love to see you speculate on the interior of the new Chinese Space Station and get your opinion on their equipment/design.

Nice mug Scott!

I once worked for an engineering firm in Salt Lake City a long time ago. Back in the late '90's they had a bunch of very expensive SGI (Silicon Graphics) workstations that kept getting huge numbers of bit flips in their DRAM which would cause problems for the long-running simulations we ran there. SGI would replace the RAM and test it back in Mountain View, CA (near San Jose @ sea level) and the DRAM sticks would be perfectly fine, then ship good ones to us in SLC and the bit flips would start right up again. This went on for quite a while until SGI figured out that the much greater (or more energetic?) atmospheric radiation flux at 4600' was _just_ enough to cause problems for the DRAM modules. They did a bunch of testing and switched to ones that were more radiation-hardened and the problems went away. I got a lot mileage over the years out of telling that story to people when cautioning them to keep an open mind about sources of error and failures in complex systems!

In the 80's I worked for the DoD's Reliability Analysis Center, which did groundbreaking forensics on failed aerospace microchips leading to critical static suppression/mitigation design improvements. Radiation hardening for space sounds sort of similar - but, of course, it would be tough to get back failed chips for analysis.

Insulating against the cold on Titan with such a dense atmosphere would be a nightmare. The thermal conductivity of the atmosphere on Titan is higher than on Earth. You would need an incredibly beefy suit to protect against that kind of cold. Your structures would have to be vacuum insulated. We're talking about a planet that averages -182.5c (-296F). You're going to have to build habitats, vehicles, suits and materials that will survive a cryogenic atmosphere. It is way harder than a lot of people seem to think. Insulating against a vacuum is easy. Vacuums have the worst thermal conductivity. An atmosphere with a surface density 1.5 times that of our own? Much, *MUCH* harder.

I would like to recover the Apollo 10 LEM so we could do some genetic tests on the mystery floating poop.... and finally determine who did it.

Back in the mid 80s, Harris made rad hardened 80C86 processors. They increased the noise margin by operating the processors at 10V, as opposed to 5V. So the components had to be modified to work at 10V, as 7V was usually the absolute maximum value. Probes launched as late as the late 90s, still used those.

I love these viewer question episodes. Great stuff 👍

can you do a computer programming in space video sometimes? with the coding guidelines etc. necessary for code that runs on space hardware? i know that video is going to be very niche, but interessting nontheless!

"I don't think gas giants will ever be candidates for habitation" Clearly you've not seen Issac Arthur.

Wow, I guess your explanation and a few very interesting top comments prompted Veritasium's latest vid about how cosmic radiation affects computers. Great topic!

Rad hardened processors cost a lot more. A NASA expert, told me, “ there’s a reason the boards we use cost two hundred thousand dollars”, and that’s for less processing power. Radiation is the most difficult problem for space travel in the long run.

Actually, flight trajectories involving the Lagrange points really is interesting. Orbital mechanics start doing interesting things when you're near them.

Regarding the question which Moon to put people on next: Remember to stay away from Europa.

One of the things to keep in mind with new CPU is when they say its a 2nm, 5nm, or 7nm process etc that actually isn't describing the size of anything on the chip. We've gone 3D in chips and stack transistors now, so when it goes from 5nm to 2nm it's describing the performance increase IF you shrunk the transistor. But we don't shrink anymore, we just add more layers or make layers better designed.

I remember reading an article at my last job when we got these aerospace journals, there was a company claiming they could avoid the bit flip problem by using just many more error correcting elements by using newer, but not radiation hardened equipment since they'd get a big processing, speed, and power boost while being able to package like 5x the hardware in the same space. Basically trying to correct the errors faster than they occur

Can you give us a deep dive into the SpaceX Starlink satelite deployment and propulsion systems? Is this information common knowledge? I'm wondering how much material is turned into space Junk too.

The fuel they should use for the mission to get mankind's gut bacteria back should be Phosphorus Oxygen Oxygen Phosphorus.

Excellent video very informative love your stuff well done sir

Love these videos. Just an addition, on earth oxygen is kept liquid by having high pressure, which makes the the boiling point higher. Since oxygen boils in air on normal earth temps, the pressure of a cryogenic oxygen tank is kept high as oxygen boils on top of added pressure.

I know surrey satellite's use of the shelf stuff from chatting to a friend at the Guildford university that works for them

I disagree, my lunch profiles are all about maximising deliciousness 😀

3:27 The Martian atmosphere is primarily carbon dioxide (freezes -80°C at standard pressure). Vacuum insulation works great until it gets punctured, then the latent heat released by solidifying carbon dioxide could quickly boil off your tank. The relief valve is necessarily uninsulated, so it may be covered in a thick layer of dry ice already. I would suggest installing two rupture disks in series with low-pressure helium between them, and of course a pressure indicator to verify that the helium is still there and hasn't gone off to explore the Martian surface.

So grateful for the free content

Time to challenge Elon to find and bring back the LM from Apollo 10 for the Smithsonian.

Nice shirt Scott!

Great session. Shorty. Thank you very much

Might I add one oxidizer, chlorine trifluoride. When a shipment spilled the quoted report was, the concrete was on fire. It burned through the concrete and nearly a meter of dirt below it. It also has other sensitivity traits that make it rather peculiar to be around. Insert x jokes as needed.

Do the L points get messed up by the other Planets, or are they balanced to take them into account like does Venus mess up L1 every now and then?

Flapping with huge wings around Titan, Gordon's Alive!!! ROAR

One thing I regret at my young years was not finishing school but sir thank you in my old years it amazing to have some one like you to make me still wanna learn I know it sounds stupid but thanks

Additional radiation hardnening techniques: Triple module redundancy with voting circuits between the three redundant systems. Additional memory technologies that are radiation tolerant/hardnened: FeRAM and MRAM. Both of which aren't reliant upon electric charge to store information and thus inherently immune/resistant to ionizing radiation. Also adding shielding to electronics or even using so called "guard-bands" that go around the individual transistors that contain induced charges from ionizing radiation (e.g. the induced charge doesn't leak from one transistor to another with the insulative guard bands).

Flapping with huge wings around Titan, sounds like Wild E Coyote. :-)

Highlight: hearing the word “poop” in a Scottish accent.

My favourite non-traditional oxidiser is high purity hydrogen peroxide. Not only is it a good oxidiser, it can be used as a monopropellant for both RCS and turbopumps. Even better, the catalytic breakdown products from the turbopumps can be injected directly into the combustion chamber because they still contain high temperature oxygen. Oh, and it can be stored at room temperature.

Fun fact, down on earth we also using radiation hardened processors. We using it mainly in medical industry (or other critical systems). Not so protected as in space but still. It’s mainly only steeplock (every operation is computed two times by two cores with small different time and compared). It is used to prevent errors happening because of different external distortions, space radiation among others.

I would like to know more about ic hardening

Have you read andy weir's new book? Would love your review and hot takes!

Since I wrote about the short story some time ago in the comments of one of Scott's videos, I'm glad he noticed it! Its Leonard Hansen reminds me a lot of the people mourning the "obvious loss to humanity" in not pursuing the Orion nuclear pulse drive. And yes, Syntin is a very good fuel, and can be used as a part of a mixture.

I actually stumbled across the Wiki on FOOF earlier today while writing some questions for Fall semester (I teach general chemistry). Amazingly well-timed! :D

Would building in caves protect computers from radiation on Mars?

Hi Scott, mam wants to know where you got those lovely shelves.

Nice video thanks Scott. Shout out to Charles Stross, Singularly or was it Accelerando? probably some of my favourite books.

id love to see a full scott vid on the radiation hardened components and history. i tried to explain to my dad why exactly its a cpu from the 90s and i need help XD

I've had a low thrust since I turned 50. Lol.

Wait, does this mean the drone has a more capable processor than the rover?!

Scott I am so a fan! You collaborated with my favorite author!!! Ann McCafferty is one of the best, Ann got all my attention with Dragon Riders of Pern. I even had a crush on her even though she was much older than me. I have been circling watching your vids. Scott Manley I am more than a fan now. I worked on the tooling team for the Space Shuttle, and many more at General Dynamics, San Diego. Your informed and imaginative videos are a breath of fresh air. Lots of likes and shares a comin your way Bro.

Thanks for the Charlie Ross reference. Now I know about aneutronic fusion!

Thank you Mr Manley.

9:19 as a long time viewer, that felt like a great payoff! lol made me smile at least cheers!

Thank you for the free content.

I think an interesting fuel oxidizer combination is HyImpulse who plan to use Paraffin wax and LOX in a solid-liquid combination with a few interesting properties.

Nice thing about Enceladus and Europa both is that yes, it is a high radiation environment ON THE SURFACE, but if you drill through the ice and place a subsea habitat below the ice, the radiation is pretty much nil.

if only i watched this video a week ago. the lag range point was one of the hardest questions on my final exam for physics

I loved the intro today Scott 😍

love your videos.....when talking of lagrange points you probably should have included the earth moon ones as these are coming into play on the going back to the moon push....

Worth noting only L4 and L5 are stable Lagrange point, L1,L2 and L3 are all lateral stable but axial unstable. In fact those 3 points are the anti-equilibrium point axially speaking, meaning any small displacement closer to one body will result in the object falling into that body. This is why all spacecrafts in the first 3 Lagrange points are not exactly at the point but all need a circular orbit around the Lagrange point to remain stable. L4 and L5 are the real stable point, so stable in fact there are large scale asteroid clusters at Sun-Jupiter L4 and L5. The famous Japanese mecha Gundam, envisioned humanity building large scale colonies at Earth-Moon L4 and L5.

There was a computer game called "Hardwar" (sic) set in a colony on Titan. The main gameplay was flying around in vehicles called "Moths" that had really fun flight characteristics, sort of modern drone-like. Still the best "Elite" style game I've ever played.

A problem with things like FOOF and dimethylmercury, even if you got past the toxicity worries, is that they are heavy. You get more energy per molecule than with hydrogen-oxygen, but less energy per gram. Just to keep things simple, think of hydrogen-fluorine (H[2] + F[2] --> 2HF) compared to hydrogen-oxygen (2H[2] + O[2] --> 2H[2]O). (Looking this up on the Wikipedia chemical bond data page, but it looks almost the same as my college organic chemistry book's table from 1980.) The hydrogen-fluorine bond has an energy of 568 kJ/mole (you make 2 of these), and the hydrogen-hydrogen bond has an energy of 435 kJ/mole (you break one of these), while the fluorine-fluorine bond has an energy of 158 kJ/mole (you break one of these). Net energy release is 543 kJ/mole of fluorine (or mole of hydrogen). The hydrogen-oxygen bond has an energy of 467 kJ/mole (you make 4 of these); the oxygen-oxygen bond in an oxygen molecule (which is very different from an oxygen-oxygen bond in a peroxide) has an energy of 495 kJ/mole (you break 1 of these) (and you break 2 hydrogen-hydrogen bonds at 435 kJ/mole as noted before). Net energy release is 503 kJ/mole of oxygen. So using 1 mole of fluorine to burn hydrogen does make more energy, but weighs 40.01 g (2 moles of hydrogen fluoride), whereas using 1 mole of oxygen to burn hydrogen weighs 36.03 g (2 moles of water). Energy to mass ratio for hydrogen-fluorine is 543 kJ/mole / 40.01 g/mole = 13.57 kJ/g. Energy to mass ratio for hydrogen-oxygen is 503 kJ/mole / 36.03 g/mole = 13.96 kJ/g. So hydrogen-oxygen is actually the better rocket fuel than hydrogen-fluorine. Unfortunately, while the above-mentioned data table gives a strength for an oxygen-oxygen peroxide bond (146 kJ/mole), it doesn't have the strength for an oxygen-fluorine bond (and I don't think my 1980 organic chemistry textbook does either), but it is reasonably probably somewhere in between the oxygen-oxygen peroxide bond and the fluorine-fluorine bond -- split the difference and call it 152 kJ/mole (and remember that this leaves plenty of energy to drive FOOF to decompose back into oxygen and fluorine even at -100 `C, because when that happens you get back the much stronger oxygen-oxygen bond of the oxygen molecule). So if you use 1 mole of FOOF to burn hydrogen to make a mixture of hydrogen fluoride and water (3H[2] + FOOF --> 2HF + 2H[2]O), you break 2 oxygen-fluorine bonds (estimated 152 kJ/mole), 1 oxygen-oxygen peroxide bond (146 kJ/mole), and 3 hydrogen-hydrogen bonds (435 kJ/mole), and you make 2 hydrogen-fluorine bonds (568 kJ/mole) and 4 hydrogen-oxygen bonds (495 kJ/mole). Net energy for hydrogen-FOOF release is 1361 kJ/mole; total molecular weight (2 hydrogen fluoride + 2 water) is 76.04 g/mole. Energy release by weight for hydrogen-FOOF is 17.9 kJ/g, which is noticeably better than hydrogen-oxygen, but it's not clear if it is enough better to be worth all the trouble. And then we get to mercury. I don't have bond energies for anything including mercury, but I can look up the fact that the average atomic weight of mercury is 200.592 g/mole, and you can only get 2 mercury-fluorine bonds with release of energy, and the mercury-fluorine bonds aren't going to be THAT much stronger than hydrogen-fluorine bonds (and mercury-oxygen bonds are positively _wimpy_ -- mercuric oxide decomposes at only a few hundred degrees centigrade). (Somebody figured out a way to cram 4 atoms of fluorine onto each mercury atom, but this _requires_ energy, and the resulting compound is only metastable in frozen argon.) So DITCH IT. Nothing including mercury in any stoichiometric quantity is going to be any good as a rocket fuel, even if you had absolutely no worries about the toxicity. Using a mercury compound for a rocket fuel would be like trying to build an electric rocket powered by lead-acid batteries -- it just isn't going to work.

"when Starliner launches..." good one there Scott! lol

19:00 Correction: *thick* not thin. High density along with low gravity is what makes flying easy. If only it weren't so obnoxiously cold (but then again that's a big part of the reason the atmosphere is so dense anyway).

Interesting that the rad hard processors etc are made by BAE Systems (British Aerospace) in the US, and the ARM processors are of course also British in origin.

Titan traversal is a great opportunity to revive all those kooky "bird man" projects. I'm thinking pedal-powered wings or rotors would be a great way to build something with enough lift to take home groceries.

5 times smaller feature size = 25 times more elements per area. Now they are going into 3rd dimension, but "feature size" is still computed as "area equivalent".