OP was talking about dark energy, not dark matter. Given that Hawking radiation is in the form of photons which have no rest mass, it's already a poor candidate for dark matter.

Then there's the minor detail that the combined Hawking radiation output of every black hole in the observable universe is less than than the power consumption of the tablet I'm typing this on. Indeed, black holes absorb far more energy than that just from cosmic microwave background radiation, so they're net energy absorbers even without actively feeding on matter.

Even if we ignore that, the the upper limit on thier total energy emited as hawking radiation since the big bang is on the order of a few joules, or less than a picogram of mass equivalence. Ignoring primordial black holes anyway, since we've got no solid evidence for thier existence.

Even if it was a signifcant amount it wouldn't matter since any amount of 'hawking-equivalent-matter' emited by a black hole will reduce the black hole's mass by that same amount.

Since the combined mass of all black holes theoreized to exist is insuffcient to explain dark matter, then so too will be any amount of hawking radiation emited by them.

My point was actually more that the gap between SpaceX and CASC/CALT is proportionally smaller than SpaceX minus Starlink vs RocketLab (or ULA if going by upmass).

By excluding internal launches, you're actually proportionally increasing SpaceX's lead.

> So numbers, while being absolutely impressive, are inflated.

If you want to discount those then you also need to effectively discount CASC/CALT and Roscosmos, who were second and third place respectively for most launches last year, since they're launching almost entirely for themselves. (Roscomos had one commercial launch last year before they invaded Ukraine)

That puts Rocketlab in second place with 9 commercial launches last year compared to SpaceX's 27. Though of course SpaceX still crushes them on upmass given the size difference - not to mention three of those launches were Transporter missions each carrying the equivalent of like a dozen RocketLab customers.

Over a small number of launches though. It's commonality with Delta IV Medium gives it some more credit, but generally speaking I'd rather fly on a rocket with 1 failure in 100 than 0 failures in 14, even though 99%<100%.

I used to make the point that Epsilon is also 100% reliable, but that since it only has 5 launches under it's belt that doesn't amount to much in practice.

I was quite vindicated when it recently failed on it's 6th launch, though also a little saddened that I'd no longer be able to use it as an example.

DC-X was a long way short of Falcon 9, or even New Shepard, and the full Delta Clipper would never have been a useful orbital launcher so long as they insisted on it being an SSTO.

SpaceX are still building Falcon boosters at a higher rate than Ariane are building Ariane stacks, despite reuse. Not to mention a lot more upper stages. Consider the following production figures:

Granted, Ariane 5 is heading towards retirement so they're winding production down, but historically it averaged about 6 per year during the 2010s, which is comparable to the rate SpaceX have built Falcon boosters at since reuse started becoming common practice circa 2018, and they show no signs of slowing given they built 7 last year and 3 in just the first two months of this year.

(Note: I'm using maiden launches as a proxy for production figures. Actual completion dates are likely some months earlier, but over a time period of 5+ years it averages out)

&nbsp;

Reuse doesn't necessarily have to reduce the number of rockets you have to build, that's stinkin thinkin.

It can instead allow you to build the same number of rockets but get a lot more launches done with those rockets - as evidenced by the Upper Stage figures for Falcon, which are 1:1 with the number of launches in a given year.

Consider also that SpaceX want to build Starship at a rate of one per month despite it being fully reusable.

I bet if you took their current internet access away and told them that Starlink was their only option for getting back online a lot of people would start changing their minds real quick.

>Didn’t Starlink redesign for that purpose once already

Brightness mitigation has generally been more of an ongoing process than a one-off, though they've just done a major revision that launched only yesterday. SpaceX are all about design iteration, and Starlink is no exception to that rule.

The details of their previous efforts can be found in this document, and the details for the new design can be found on page 3 of this document.

Where'd you get 28km/s?

My handy dandy Delta-V map says LEO to Callisto's surface is only ~14.2km/s. I know Delta-V maps aren't super accurate but I find it hard to believe it's that wrong.

For reference, the map also puts LEO to Mars surface at 9.5km/s, which is within a reasonable margin of error of 10km/s.

There's a huge gap between knowing enough about rockets to critique their merits and knowing enough to build one entirely yourself. I doubt many car reviewers could build a car from scratch, but they can still make an informed choice between two options by analyzing and comparing various criteria.

As it happens though I do know a fair bit about rockets, so I'm quite happy to discuss any specifics you want.

&nbsp;

I'd also point out that it's not just random people on reddit levying criticisms at the SLS program, there are some people with serious credentials saying similar things. The Government Accountability Office issued a pretty harsh report on it, and Lori Garver, former Deputy Administrator of NASA, has also had some choice comments and insights.

Perhaps the most notable example is Charlie Bolden, who was the head of NASA under Obama, and as such oversaw the SLS program for about 5 years. In 2014 he (in)famously said:

>“Let’s be very honest, we don’t have a commercially available heavy-lift vehicle. The Falcon 9 Heavy may some day come about. It’s on the drawing board right now. SLS is real.”

At that point in time he clearly had a very high opinion of the SLS program and believed it would come to fruition well before Falcon Heavy. However, after Falcon Heavy ended up launching before SLS (by almost 5 whole years), he had to reevaluate his stance. In 2020 he said:

>“SLS will go away. It could go away during a Biden administration or a next Trump administration, because at some point commercial entities are going to catch up. They are really going to build a heavy lift launch vehicle sort of like SLS that they will be able to fly for a much cheaper price than NASA can do SLS. That’s just the way it works.”

Note that comment was made before the 2020 election results, so when he says "during a Biden administration or a next Trump administration" he's talking about the period from 2021-2024.

And while he doesn't explicitly say who or what rocket, SpaceX and Starship are really the only ones that fit his description.

>If you don't like it why don't you make your own rocket

People criticizing the SLS aren't usually contending that they themselves could do better, but that someone else could - typically SpaceX, though I personally also like making comparisons to the Saturn V.

People criticizing the SLS program are usually pointing out the corruption/pork barreling, rather than anything technical about the rocket itself. Though of course, these two criticisms aren't mutually exclusive.

>Armchair astronauts

This should be armchair engineer or at least armchair rocket scientist, no?

Astronauts ride rockets. They don't build them.

>Starship would be better utilized to build a fast human Mars transport vehicle in LEO than being used directly as the crew transport.

I agree, but the way to do that is with NEP, not NTP. Or at least not of the solid core variety anyway; gas core NTRs might do the job, but it doesn't seem likely they'll be a thing anytime soon.

Solid core NTRs don't really get you to Mars any quicker than Starship. With an Isp of 900s and a mass ratio of 5 you're looking at 14.2km/s. Starship with it's 380s Isp and ~12 mass ratio only gets 9.2km/s - 5km/s less.

On the face of it, that seems like a 50% speed increase, which is nothing to scoff at. The problem lies in slowing down at Mars. Starship is able to aerobrake, saving it ~4km/s of delta-v as compared to propulsively braking into orbit.

So there goes most of that 5km/s lead for the NTR stage, and the remaining 1km/s has to be split in two - a 500m/s higher cruise speed also incurs 500m/s more braking requirement - a much more modest 6% speed increase.

You could fit the NTR stage to aerobrake as well, but given the massive size of the tanks and the high dry mass of the NTR, it's likely to suffer a proportionally larger performance hit than Starship paid for the same capability.

&nbsp;

Now, to an airless body such as Ceres or the Jovian moons it's obviously a different story and the NTR takes a considerable lead over Starship. However, NEP's advantages over NTP grow even more pronounced over larger distances, bringing me back to my original point.

The only real exception is of course Luna, where you can't aerobrake, and it's very close proximity puts NEP at a huge disadvantage.

NTRs perform a bit better if you focus on payload capacity instead of speed, but only in terms of mass fraction. In terms of cost the pure hydrogen and enriched uranium might well cancel out any mass savings if you've got cheap LEO lift. And in terms of ISRU, pure hydrogen is an order of magnitude more energy intensive to produce than hydrolox or methalox.

I'm definitely a Starship enthusiast, but I don't think Starship 'solves all problems'. It solves many problems, and NEP solves most of the other ones, while NTP only solves a few niche ones in between, so I have to wonder if it's really worth the trouble.

NTRs are for getting stuff around the solar system more efficiently - though in many cases NEP would be better, I wish that was getting as much attention.

For interstellar travel though you want a fission fragment engine as a bare minimum, preferably fusion propulsion of some sort.

It's not a practical limit, it's a theoretical one. The original calculation that the 100km/62mile definition allegedly stems from was done by Theodore von Kármán, who determined that at ~84km/52miles, an airplane would have to move so fast to produce sufficient lift that it's speed would place it in orbit. The number was rounded up in most countries to 100km, though down to 50 miles in the US.

The fact that noone has actually flown a plane in sustained flight anywhere near that high doesn't change the math. Though I'd note that there have been unpowered flights at such altitudes - the Space Shuttle, Buran, X-37B, various hypersonic glide vehicles, etc.

Perhaps the best example is the Apollo capsule - not typically something you'd think of as an aircraft, but it did produce lift, and if you look at it's reentry profile you can see that it managed to maintain (approximately) level flight at around 200,000ft for a fair distance, before finally bleeding off enough speed to continue descending.

If the Apollo capsule, or Space Shuttle, or whatever had been fitted with some form of propulsion to maintain speed, then they could theoretically have sustained level flight at over 200,000ft - at least until their heatshields gave out anyway.

S-IC nominal burn time is 150 seconds, or exactly 2.5 minutes.

Superheavy nominal burn time is 169 seconds, or 2.8 minutes.

Though that figure may have been for the 29 engine version, the 33 engine version might be a bit less since it burns fuel quicker.

Either way, it's close enough to the Saturn V that I think it's fair to call it a 'similar' duration.

I doubt the launch pad would survive such a test. I'm not sure there's a facility in the world that could.

The test stand at NASA's Marshall Center was able to handle the Saturn V first stage static firing for a similar duration to Superheavy (~2.5 minutes), but Superheavy's energy output is about 2.7 times greater.

It's still extremely inefficient. 100kg worth of plutonium pellets in some RTGs will produce about 50 kilowatts of thermal power. 100kg of plutonium in an SMR on the other hand could easily provide 500 megawatts of thermal power.

The average US household has an average power draw of about 1.3 kilowatts, so assuming 100% conversion efficiency in the above cases, the RTGs could power about 38 houses while the SMR could power about 38,500 houses.

Given how expensive plutonium is, that thousand-fold difference makes RTGs a complete non-starter. And in practice an SMR would actually use uranium which is much cheaper, making things even worse.

Sometimes it takes quite a while for a theory to be practically implemented.

Work on Scramjets started in the 50s, with engines working in laboratory conditions in the 60s, but didn't operate in real flight conditions until the 90s and have only recently started to approach practical use - it's hard to say exactly where they are today since most such projects are classified.

I'm not sure when the theory for FFSC engines dates to, but the first example was built in the 60s. However it was unable to sustain stable combustion and the first stable engine wasn't tested until the early 2000s. The first test flight of an FFSC engine wasn't until 2019, and the first practical use will probably occur this year.

The basic theory for fusion dates back to the 1920s, with proposals for fusion power specifically dating to the 1950s, but it still hasn't gone anywhere, yet. We have been making steady progress, so it may still go somewhere given more time. The recent scientific breakeven at the NIF was a significant, if not directly applicable milestone.

Advances in computer control technology have been instrumental to a lot of the recent progress in the aforementioned applications. Having the theory is one thing, being able to control a complex and delicate process in practice is another.

Maintaining continuous rotating detonation has proven quite challenging in the past, typically breaking down due to instabilities in a matter of milliseconds. The fact that NASA were able to run this engine for what looks like about 8 seconds is very promising indeed.

By my count there have been 9 solar sail demonstration missions launched, the first being Cosmos 1 in 2005 and the most recent being the NEA Scout launched just a few months ago.

However, the vast majority of these missions have either outright failed or been a dubious success at best. The only one which has been truly successful and demonstrated practical use of a solar sail was IKAROS in 2010, which operated successfully until 2015.

It only takes one success to prove that the concept works, but the large number of failures seems to have tempered expectations and dampened enthusiasm somewhat.

There was a proposed follow-up to IKAROS called OKEANOS intended to go to the Trojan asteroids, which would have been the first real use of a solar sail for a scientific mission.

It was a finalist for ISAS's consideration in 2019, but ultimately lost to LiteBIRD. Unfortunately there just isn't enough space science funding for all the missions people would like to do, so a lot of neat stuff gets passed on.

We've had plenty of solar powered electric propulsion since. The Dawn mission was a great example of what electric propulsion can do. Solar power just doesn't scale up well to larger vehicles, or work very well as you get furthur from the sun.

A nuclear electric system has the potential to be much faster than a chemical rocket over long distances, i.e to Mars or especially beyond.

The real issue has been the reluctance to put nuclear reactors into space. SNAP-10A remains the only example the US has ever launched, even though much better designs like the SAFE-400 and KRUSTY have since been developed.

>Or just with 9x the mass flow rate

You need 9x the molar flow rate, not the mass flow rate. And since hydrogen has 1/9th the molar mass of water, it ends up cancelling out.

You seem to be ignoring the minor fact that lower molar mass also means more moles, so it cancels out.

If you pump 1kg of water into the engine, that's 55.5 moles. If you pump 1kg of hydrogen into the engine, that's 500 moles.

So hydrogen produces 1/9th as much force per mole, but it also has 9 times as many moles per kg of fuel. The end result is that both produce the same total force when that kilogram is expelled from the engine.

Or at least, they would if they were both expelled at the same speed. Since hydrogen actually comes out twice as fast, it produces 1/4.5th as much force per mole, while still having 9 times as many moles, and hence produces twice as much total force.

It would work if you added some propellant to the mix. For example, add a big water tank and use that electric motor to drive a pump that sprays the water out the back at very high pressure.

Realistically you're not going to get a very good exhaust velocity with that method, so you'd instead use a different kind of electric engine to accelerate the propellant; electrostatic, electrothermal, or electromagnetic.

Indeed, the only example to date of nuclear propulsion actually being used in space was on SNAP-10A, which featured a nuclear reactor powering an electrostatic engine with cesium as the propellant.

Granted, it only worked for about an hour before it broke down, but it did work. It's a shame there hasn't been any followup in the 58 years since then.

>When approached by NASA in 2018 for potential SLS replacement, they stated to NASA and press questions that the best theoretical max for Falcon Heavy disposable TLI payload is 18,000kg, but only a realistic 16,000kg to lunar orbit for a crewed vehicl

Source?

>The Mars injection is using the 6 month Holmann transfer window,

Even the lowest energy Mars transfer still needs more energy than fast TLI.

A best case Hohmann transfer during an ideal window like the 2033 window would be about 3500m/s. However, the upcoming 2024 window will be more like 4100m/s. On average it tends to be around 3900m/s, so I expect SpaceX's payload figures to be based on something like that.

By comparison, Apollo's fast TLI burns were nominally 10,400fps, or ~3170m/s. Call it 3200m/s. No matter how you cut it, the Mars transfer needs several hundred m/s more delta-v.

It seems very odd that an extra ~700m/s to get from GTO to TLI reduces payload from 26.7 tonnes to 16-18 tonnes, or a whopping 10 tonnes less, yet another ~700m/s to get to TMI results in virtually the same payload.

Even if SpaceX are using the best case Mars transfer, you'd still expect it to be a few tonnes less.

I mean nuclear fuel is typically comprised of uranium-235 and uranium-238, both of which occur naturally and can be found in low levels within all rock, soil, and water. Soil for example contains about 12 milligrams of uranium per kilogram on average.

A typical space-based fission reactor contains abut 30 kg of uranium. If a rocket blew up halfway through the launch and scattered that perfectly over a radius of say 50km, and it was all absorbed by just the top 1cm of soil, that would amount to an additional 0.3 milligrams of uranium per kilogram of soil.

Of course, if it spread over a smaller area the concentration would be higher, but it would have to be a pretty small area for there to be enough to matter, so it's not likely to be a major hazard in the grand scheme of things.

&nbsp;

In practice it's more likely that the fuel rods would remain largely or entirely intact and end up at the bottom of the ocean somewhere. The ocean contains approximately 4 billion tonnes of uranium, so even if the fuel rods were gradually eroded, they'd quickly be diluted into irrelevance.

Now, there are some ways that it might be possible for someone to be exposed to a dangerous quantity - for example, say something like a gram of uranium being chipped off and somehow ingested by someone, my point is more that it's not going to be a widespread ecological disaster.

Whereas in the case of a disaster like Chernobyl, there were a lot of nasty isotopes present in the partially spent fuel rods, most notably iodine-131, caesium-134, caesium-137 and strontium-90. These isotopes are tens of millions of times radioactive than uranium-235 or uranium-238, so even the most miniscule quantities are dangerous.

I'd also point out that we already regularly launch other dangerous substances on rockets. Hydrazine for example has comparable toxicity per milligram to uranium, and large satellites are regularly launched with literal tonnes of that onboard.