K^2

Definition of a moon is too broad. I would require the same of a moon as any dwarf planet. Namely, enough mass to make a tidy sphere. Everything else can be classified as a moonlet. That would, of course, mean that Mars has zero moons, and Saturn and Jupiter have way fewer. Additionally, I stand with Issac Asimov and a handful of astronomers in that a moon is an object that orbits its planet first and foremost. As such, Moon isn't a moon at all, since it primarily orbits the Sun, merely sharing orbit with Earth, and is therefore a part of a double planet Earth-Moon system. To avoid confusion, I would rename Moon back to its Latin name Luna, and therefore, all five inner planets, Mercury, Venus, Earth, Luna, and Mars would have no moons under this classification. Nice and tidy. I do not appear to be in the majority with this point of view, but I'll keep pushing for it.

Absolutely all of them had inertial navigation of some sort. Even the olden-days ballistic missiles didn't execute the program on the clock. Instead, they had series of gyro-integrators that would help keep rocket on course, pull the program tape or equivalent, and be capable of triggering warhead or self-destruct as necessary. Soviets stuck to that very heavily until the end, while US has added more and more electronic components in later models. And while some rockets weren't stored fueled, as far as I know, they were all stored pre-programmed with targets. Targeting information was just too sensitive to truck it out to the rocket at the last moment. Of course, modern computers made that bit entirely redundant. Still, I'm not sure if Russia ever switched over to computer guidance. Back in the days of Cold War, Soviets considered electronic components a weakness, potential for an attack, and avoided them even when ICs became available. There wasn't even a radio on board. Once it lifted out of silo, it was heading to its target and that's that. Part of me wouldn't be surprised if Russian ICBMs currently deployed still use punch tape with mechanical gyro-integrators.

That would actually make a huge difference. Do you have a list of requirements somewhere? Mostly in terms of mechanical reliability testing something would have to go through before getting a ride.

Unfortunately, I no longer remotely have the amount of time required to make but an incidental contribution. (And for a while, I was locked out completely by employment contracts.) I can probably assist in getting guidance and tracking worked out on the mathy side, but there are tons and tons of mechanical and electrical engineering that needs to be done.

Out! () Anyways, since this was just a 3% event, my 756 day estimate is slightly off. But the degree to which this matched the last 3% event is astonishing. This suggests another 20% event is incoming. If anything, that makes orbit more Earthlike, in terms of solar flux. As likely as that is to be a coincidence, that's still amazing. I hope, all doubt of this periodicity being due to orbital period of something is burried at this point. That kind of perfect repitition would have to be artificial otherwise. But the fact that we missed a whole event in 2015 is highly unusual. The only explanation that comes to mind is a gas giant orbiting at high inclanation w.r.t. orbit of whatever this cloud is, causing the star to shift out of plane of orbit. Do we have red shift data on Boyajian's star? I would love to run it through a model to see if it happens to confirm such oscillations. If we are lucky on inclanations, of course. P.S. No change in spectrum observed so far. So solid objects then, at least hundreds of microns range. That mostly excludes dust and gas. So much for comets. But also excludes fresh impact. I've ran out of ideas that don't involve aliens, and I'm not ready to accept aliens. P.P.S. Scratch that, I have a plausible explanation! The 20% dip is caused by a dense asteroid ring. The 3% dip by a brown dwarf. It is massive enough to wobble the star in and out of the plane of the belt. 3% dips are half as frequent as 20% and later can vary in intensity based on density of the asteroid field. It matches, and you've read it here first. Unfortunately, totally not habitable zone then, nor aliens. But might be publishable. I'm going to go check numbers.

It is when it is brief and periodic. Dips observed so far are exactly timed in duration and periodicity to a large cloud of something orbiting that star at 1.83AU with low ecentricity. Anything having to do with star itself would either be not nearly as pronounced or not periodic. The only thing worth discussing is source of the cloud. Debris field of some sort is most likely, with star impact leading the polls, but I don't think planet-planet has been properly excluded. A terrestrial world passing too close to a gas giant can produce exactly this sort of field without leaving a thermal signature of an impact. Yes, prior is low for such event, but it is a much better fit for orbit of the cloud.

And on schedule! That puts the period at 756 days, almost exactly the value expected from the first two dips. Whatever it is, it appears to be roughly 1.83AU from the star. For that star, that puts spherical black body radiation equilibrium temperature at 302K. That's 29°C or 84°F. If it's painted a light color, might actually be closer to 20°C So not only is this definitely in the habitable zone of the star, that's probably where we'd actually build a habitat if we had the tech for it. This is starting to get interesting now. Edit: Alternatively, if this is a debris field of a formerly habitable world, this is getting terrifying. We might be looking at a graveyard.

You get 30cm of wire for 1ns latency. You don't get that much distance round-trip on most boards. Given that RAM latency is absolutely glacial on that time scale (~15ns), it's simply not worth the effort to migrate northbridge into the CPU. You get significantly more bang out of dedicating more of the die to cache instead.

Not an expert, but I would guess that the biggest simplification is in the pump mechanism. These rockets use your basic fuel-rich staged combustion setup, and hypergolics make it almost laughably simple. No starter mechanism required. You just open the valve, the tank pressure forces a bit of fuel and oxidizer into pre-burner, and that's enough to start spinning the turbine, ramping up the pressure at which the fuel and oxidizer are fed until you have proper flow into the main combustion chamber. The other simplification is the fact that both fuel and oxidizer are liquids at room temperature with just a touch of pressure in the tanks. This means you don't need to have any kind of insulation or low-temperature materials involved. All in all, nitrogen tetroxide with some kind of hydrazine fuel allow for very simple, very light engines. The drawbacks, of course, are that both are quite toxic, and between that and aforementioned hypergolic quality, failures tend to be extra dangerous even by rocketry standards.

Not going to be that cheap. ~1kg cubesat rides are going for $50k+. It's highly non-linear, because a lot of work goes into verifying your payload is safe and can be properly deployed to a safe orbit. I can't imagine you'd be able to launch anything at all for less than $10k.

Take a look at operational history of V-22 Osprey and similar aircraft. Personal attitudes of members of the military can vary, but the net result is the same. They accept loss of human life as part of development and operational cost of the hardware. It's not even wrong. Looking at it from perspective of cold economics is preferable to loss of efficiency over misplaced sentimentality. Because however many lives are at risk at any given moment, countless more could be at stake. Treating soldiers as monetary investment, at grand command scale, is the best way to minimize loss of human life overall.

You lose an input DoF with a ball joint. You can still get any 3DoF orientation you want by combining rotations, but you aren't necessarily going to be able to track an arbitrary curve. In addition, you're going to be limited to less than 180° cone. The other classical solution is many-jointed arm. Again, you typically want at least four, possibly five joints to allow for full tracking, and you'll have to be solving inverse kinematics problem along the way. Of course, that's really just a generalization of gimbal concept, as anything with joints will be. The only other thing I can think of is magnetic suspension which avoids joints entirely.

Classical solution to gimbal lock is fourth gimbal. Tracking solution for all four gimbals becomes more complex, but it's not a problem for modern computers or even MCUs.

I suspect, this thing is actually less safe than a capsule overall. For a capsule with parachute/retrorockets, it doesn't matter a whole lot where it's going to land, so long as the chosen braking mechanism works. It's a single point of failure that can be made sufficiently redundant, and you pick your landing spot in advance. Military might be a lot more picky about ability to change landing location at the last moment. I mean, if you were going to land a squad of spec-ops on a friendly airport to help conduct a clandestine operation, and suddenly that airport is no longer so friendly, you might not want to deliver your secret hardware and men to the enemy. If the above means that you are more likely to suffer a heat shield failure, over-g on landing, landing gear issue, or some other mishap, then it's an acceptable compromise. Loss of crew during military operation isn't going to be a PR disaster for the military. It's just loss of money invested in hardware and training for them.

Slightly off. I'm getting 99.956m to the same precision. The part under sqrt comes out to a hair over 742850. So you should be getting: r = 1600 * 450 * 110 / (742850. + 450 * 110) And yeah, I'm extremely happy with how easy Mathematica makes it not only to solve the problems, but also generate visuals. Sure, I can set everything up by hand and then make images in CAD, but I'm entirely too lazy to bother with all that.

You aren't actually getting much speed in the first 10k. Most of the thrust is spent fighting atmo and gravity to get to thinner air. 10km up, a rocket is going to be through about 1.5km/s of its 9km/s delta-V to orbit. That's going to be between a third and half of its fuel, depending on ISP. So 90% thing is definitely crap, but because it's simply wrong, not on technicalities.

In general, you find the size of a sphere inscribed in a convex polytope by solving a system of inequalities. But this problem is highly symmetrical, which makes things easier. Consider such pyramid placed in a 3D Cartesian coordinate system with center of the rhombus (diamond) at the origin and apex of the pyramid on the Z axis. Maximum inscribed sphere will have its center on the Z axis as well, thanks to the symmetry, and that means we only need to consider contact with one of the 4 sides and the base. Base is easy. Sphere or radius r will have its center at (0, 0, r) so as to touch the base. Now we just need to know the distance to the side faces. Consider face that passes through (450, 0, 0), (0, 110, 0), and (0, 0, 1600). The equation that describes this plane is given by: 110 * 1600 * x + 450 * 1600 * y + 450 * 110 * z = 450 * 110 * 1600 Consequently, distance between this plane and point (0, 0, r) is the following: d = (1600 - r) * (450 * 110) / sqrt(110² * 1600² + 450² * 1600² + 450² * 110²) And for an inscribed sphere, all we have to do is substitute d = r and solve for r. I'm sure you can solve a linear equation, so I'm leaving that up to you. P.S. As one might expect from the dimensions, it's not very big.

Hydrogen won't follow ideal gas laws at high pressure. It's a notoriously difficult substance to compress. Basically, don't count on density going up much above 0.1g/cm³.

On average, cross-section of a proton divided by surface area of a sphere with radius equal to inter-atomic distance. But I suspect variance will be rather high. Most of the energy will be carried away by relatively few high energy particles. Picture trying to hit a pea with a shotgun from across a football field while blind-folded. If you hit it, won't be much left of a pea, but your odds aren't great. What I don't know is how much low-energy radiation is released. Anything in 1eV range is likely to be absorbed, causing that proton to go flying, but not terribly fast. These are ordinary chemical energy ranges.

That's actually completely wrong. Matter-antimatter reactions can happen with many different pairs of particles. The only rule is that all conserved quantities are, well, conserved. Lets start with the fact that proton isn't elementary. It's made up of a crap ton of particles and antiparticles. I'm going to let that sink in for a moment. Ok, so the interior of a proton is filled with quarks, antiquarks, and gluons. Gluons are massless, so they are their own antiparticles, just like photons. And together with quarks and antiquarks all of this mess is a constantly bubbling soup of quantum probability with new pairs of quarks and antiquarks constantly created and destroyed. However, no matter at what point you do the counting, there are always going to be precisely 3 quarks without an antiquark pair, and these are known as valence quarks. Moreover, in a proton, these are going to be two up quarks and a down quark. These two flavors of quarks (yes, they are actually called flavors) are different from each other in their electric charge and isospin. Both of these quantities are conserved, and really, what makes a proton a proton. In contrast, neutron will have two down quarks and one up quark. All other quark compositions are highly unstable but do have corresponding particles. With all of this in mind, you can probably see that annihilation event isn't going to be a clean affair. In proton-antiproton collision, all of these particles will start interacting with each other. Quarks from one hadron will start annihilating with anti-quarks from another, all the while new pairs are still being created out of quantum vacuum. But what's interesting is that there is now an anti-up quark for every up quark and an anti-down quark for every down quark. So they can all find each other and annihilate the proton and anti-proton completely. Can, but usually don't. The moment proton and anti-proton meet, the amount of energy released blasts the whole thing apart while the annihilation is still in process. One possible outcome is that two valence quarks meet, say a down quark met its anti-down match. Lots of energy released, new particles could be created, but among remains are the old valence quarks. And suppose an anti-up quark got blasted out with a down quark. And the up quark with the anti-down. These are pi mesons, aka pions. (Or rho mesons, depending on their spin.) For simplicity, lets say they are pions. π+ and a π-, with charges +1 and -1 respectively. So with up quark not having a corresponding anti-up in the π+, do we expect it to never decay? No, actually, pions have a very short half-life and a number of possible decay modes. The most direct is for the up quark to annihilate with the anti-down quark. That leaves us with an uncomfortable remainder carrying not only energy but electric charge. What can it be? Well, there are some options, but the most likely product of anti-up and down quark annihilation is a W+ boson carrying the charge and a photon carrying some of the energy, recoil momentum, and counter-balancing the spin of the W. Photon's stable, we're good. But not the W. That needs to decay. And just like its origin, it produces a pair of particles that aren't each other's antiparticles. Most likely case in pion decay chain is an anti-muon and a muon neutrino. These are also two flavors of the same kind of particle. But nonetheless, they aren't a particle-antiparticle pair. Neutrino flies off. Anti-muon decays again. This time, finally, to stable particles. Positron (anti-electron), electron neutrino, and a muon anti-neutrino. All with considerable energy. This is just one branch of one possible chain in proton-antiproton annihilation. Which brings us now to a question of what happens if anti-proton encounters something else, like a neutron, instead? Well, all of the rules from above apply. Anti-quarks will look for quarks, not necessarily of the same flavor, and neutron has these. So there will be an annihilation event. Just like before, it's going to blast the quark-antiquark mix into a bunch of new particles, and these will eventually decay to some stable particles. After everything calms down, you'll just have a bunch of photons, various flavors of neutrinos, and maybe leftover positron until that encounters something interesting. If you really start digging into particle physics, we don't even talk about creation/annihilation events. The quantity of interest is a vertex. There is very little difference between electron scattering and electron-positron production. They are both vertices with two electron currents and one photon current. The only difference is time-ordering, and that's a fairly flimsy thing at these energies. Likewise, two particles of different flavor being annihilated is similar to a flavor-changing weak interaction. So if there is a way for particle A to become particle B, then A can annihilate with anti-B. And that opens up a whole bunch of possible combinations.

Not particularly. The main reason to have large aperture on individual telescopes is to collect more light. If you are trying to look at a really distant object, inverse square law really bites you. And with interferometer, it tends to be much harder to have steady tracking, which is what something like Hubble would use to take a long exposure picture. To reduce exposure time to something reasonable, you need rather large telescopes, which is equivalent to having telescopes with large resolving power, even if you aren't actually using that aspect of it. You can always think of an interferometer telescope as one giant telescope whose entire lens is covered by something opaque except in a few places where you have your component telescopes. As you probably know, covering parts, even significant portions of the lens on a telescope doesn't affect resolving power, but does reduce amount of light collected, requiring longer exposure time. This probably won't be as big of a problem when we start building space interferometers. An interferometer satellite in space should, in theory, be able to track just as well as a smaller space telescope, allowing for very long exposure images of very distant objects, finally letting us capture images of various nearby exoplanets.

This type of problem is solved by an unguided learning system, which requires a ton of training input. Even if you are using an NN approach for the guidance system, your best bet is a genetic algorithm. With that in mind, I wouldn't start with an actual game. There is no way to run hundreds of thousands of trials in short amount of time. I'd start with a good simulation, which can be ran much faster if you don't wast time on rendering and approximate rocket as rigid. Simulation like that can run an entire launch to orbit in a fraction of a second. This is where you can start learning. You can then hook it up to the game to display some of the better results with an actual flight. P.S. For anyone who wants to do NNs or Deep Learning in Python, I strongly recommend TensorFlow. It's a modern tool, far superior to most Python NN libraries.

Good call on Doppler Effect. My mind was locked on a static picture, where distances don't change.

No, this is absolutely wrong. Once sound is produced, its frequency will stay the same. The wavelength will change, certainly, but frequency has to stay the same. If you received more oscillations than were produced, where did additional ones come from? The future? Unless you are dealing with relativistic red/blue shifts that actually have to do with time behaving in a funny way, frequencies never, ever change. Be it light or sound. On the other hand, certain means of producing sound will change in the source frequency as the speed of sound changes. Resonant frequencies of your voice will shift in Helium. But you'll hear that shift even if you are in a room filled with nothing but helium. Sound produced by your voice is already higher pitched in helium, and whether it transitions to normal air or not makes no difference. Likewise, any wind instrument will change pitch in a different atmosphere. And by the way, @Green Baron, speed of sound, and consequently certain sounds, will change with altitude. It won't touch things like engine sound, though, and won't be noticeable enough with voice, so it's easy to miss. Indeed, to get a sufficient shift, you need to climb to the point where you need an oxygen mask.

That's definitely a thought. Falcon Heavy should be capable. They'll probably want to launch something useful as well, like maybe a GTO stage with a telecom sat, but that still leaves space for a minivan at least. Maybe even a short school buss. What caught my eye in the article, though, is the fact that they plan to use flown 9Rs as boosters. That seems like a brilliant move. They can cut back on throttle to the boosters for early launches, dramatically reducing the durability threshold. So they can get away with using flown stages with significantly less refurbishing than it'd take to get them ready as cores for another 9 launch. Eventually, they'll have to get to a point where they can re-fly each 9R core as many times as they like and get them back to as-new condition, but using new core for central stage and flown cores for boosters lets them launch Heavy this year and within an on-target budget. That's fantastic news in my book. It's been ages since we've had a decent heavy lifter.