Z-Man

ZetaX: Momentum and energy are linked by Lorentz transformations. If you violate momentum conservation for one observer, you automatically violate conservation of energy from the perspective of another observer moving relative to the first one. M Drive: One way to accidentally detect thrust with a torsion pendulum is if the device produces a torque. Of course, in a vacuum, that is equally impossible in standard theories, but a magnet will do just that in the geomagnetic field. I'm quite sure they eliminated the possibility of that interfering with the result by shielding/countering the field, though.

Ah, at least they have videos: http://vimeo.com/cannae/videos The introduction ones I watched. Note that the device is called QDrive there, the name was later changed. It's pretty much identical to the EMDrive in principle. An asymmetric cavity pumped with microwaves. And like the EMDrive, their rationale for it working is purely based on regular electrodynamics. That is wrong with absolute certainty. Provably, the only way to propel yourself in a vacuum with electrodynamics is by emitting radiation, and if you do that, you get a 1/c thrust to power ratio. That does not mean such a device can never work, it just means their theory is rubbish. And if you build a device on a rubbish theory, it's pure mad luck if it works anyway. Hilarious: In video 1, they boast how this would be so much better than traditional stationkeeping for satellites, which takes up to 20% of the full weight and at some point runs out of fuel, after 15 years... In video 3, they describe how to fully assemble the device, crucial component being liquid helium cooling. Guess how heavy that whole construction is and how long cooling can be maintained in space. Sort of. The Alcubierre Warp, if it works, would be reactionless in the sense that it moves the ship without losing mass or energy to the surroundings. As soon as you turn it off, though, the ship stops; the drive makes no permanent change to the ship's momentum. It's using a loophole in GR.

The device tested is NOT the EMDrive, the Wired article is not very clear on that. Instead, they tested the Cannae Drive. 404s on the theory and numerics pages, very useful. From what little can be gathered, they tested the actual device and a null device, almost identical but designed to not produce thrust. Both did produce thrust. With the limited information, my best bet of an explanation would be near field effects. There was plenty of metal around the device (the test chamber, for starters), it operated at around 1 GHz which gives near field effects a range of 30 cm (the wavelength). That should suffice to let the device give itself a good push while pushing the chamber back; the measured force was only 50 micronewtons at best, not enough to lift a common housefly.

Baloney checklist: They do not publish in peer reviewed journals, instead they want to sell you books or DVDs: Check. They bring up Einstein as a person, either to defame him or as an all-knowing sage: Check. (Bonus points if you manage both!) They misrepresent scientific history, especially attributing motivations and feelings of scientists they have no evidence for: "They did experiment X; they could not believe the results, so they did experiment to Y and were very surprised they got the same result!": Check. "These simple assumptions explain the whole universe!": Check. Providing zero testable predictions: Check.

If stellar black holes are very small by your standards, then yes. Those still rip you apart long before you cross the horizon. You may just be able to make out their gravity lensing effect with the naked eye if you fly closely past one, the stars would twitch a little bit.

For a classical black hole, all of the inside apart from the singularity is empty (curved) space. If quantum effects are allowed, they may reach out a bit outside of the horizon and fill the entire inside even for big holes. Nobody knows. As you may imagine, testability is somewhat of an issue here There is, for example, the Firewall hypothesis that has a radiation flash near the event horizon that fries you right as you enter.

That's a good general idea. However, something much wider and/or longer than the device itself would be even better, it has a better inertial moment to mass ratio. I'd try the longest curtain rod I could find and, if it's not heavy enough on its own, attach small weights to each end.

That formula appears to be correct. An important number to extract from this is L0 = t/(d*h); if h = L0, then m = mh. When I plug in your numbers into google's calulator, I get this: L0 = 5000 km So a 5000 km cable can just about support its own weight, just what Wikipedia claims is needed. You probably got your units wrong somewhere. Densities in g/cm^3 always confuse me, I know that much. Additionally, you can make the cable even longer by making it thinner at the bottom; the diameter will grow with e^(h/L0) (it's more complicated for the actual case where you want to go up to GEO, but similar enough. Edit: actually, it is just as simple. Just plug in the effective height in for h instead of the true distance to solid ground.). Nasty exponential! That means that a small difference in tensile strength of a factor 2 can mean the difference between "feasible" and "absolutely impossible". We we don't know the properties of longer, woven fabrics of carbon nanotubes yet (or, for that matter, how to produce them), so it's too early to tell which side they fall on.

Number checking again. One climber, 10t, 10m/s eats up 1 MW while under full gravity. That's about what a medium-large train uses (peak power, of course, as opposed to continuous), so not all that much. No need to plonk down a nuclear power plant, though you probably do want a couple of gas turbines nearby. PakledHostage: A million tubes gets you a ribbon a couple of cm wide easily. If a space elevator is possible at all, a climber should scale down to be usable with that. It'll be small, of course, but it would not need to be big. It just needs to carry up more ribbon and mass for the counterweight above. Ideally, it would "stitch" the new ribbon up to the one already in place to make it wider and wider, but if that's not possible (it probably isn't), you can first pull up many smaller tethers and then finally use multiple climbers in unison to pull up the real tether to use later. If it is possible to build a small elevator (big if), there is very little reason to assume you can't use it to build a bigger elevator.

About 50 megajoules per kg for the whole trip (speed cancels out, naturally; I estimated by calculating the amount of energy required to get the mass up to 10km/s.), that are about 14 kwh. Production costs for that are in the 1$ range today. So several thousand for a multi-ton climber, ignoring that you may be able to reclaim a large portion of the energy on the way down. And earlier: A single carbon nanotube? About half a gram, tops. You don't need that many of them for the boostrap construction. A million should suffice fine. Lesson: when using "do you have any idea how much..." as an argument, make sure the answer actually supports your side. Yes, everything else is still a big problem. Will the tether be strong enough? Can we actually produce a suitable material at reasonable cost? How do we get the energy to the climber? Etc. But raw energy cost and tether weight are not.

The staring position is the one shown in the overview video? With the (very heavy, I presume) gyros to the rightmost position? If so, that would mean the average COM would be more to the left. And horrifyingly, I remembered what else gyros can do (and that drill, too). Their torque can shift the weight between the wires without actually shifting the COM around. So even though you never move the COM forward and backward (camera view), the torque can put the weight of the device to the front and back wire pair. Just like a slowly precessing gyro on a string puts almost its whole weight to the pivot point. Let me emphasize what I tried to bring across previously, poorly: It's really all about the wires. The deflection of the wires determines the direction of the force in the xy-plane they can apply to your device. And if you want to be strict, only the direction, not the magnitude. The magnitude also depends on how the device throws its weight around. If all wires are deflected to the right simultaneously for significantly longer that the pendulum would take to swing, I'll be forced to shut up and concede (or, more realistically, demand you do the same in a vacuum, without the power cable ). To achieve that, you probably need to get rid of the rotation around the vertical axis. Which is difficult. Two right ways to reduce it is to a) build a second device that is an exact mirror image of your first one even in operation so that it produces opposite torques, but thrust in the same direction, and mount both on the same frame and operate them synchronously Suspend the device from a much, much larger frame, with the wires several meters apart (still perfectly vertical, of course) and weights in the corners to increase the frame's inertial moment so it reacts less to torques Neither is really feasible, though to really turn the device into a workable space drive, you absolutely will need to do a) or something equivalent. There is no use having a reactionless drive if it makes your craft spin uncontrollably. Best practical way I can come up with is this: c) Add a long rod to your device, sticking out two meters left and right (camera point of view). Ideally, that rod would just move along its axis and up and down a little. Constrain it to that motion with two pairs of vertical rods on each end, as far away from the device as you can. Yes, this introduces friction again, like your earlier rail system. However, the friction now is proportional to the force between the rods, which in turn is given by the torque your device produces divided by the distance of the constraining rods from the device. So you have a parameter that controls the amount of friction you get: the distance of the constraining rods. Do different runs with different configurations, make sure the friction is low enough you never get sticky friction, and if the different runs produce similar enough results, friction can be reasonably ruled out as source of thrust. Of course, to have the runs be comparable, you need to keep the device on continuously, or driven automatically and not by a manual switch. Lastly, nothing I say is meant to discourage you!

Impressive indeed. I do have to remain skeptical, however, because of two observations: 1. The device turns a lot about its z-Axis 2. The laser pointer is clearly mounted way to the right (from the camera's point of view) of the average center of mass during operation, both the gyros and the drill turning the central axis are to its left The strict and irrefutable version of the pendulum test requires that you attach one laser pointer to each rope pointing in the direction of the rope, that the ropes are perfectly vertical at the start and that during operation, all laser dots are displaced to the left for an extended period of time simultaneously. Averaging the position is somewhat acceptable in this situation, but only if the device does not twist. A single laser pointer is also somewhat acceptable, but only if it is in the center of the suspension ropes AND aligned with the average center of mass (ideally, you'll also have the center of mass not wobble around, but it's OK if the center of mass moves back and forth, not sideways, on timescales small compared to the natural pendulum frequency, which is the case for your device). The reason for these rules is that it is very easy to deliberately cheat without them. Take your exact setup, but instead of the device, I step on the harness. I put my feet close to the two left ropes, facing left myself. I put all of my weight on my left foot and push my right foot forward. The harness will turn (clockwise when seen from above), your laser pointer and the three wires that don't support my weight all move to the left (forward from my point of view). Only the wire next to my left foot is going to move to the right, and far less than the others moved left. I can hold that position for a while, but eventually, the torque from the twisted wires is going to undo the turning. No matter. I shift my weight to my right foot and push the left foot forward. As long as you allow me to push one wire to the right, I can abuse that to push pretty much any reference point to the left (on average). Having a laser on each wire and demanding they all are displaced to the left continuously would catch me. Forcing me to stand in the center of the harness and not lean to the sides would foil my plan. Your device is doing something similar without the malicious intent. Its weight is mostly on the wires on the left (and shifting weight from one wire to the other is not required to cheat, it just enhances the effect). If you look at those in 00029 0019-0054 !!!.MTS, they seem to be displaced to the RIGHT on average during operation, and thus they will apply an average force to the left (compensated on average by the forces from the other wires, of course). I think that is the thrust you would like to see. When it stops raining cats and dogs here, I can try and make a video of me dancing on a swing as described above. Sadly, we have no super-long swings anywhere near, a short one may have to make do.

That's another way to view it, yes. The promised crudeness: Time goes up, reference frame is the center of mass frame. Center of mass is the thick black line in the center. Before the collision, the rods are painted blue, L0 is the length of one of them while it moves freely (already including Lorentz contraction, its rest length would be larger still). At point A, the tips collide. The information about the collision can only spread along the black lines going at lightspeed from there. In the whole blue area, even above A, the rod material is blissfully unaware of the collision. At point B, the information arrives at the end of the rod, only then it can react. But by then, it has already contracted to length L1, clearly shorter than L0. In the yellow area, the rod is heavily compressed, probably oscillating and angry. What happens above B in this picture is simlply the easiest way you can imagine an elastic collision: it's just the time reversal of what happened before. I now can give an example of an "object" that never gets "destroyed", but is partially swallowed up by a black hole. It goes like this: Take a full gas tank. Make it hover above the black hole in a safe distance. Slowly release the gas. Do so that you never fully empty the tank, but always release some (say, let the amount left be m_0 + a/(t+t_0) at time t). The gas will never at any point in spacetime be torn apart and there will always be some left outside. Coming up with a hypothetical more solid material, object shape and movement pattern outside the horizon is left as an excercise for the reader An elastic, unbreakable rubber band you slowly lower into the black hole should do it. Of course, it is debatable whether the infinite stretching at and inside the event horizon the material is inevitably going to suffer from is to be considered "breaking". PS: In case it's not clear: in all of this we need to assume the material in quesion is not actually made of atoms. Atoms will always get torn apart. Instead, we need to assume it is truly continous.

The finite propagation speed of all information, yes. It works in any reference frame: before the tips of the rods collide, they can be perfectly rigid. However, it takes time for the information that a collision happened to propagate back to the rod ends. Until that happens, the rod ends need to keep moving forward, toward each other, at the same speed they had before. When they finally can react (and yes, for an as-rigid-as-possible material, they will directly bounce back), there simply is less space between them than the two rods would require, which means the rods must be compressed at this point.I can make and upload crude time/space drawings later.

Think about what would happen if you let two of your rods collide end to end at relativistic speeds. Think about how an observer stationary with respect to the center of mass sees the event and one observer sitting on the far end of each rod. From the perspective of the rod riders, the colliding ends will have to move and give way in reaction to the collision, no matter what.

Correct. "Inflexible" is incompatible with special relativity already, there can be no rigid bodies if all information about impacts and forces can only travel at the speed of light. One can theorize about indestructible materials, though, they just need appropriate reactions to deformations. Also, mind that for an outside observer, nothing ever passes the horizon. So while from the perspective of the bit of the object that has passed the horizon, the rest is inevitably pulled in after it*, from the perspective of the bits outside, that may very well happen only after infinite time has passed. Whether that's the case may depend on the specific properties of the hypothetical material, though. *: Edit, thinking a bit more, it also can't be ruled out in general that from the perspective of every bit of the object that crosses the horizon, it hits the singularity (and not even fantasy material survives that) before all of the rest is pulled in. Again, this may depend on the material properties.

That is a good question. The answer appears to be yes. Rationale: 1. We know that locally, the horizon is nothing special. No infinite tidal forces appear to a local observer. So the object will not be torn apart there. For large holes, that is true even for regular materials. 2. We know that anything going inside (or past) the horizon can't come back. Since the part "dipping in" can't come back and the object will not be torn apart, it follows that the whole object is swallowed. And it doesn't matter how large the object is. What pulls the bits far outside the horizon in is not the (supposedly) infinite force from the bits inside the horizon, that force does not exist (the bits inside cannot have a causal effect on the bits outside), it's the infinite force from the bits very close to the horizon on the outside. What you need to understand here is that the horizon is not a surface like a wall in your house; it is 'lightlike'. Locally, as you cross it, the horizon swoops over you at the speed of light. Therefore, to get away from it once you gotten too close, you essentially have to run away from something moving at the speed of light. Which is possible with finite acceleration. Even in flat space, you can run away from a wall of light chasing you by applying finite acceleration for the rest of your life, and in the black hole case, the curvature of space allows you to get away completely after a bit. But: The closer you are to the horizon, the more total delta-v you need, and it diverges to infinity as you approach.

Dunno about you, but for me, "walking" implies "touching". You would be touching the ring from time to time, every half orbit at least; no orbit can stay in the "upper" half space indefinitely. If you are careful*, you could skip along (but one step every hour or so is not walking). *If you are not and the ring is fully solid and rotates rigidly like a record, you would get flung off either on the inside or outside.

Orbits don't work that way, they are not static things that force your trajectory. Assuming your solid ring and no unpleasant collision effects, as you push "down" on it, the ring pushes "up" on you, changing your orbit until it aligns with the ring. At that point the pushing would stop and you would just float alongside the ring.

Humans are animals, too. With a slightly better food supply management, granted, but they would still all starve before they suffocate. Basic estimate: Imagine you burn all organic material, calculate how much CO2 that is going to release. It's not enough to get to toxic levels. And no matter what you do, no biological process would be able to release more CO2. If you want suffocation to be a risk, have your virus "reverse" photosynthesis so it extracts oxygen from the air somehow (and faster than animals would do it) or modify it so that all plants now produce a more toxic waste product, like cyanide acid. Or just pick a number and apply suspension of disbelief.

Would they still magically be alive and edible? Because if not, all animals would easily starve way before carbon dioxide poisoning (let alone oxygen deprivation) becomes a problem, and then only geological processes are left to influence the atmosphere.

Not sure what you mean by "quantum levitation" (Meissner efffect?), but if you want to essentially do maglev without a track, the answer is no. To levitate magnetically you need inhomogeneous medium-strong external fields. The geomagnetic field is too weak and locally too homogeneous. In a homogeneous external field, a magnet can only produce torque, not force.

A round trip to 20 asteroids, maybe. No, I don't know what that would be good for, either. Gathering detailed data to select them and prepare future mining missions?

Yep. And if you apply path integral formalism to calculate the S-matrix from one spacelike slice to another with a noncausal region with closed timelike loops in between, your result is not unitary (unless your fields are free and non-interacting). And not just in the black hole paradox kind of way where quantum information is lost, but in the sense that probability is not conserved. That is a bit of a problem and none of the suggested resolutions I am aware of are particularly satisfying. Specifics would be for another thread, I sidetracked this one enough.I found some good places to start in the meantime, for example this old thing by Kip Thorne. Walking the quotation graph should get me up to speed. The last four references are about the loss of unitarity. No, not really. Renormalization deals with local problems on small timescales and lengthscales, and as you correctly say, non-renormalizability is not a problem for an effective field theory. So what if you need infinite parameters to reconcile theory and experiment as long as finite parameters suffice to reconcile theory and all experiments you are interested in to good enough accuracy? The problems with closed timelike loops, on the other hand, are nonloclal problems on the lenght and timescales of the loops in your chosen metric. And yeah, I'm thinking about QFT on a fixed background metric, possibly including gravitation on the perturbative level, but that is not a requirement.

Interesting, can you point me to some material? All the prescriptions I am aware of require a time-ordered slicing into spacelike hypersurfaces before you can apply QFT. And if you ignore that and apply path integral formalism naively to a spacetime with closed timelike loops, you get nasty infinities from all the particles that can inhabit those loops. They are somewhat similar to the infinities from virtual particle pairs, and in some special geometries, you can get rid of them in similar ways (first google result, not read yet), but AFAIK not in the most general case. Edit: Single particles without interaction would be fine, that much I know. As long as local time development is unitary, global time development can be unitary too (I'm talking about the case where the closed timelike loop only exists temporarily and there are spacelike hypersurfaces in its past and future you can use to even define global time development on); simply put, your particle can traverse the loop an arbitrary number of times and you sum up the probability amplitudes. But that no longer works once interactions are allowed.