Enthalpy

You expected it: light rocket structures make Single-Stage-To-Orbit (SSTO) possible, so here's an example. The single RS-68 is widened to D=3m (weighing 6.9t hence) but has a temporary restraint insert to maximize thrust at lift-off. An added deep throttle mode around 20% thrust would be very useful. Lift-off mass is 234t, leaving 23.1t at propulsion end, or 9500m/s performance for Low-Earth-Orbit (Leo). The central tank contains oxygen. D3.2m H20.5m made of AA 7022 profile, t1 = t2 = 1.5mm, a = 45°, with the extrusion direction tangential. Walls weigh 2.9t, heads 0.2t, foam 0.1t. Six lateral balloon tanks contain hydrogen. D2.0m H18.8m made of hot-rolled and welded Maraging steel, 170µm thin. The engine pushes their base. They hold only the internal pressure: 1,0b+0,5b+0,2b. Each skin weighs 193kg, its foam 71kg, six tanks total 1.6t. Thin aramide can cover the foam at the top, but something must prevent an impact during the fairing's separation. This leaves 11.4t for the payload, adapters, fastening rods for the side tanks, electronics, leaving around 9t payload. A Soyuz at 7.5t would even afford some heat shield and parachute to recover the engine if it were reusable. Tank mass is only 23kg per ton of propellants, and the payload is 3.8% of the lift-off mass, nearly as good as a multi-stage launcher. If this stage were to serve as a side booster, it would have two bigger hydrogen tanks instead of six small ones. Marc Schaefer, aka Enthalpy

At the oxygen tank of Delta IV's Common Booster Core I gave as one example, hydrostatic pressure produces a stronger stress in the walls than thrust does, so putting the profile's extrusion in the tank's tangential direction is better. The profile can consist of AA 6082 with t1 = t2 = 1.6mm and a = 45°. The booster's wall weighs then 8.1t instead of 5.9t, but if keeping the thinner axial extruded profile at the hydrogen tank only, the wall weighs 6.6t. To roll the extruded profile to the tank's radius, one known solution is a temporary filler in the profile. Other prospective ways: The outer skin could be warm and the inner cold when rolling. This could be done right at the extruder, by cooling one side first. Could the extruder's die produce a curved profile? Electrical discharge can machine a curved die. In both cases, the profile could be stored as a coil immediately when produced. An extruder bigger than usual might produce all wall material for a rocket stage in one step. Instead of storing a coil, the profile could be welded immediately into a wall by a continuous process. Some alloys are better welded hot. Later weld joint corrections are possible with the healthy alloys cited. Marc Schaefer, aka Enthalpy ================================================================== In some cases, profiles lighter than t1 = t2 = 1.6mm would be preferred. Extruded profiles exist with t = 0.25mm. They use to demand weak alloys like AA 6063 and small lateral dimensions. The extruder's force seems to limit it, not the alloys' capability, so a specially built machine may produce thinner skins of stronger alloy. Open profiles, having for instance T-shape stiffeners on a single skin, are easier to extrude, need less material and can be thinner. Such panels made by extrusion demand additional stiffeners in the other direction: a rocket tank can have tangential rings and axial extruded panels. While I trust them far less than closed profiles at rockets, such open profiles build already the fuselage and the wings of most aircraft; some of them might be extruded, to save money and maybe improve reliability. Marc Schaefer, aka Enthalpy

Nearly all launchers have isogrid tanks. They do so to avoid buckling and nothing else. Are you sure you're speaking about liquid rockets? Since liquid rockets with composite tanks are extremely uncommon. I should have made it clear at the beginning of the topic: it's about liquid propellant launchers, not solid missiles, which is a different world.

Buckling IS the only reason why rockets have isogrid panels, added rips and the like! Stiffness against axial bending, longitudinal compression, and others, would be best overcome with plain sheet. The corrugations, or zigzag, or call them as you want, are vertical. They follow the extrusion direction which is parallel to the rocket axis. Though, we could wrap the extrusions in circles or in a helix (possibly with several threads) in some cases where pressure in the tank creates more stress than acceleration does. No worry for the few G then, their direct effect on the sheets is really negligible. The most common shape (for 40 years) is an isogrid. From a thicker aluminium sheet, a milling machine leaves thick ribs on a (generally triangular) pattern, and thin sheet between them. The resulting panels are welded together. And, my mistake, T-profile ribs exist for decades: http://en.wikipedia.org/wiki/Isogrid http://femci.gsfc.na...grid_Design.pdf or Google isogrid, then select "pictures" which illustrate my obscure formulation of non-uniform thickness. That is: to obtain flexural stiffness from a given material volume, you better concentrate matter at narrower thicker places. And dome-shaped honeycomb as well has been around for a long time.

Ask on a computer forum like Hardware.uk or .de or .com. People who overclock their CPU have such closed-cycle refrigerators, which can't be very expensive. They want seriously cold temperatures because CMOS components are faster at cold (record overclockers let them run at 77K+), and if your component heats less than 100-200W, the refrigerator will achieve a lower temperature than with a CPU.

A magnetic field with negligible electric field varies very slowly, and then the oscillation has weak effects. The effect on protons is small, like 50%+1ppm of the protons taking the favourable direction in a 1T field at room temperature. The direction if these protons can the fluctuate at a frequency that depends on the polarizing magnetic field. Measuring the frequency gives you the magnetic field, and the attenuation of the oscillation tells the influence of the nearby atoms. Both effects are used to measure hydrogen abundance, field intensity, and make maps of images of hydrogen, for instance in the human body by Magnetic Resonance Imagery. At the electrons, it's more complicated. They would respond more to magnetic fields than protons do, but they are paired in all normal materials, especially near hydrogen nuclei. A few materials have unpaired electrons and are used for instance to make oscillators whose frequency is controlled by a current in a coil, but none uses hydrogen as far as I know. You'd already need to produce atomic hydrogen for that, and it takes exotic conditions, like 3000K at low pressure.

A few arguments against antiparticles repelled by gravity: - Their inertial mass is absolutely normal. Shall their gravitation mass differ from the inertial one this time? - Intense gravitation fields, like the horizon of black holes, would create pairs and eject streams of antiparticles and swallow the particles. - When a pair is created from energy, some mass+energy would disappear. And why would the gamma ray need a minimum energy to create a pair? Experiments do isolate antimatter for some time! Antihydrogen has been produced and stored for long enough to measure its optical spectrum. Maybe weight could be observed in these experiments. Thanks to have cited the source of this strange idea...

Oops. The extruded panel for the Common Booster Core would have a=60°, not 30°.

Hi Matty and all, thanks for your interest! That's perfectly true, many stiffeners have been invented, tried, and several types are used presently. They all make the panels hollow, to get thickness hence stiffness without the weight of a solid panels (that's because stiffness varies as the third power of thickness, so a non-uniform thickness is better). In the nice link you give, fig 1 type is used at airframes. A reliable bond between the skin and the sheet stringers is difficult to obtain, and I suppose this bond would require too much work for a launcher used one or few times. Though, inter-stages use often stiffeners of folded sheet, rather in an omega shape than a T. The sandwich panel at fig 3 has even higher performance because it puts most material at the extreme locations of the thickness. Because of accessibility, the skins are glued to the central honeycomb. Glueing is mistrusted in rocketry - much more so than welding - and it makes difficult to weld such panels to other parts, because glue burns at heat and pollutes the joint. The bond between the skins and the honeycomb is little stressed and glue is good enough for it, but for a tank, the panels (the skins) must also be bonded with other parts like tanks heads and inter-stages, strongly and tight, and I know no good solution for it. I haven't seen cylindrical honeycomb panels but they probably exist. Shapes neither cylindrical nor conical are much more difficult and I doubt they've been made, so the tank heads would use a different material. One further difficulty at cold tanks is the dilation difference. When the tank is full of liquid oxygen or hydrogen, or is being filled, near parts of the panels have very different temperatures, and I distrust glue under such conditions. Yet alone cold is bad for glues. For that purpose, a milled panel, or I say a welded one, is safer. It's done on combat aircrafts, where the wing's skins have "integral stiffeners" in one direction (a single part, 10m*20m, milled by a machine bigger than that, nice toy...), and on rockets, as on fig 2. As an improvement over fig 3, the rips are better on a triangular pattern, making an "isogrid". Easy now with CNC milling machines. This isogrid may improve a bit if we give the milled rips a T shape. This is possible with the cutters that produce woodruff keys (or better, an optimized shape with central cut): http://en.wikipedia....utters-Keys.jpg http://en.wikipedia..../Milling_cutter but milled shapes must still be open where the cutter plunges, limiting the rips' resistance against buckling. I have confidence in weld joints for having welded by myself aluminium alloys without worries. I suppose aeronautic and space designers distrust welds because they use unsound aluminium-copper alloys like the AA 2219, but the AA 6005A and AA 6082 are really safe, and the AA 7022 allegedly as well. Hence, I'm happy that the extruded profiles can be assembled by welding. Further, extrusion puts the welds in a little stressed direction and can make the material thicker there, nice. Marc Schaefer, aka Enthalpy

Hello everybody! Rockets must be light to attain a high speed, but their thin structures, pushed by the engines, are threatened by buckling. Usual tank construction includes sheets milled down to an isogrid structure, or omega stiffeners welded on a sheet. I propose to assemble rocket structures from tailor-made extruded profiles instead. http://en.wikipedia.org/wiki/Extrusion Extrusion can produce thick profiles consisting of thin walls that build closed channels. Such walls are well supported hence strong against buckling. The extrusion direction parallel to the rocket axis has advantages. All walls bring strength against rocket bending moments and axial compression. And since extrusion companies can deliver parts 30m or 50m long, a complete stage can consist of extrusions assembled only side-by-side, with little stress at the joints. Many materials can be extruded. These aluminium alloys are interesting: - AA 6005A. Yield>215MPa, Young 69.5GPa, 2710kg/m3. Easier to extrude, keeps its strength at weld joints. - AA 6082. Yield>250MPa, Young 70Pa, 2710kg/m3. Still easy to extrude, keeps a good strength at joints. - AA 7022. Yield>420MPa, Young 72Pa, 2760kg/m3. Harder to extrude, loses some strength at joints. http://aluminium.mat...ect/default.asp These alloys stay ductile at cold, even 20K. Their weld joints are ductile even with low-tech methods like TIG or MIG: easier, safer. AA 7020 would even regain its full strength after a week at room temperature. Extruded profiles exist with walls <0.25mm thin, width >700mm, hard alloy like AA 7075... To my understanding, each of these parameters just increases the necessary extrusion force, so I avoid to demand all extremes simultaneously. And since an extrusion machine costs a few 100k€, it can be made specially for the rocket production plant. More to come, including drawing(s). Marc Schaefer, aka Enthalpy ================================================================== Here the structural skin of a stage consists of extruded profiles welded side-by-side: The view is parallel to the rocket's axis and the extrusion direction. A propellant can be at the inner side at tank sections, or something else at inter-tank sections. Any insulating foam would probably be outside the skin. Depending on the extrusion machine, the width of a profile can be 200mm or more, so the profile can be curved or flat. A D=5m body would consist of 80 profiles for instance, welded side-by-side on one tank height, or better, on one stage height. If you feel this is much joint length: - These alloys like welding, as opposed to AA 2219 - The process is automated, and the quality of apparent joints is easily checked - Think and the joint number and length at omega stiffeners - Don't travel by ship The profile can be thicker at the joints, build a groove where the joint will be, and have keys to hold the parts precisely during welding. Very useful at least for manual TIG process. The keys also protect the joint's rear side against air when welding. Welding things like tank heads, interstage adapters, engine thrust rings... on the profiles weakens them a bit, but local reinforcements can compensate this. Reinforcements parallel to the extrusion direction and fitting between the profiles (machined narrower there) lose no strength at the profiles and take the full stress over, to cross the skin's weaker position. Marc Schaefer, aka Enthalpy ================================================================== Let's take a Zenit first stage as an example. The RD-171 engine pushes 7.9MN, the stage has D=3.9m, h=32.9m and carries 327t of kerosene-oxygen. AA 6082 can have t1 = t2 = 3mm, a = 45°, B = 40mm and b = 34mm. Then it withstands - without any safety factor - 31MN compression and 30MN*m bending moment. This skin weighs 11.1t. Better: AA 7022 with t1 = t2 = 2mm, a = 45°, B =44mm and b = 40mm. With no safety factor, it withstands 35MN compression, 34MN*m bending moment, and weighs 7.6t. Tanks heads may add 0.5t, foam 0.2t if any, the engine 9.8t. Even with the engine-to-skin transition and the interstage, this can be lighter than the original 27.6t stage dry mass. ----------- More difficult example: Delta IV's Common Booster Core. The RS-68 pushes 3.3MN. D=5.0m and h=34.9m (cylindrical section) to carry only 200t of bulky hydrogen-oxygen. AA 6005A shall make t1 = t2 = 1mm less difficult. a = 30°, B = 34mm and b = 32mm. With no safety factor, it withstands 13.5MN compression, 17MN*m bending moment, and weighs 5.9t. Add 0.7t for the tank heads and 0.3t foam, and the tanks weigh 35kg per ton of propellants. That's as little as the Shuttle's external tank, which didn't have to withstand the push of a first-stage engine. Marc Schaefer, aka Enthalpy

Maybe if some kind of gold oxyfluoride is soluble? Chances are slim, but it worked with silicon, at the time semiconductors were processed in liquids - that's paleotechnology now. HF is known to dissolve SiO2 by producing a soluble silicon oxyfluoride; to etch silicon (nothing trivial) we used HNO3 to first oxidise it, and HF to "dissolve" the oxide.

I do agree that demagnetization is a story of the past. AlNiCo was plagued with it, FeNdB isn't. FeNdB can be put front-to-front so that zero flux passes and they don't lose strength. No idea where the often cited figure of 400 years comes from: manufacturers give no limit. I'd say varying fields won't do anything, because alternators let Nd magnet work in varying fields (created by the induced current in the stator windings) and they keep their strength. From the distances observed on your video, your varying component must be smaller than theirs, and at wind turbines, the induction varies around 50 times per second over decades (1.5e9 times a year). Any loss must be hard to predict, since manufacturer give no data about it. In case of doubt (I have none) or if you need a proof for someone, I suggest to experiment it with a couple of magnets that you rotate with a motor. At 1500/min over 46 days, you have your 100M passes. The best method I found to measure the magnetization intensity of magnets was to hold them with a thin wire so they rotate freely, and measure the oscillation period in a known field. Earth gives one for free, or Helmholtz coils give a stronger one - add and subtract Earth's field over two measurements to compensate it, just use the squared frequencies.

Could they radiate little once their matter has reached a near-equilibrium, and light only when new matter falls down from a direction out-of-plane?

Under the "normal" conditions, no influence at all. Then, you can use heavy means. If you bombard 14C with 10TeV protons it will break, but (*) that's cheating (**) it's not radioactivity (***) any nuclide would break. Somewhere in the middle: - One beta minus emitter (forgotten which, the experiment wasn't with 14C) decays much faster if the atom is completely stripped from its electrons. The emitted electron jumps to the then available 1s orbital, through a tunnel length and depth more favourable than to vacuum. - An absorbed gamma ray that excites 14C will necessarily change the decay rate. Half-cheating again, as people consider the excited nuclide is no 14C any more. Not at beta emission, but electron capture: one paper claimed an increase in electron capture rate at sonoluminescence, but this has never been observed by the other teams and is widely considered a mistake.

Hello you all! Rocket engines often pump liquid propellants in their combustion chamber, and many cycles are used to power the pumps, the best known being: http://en.wikipedia....le_%28rocket%29 http://en.wikipedia....le_%28rocket%29 http://en.wikipedia....le_%28rocket%29 but here I'd like to describe uncommon cycles, which may be new and of my invention. ======================================================================== The first sketched cycle cracks in a pre-chamber a hydrocarbon or an amine with hydrogen, which produces methane (chemists call this hydrogenolysis), some excess hydrogen, and enough heat that the following turbine powers all pumps that achieve for instance 440b in the pre-chamber and some 200b in the combustion chamber. Logically, fuel density and specific impulse are between methane-oxygen and hydrogen-oxygen, thus filling the gap between kerosene and hydrogen engines, as the diagram shows (click to magnify). The pre-chamber and turbine run with fuel-rich hot gas, far easier than the oxygen-rich staged combustion cycle which was needed to avoid soot with hydrocarbon fuels. Or if one oxygen-rich pre-chamber is kept, hydrogen pumping can go through a separate cycle like the one just described, in a kind of full-flow cycle that achieves a higher pressure but would not soot with hydrocarbons. Maybe an adaptation from existing engines like the RD-180. More details to come, as well as other cycles. More lengthy details, in a less ordered fashion, at an earlier thread there: http://saposjoint.ne...start=60#p34195 and followings, on Sat Sep 24, 2011 7:05 pm and Sat Sep 24, 2011 8:09 pm and Wed Sep 28, 2011 10:43 pm Marc Schaefer, aka Enthalpy ======================================================================== Now, this other cycle shall decompose an endothermic oxidizer in a pre-chamber to obtain hot gas for the turbine: It shall be simple and reliable: - No mix is required at the pre-chamber, which runs at a safe fixed temperature (but in oxidizing gas, true) - Pump and turbine speed is moderate - If you have some limited pressure in the tanks, opening two valves starts the engine, and the attitude control can be pressure-fed - The oxidizing gas lights the fuel by its mere temperature - The moderate decomposition temperature permits a sort of glow-plug igniter This cycle burns storable propellants and is more efficient than tetroxide and toxic hydrazine, like 350s with a good expansion, and its tanks are lighter. One oxidizer is Mon-33, or 33% NO dissolved in 67% N2O4. It freezes at -107°C, so if paired for instance with 2,4,6-trimethyl-tridecane (freezes at -102°C), they stay indefinitely on Mars or an asteroid or a Moon just in white tanks. Less NO lowers the vapour pressure at terrestrial temperatures, lowers the pre-chamber temperature, and loses little performance. Again, a messy earlier thread contains more details: http://saposjoint.ne...6&t=2272#p27535 Mon Jul 05, 2010 2:39 am (cycle description) http://saposjoint.ne...start=40#p30642 Sat Feb 19, 2011 3:04 am (improved cycle figures) http://saposjoint.ne...start=20#p28807 Sat Sep 18, 2010 9:48 pm and Sun Sep 12, 2010 5:28 pm (igniter) I had considered hydrogen peroxide H2O2 as the endothermic oxidizer, but the RD-161P uses it already, what a shame: http://esamultimedia...ocs/EMO/LPE.pdf so here are the reasons that make this choice bad: - It's less efficient than oxygen-kerosene - But peroxide isn't storable neither - And peroxide is seriously dangerous HandlingPeroxide Marc Schaefer, aka Enthalpy ======================================================================== And that other cycle recomposes in a pre-chamber a mix of amines to produce methane, nitrogen, little hydrogen, and heat: Few amine mixes don't soot. I consider Ethylenediamine dissolving 358:1000 of Guanidine. The pre-chamber, turbine and pumps work then at comfortable temperature and speed, and the hot gas is fuel-rich. The sketch suggests an optional pressure-fed attitude control, but since only oxygen gives good performance here and is more difficult to store, this cycle would rather fit some launcher's lower stages. Attitude control uses to gimbal engines there, and booster pumps accepting a low input pressure save tank mass. This cycle is as good as oxygen-kerosene in a staged combustion but far simpler. Previous description, less ordered but more complete: http://saposjoint.ne...6&t=2272#p27477 Sat Jul 03, 2010 4:45 am http://saposjoint.ne...start=40#p30666 Tue Feb 22, 2011 1:02 am Marc Schaefer, aka Enthalpy ======================================================================== A few rocket engines decompose a third propellant just to power the turbo-pump for the main propellants. Slightly old-fashioned (hi Wernher) but well-proven, for instance by Soyuz. http://www.lpre.de/e...D-107/index.htm The standard choice here is hydrogen peroxide, which is dangerous (link above). My suggestion is to recompose an amine instead, to produce hot methane, nitrogen and possibly soot. The previous Ethylenediamine - Guanidine solution would work, but I believe soot is acceptable here, and then very safe amines are good: DETA (diethylene triamine NH2-C2H4-NH-C2H4-NH2) TETA (triethylene tetramine NH2-C2H4-NH-C2H4-NH-C2H4-NH2) TEPA (tetraethylene pentamine NH2-C2H4-NH-C2H4-NH-C2H4-NH-C2H4-NH2) Gas temperature is mild and constant. Expansion speed is better than 82% peroxide, without the dangers of peroxide or hydrazines. More details there http://saposjoint.ne...start=60#p34071 Thu Sep 15, 2011 9:09 pm Marc Schaefer, aka Enthalpy

Plants are little efficient at creating biofuels, under 1% of the incoming Sunlight. So while plants make a good use of their production cost, their use of farmland can be improved a lot. Up to now, untaxed ethanol or oil or biodiesel from farming are competitive against taxed gasoline. A truly sustainable scheme would need biofuels to be as good under fair conditions. (Please don't misunderstand me, I use biofuels). Our farm plants are optimized for food production, so better choices will improve this quickly. Maybe single-cell plants are better at producing chemicals. But if ATP or some part of a plant works better than a complete organism, fine! Very useful as well: if we could feed them with seawater, and grow then in glass tubes in deserts. Taking only unused space, they would just need to be cheaper than Solar thermal energy for instance, which is stored and produces affordable electricity ("affordable" is not: "as cheap as coal").

Since my post of 24 September, I've looked again at dispersion in coaxial cables. They won't propagate a lazy step slower than a steep one. But the 100m (or 116m? As they claim 580ns delay) thick cable used at Cern must widen all steps by something like 15ns, or by about 50ns is the cable is a thinner one, and delays the slow signal additionally by that same value. Whether this additional delay was computed in is unclear, but rather probable. An other source of dispersion, probably worse, and more difficult to estimate, is the transformer that measures the beam's proton current. This one can deform the measured signal a lot. More details there http://saposjoint.ne...&p=34212#p34208 To overcome both sources of error, I suggest to measure the amplitude of the oscillation at 200MHz of the beam's current instead of the mean current, as this oscillation around a narrow frequency band is less deformed by the transformer and the cable. It needs a filter around 200MHz, a radiofrequency detector, a lowpass filter. As the measurements must be recorded digitally, these operations are best done by software. If the 200MHz oscillation wasn't recorded then, it can still be done just to compare the now deduced envelope with the one previously used, to search for a possible time lag. Here under is an example of a 200MHz filter with the 1000Msps sampling rate, it's a crude FIR. Apply it several times to filter better, or design a better filter. Detection can be done by squaring each sample. Post-detection filter can be a sliding average of 5 samples, done for instance three times. Few more details there http://saposjoint.ne...&p=34212#p34209 Or take time to have a signal specialist do it properly. Or use some standard software for it. Marc Schaefer, aka Enthalpy

Between the current transformer that observes the proton beam and the digitizing instrument, they have a 100m coaxial cable which explains well the signal attenuation at 200MHz and, according to my first hand estimates, a shift of the signal timing. That is, the cable introduces a delay of 580ns. Any reasonable experimenter would use a steep wave front to measure this delay precisely, but this gives a delay for steep fronts, or for high frequencies if you prefer. Unfortunately, cables are dispersive; less so than wave guides, but the 500ns wave front at the measurement signal is at unfavourable frequency, bad luck. This slower wave front propagates a bit more slowly than a steep wave used previously to characterize the cable. When estimating the neutrinos' time-of-flight, the signal from the current transformer (at the source) arrives at the digitizer later than believed, giving an estimation shorter than it should, and a speed too high. I'll come back when I have harder figures. Marc Schaefer, aka Enthalpy. ----------------------------------------------------------------- I suggested to compute a correlation between the current transformer signal and the neutrino detector's signal around 200MHz only, but maybe the neutrino detector doesn't have this bandwidth, and the current transformer signal wasn't recorded over years with such a sampling rate. There are solutions without a fast neutrino detector, apart from checking the cable's propagation delay at varied frequencies using known - though not usual - electronics techniques. For instance, a piece of software could work on the wide-band digitized signal from the current transformer, keep only a band around 200MHz (a simplified FIR avoids additional time lag and runs in near real time on a PC), and perform an envelope detection at 200MHz. This signal resulting from amplitude detection won't suffer from low-frequency dispersion in the cable. Then the new signal can be compared with the previous LF signal from the current transformer to chase a time lag between them (doesn't need wide-band records over past years), or be used to compute the correlation with the signal from the neutrino detector. Marc Schaefer, aka Enthalpy -----------------------------------------------------------------

Already a discussion there on this topic http://www.scienceforums.net/topic/60029-tachyonic-neutrinos/

If you read my last post you see I put there the address of the science paper. And I suggested an explanation to the effect because I had found the paper mranwhile.

BUT the Cern's website directs to the scientific paper, on ArXiv hence freely available: http://arxiv.org/abs/1109.4897 Read (too quickly) the paper from arXiv. My first comment is that I'd strongly prefer the propagation time to be measured between two neutrino detectors, one at Cern and the other at Gran Sasso. Presently it's measured between a proton beam current detector at Cern and a neutrino detector at Gran Sasso. As the neutrino beam is 3km*3km wide at arrival, a small detector at the source would provide as many event there for a more direct comparison - err... IF the mu neutrinos can be detected with the same inefficiency as the tau neutrinos are, which I ignore to a high degree of precision. GPS signals are jammed but many techniques, especially differential GPS, overcome it. From the comments in the paper, scientists there obviously know that better than I do and took care of these clock and position measurements, end of paragraph. The signal from the proton beam intensity shows a decline instead of a steady plateau. Could it be that a fuzzy signal from the neutrino detector correlates better with the inclined reference if it's shifted forward, just as a result of the waveforms? Now, things I'm easier with. The 200MHz modulation of the proton beam brings no certainty at all to the discussed 30ns. If the slower beam current envelope, lasting 2µs, could be measured with 5ns certainty, then the 200MHz modulation would improve the correlation precision to about 10ps, which isn't the case here. The measurement relies only on the 2µs envelope. I believe to understand that 200MHz is the frequency of the accelerator cavities, and modulate the beam intensity fully, something like 0% to 200% of the mean intensity. Though, the diagrams on page 6 show only +-15% modulation depth at 200MHz, so even though the beam current transformer and supposedly the acquisition device have a broader bandpass than 200MHz, something attenuates the 200MHz component, be it a medium to long cable or something else. Unfortunately, the thing that attenuates at 200MHz is probably dispersive, that is, it introduces a propagation delay that depends on the frequency. A cable for instance delays precisely at 200MHz by its known speed but gets slower at lower frequency as its series resistance adds to the inductance, and here we're talking about 30ns precision over a 2µs waveform with 500ns transitions - that is, the measurement results from a rather strong statistical interpolation. Hence I wish this possible dispersion be eliminated. Fortunately, this looks easy, thanks to the 200MHz modulation. It just needs to suppress the DC and LF components of the signals, both at Cern and at Gran Sasso, and compare only the tone-burst envelope. It needs a filter around 200MHz, a broad one like 100-300MHz to minimize its propagation time. Over this favourable and limited frequency band, all cables and transformers will show their normal delay. Maybe these filters can be made by the same piece of software, introducing the same delay. The correlation will oscillate at 5ns, but this is meaningless. The envelope of the correlation will be meaningful and independent of LF dispersion. Marc Schaefer, aka Enthalpy

I could drill D=10mm holes in a Champagne bottle, rather easily. I took a drill bit for concrete and made it seriously sharp by grinding. As well, but maybe it wasn't necessary, I made a pool around the future hole, with modelling clay, and filled it with water, to cool the glass and the bit, and absorb the vibrations. I even got thin glass chips. It doesn't take very long. Moderate cutting speed. Small vertical home drilling machine, feed by hand. Mind your health. Cloth around the glass?

The general press reports this story, but I've found no science paper about it. I went to the website of the experiment, there is absolutely NOTHING there about said overspeed. Sorry folks, this is a hoax, relayed by newspapers which don't even make quick checks before publishing.

I read it in the general press as well but found nothing in science publications. I went to the website of the experiment: absolutely NOTHING there. So until I get credible information, this is strictly unfounded. A hoax if you prefer. Regrettably, many newspapers publish it without even checking if an other source tells it.

Density changes radically the process. The orbital energy levels remain nearly the same, BUT: - The time between two shocks gets tiny, so you won't see any emission lines more. They'll get so broad you can't separate them. - The distance between two shocks gets tiny, so it needs a much stronger field for an electron or ion to acquire the energy needed to ionize an other atom and maintain the spark. - The very concentrated spark vaporizes the liquid locally, making the process unstable. In fact, neon discharge tubes use a low pressure to obtain a stable discharge, not just to permit a 100V or 200V spark over a significant distance (only 1/10mm at 1 bar). Sparks in liquids exist and are used, in electrical discharge machining (see Wiki). Very short distance, very destructive.