Enthalpy

"one TEM00 mode but multi longitudinal modes" and "substructure with periodically spaced longitudinal modes" are other words for "superradiance modes". I suppose electrons and holes in a laser diode have energies too varied, so they don't engage all on a common transition energy. Instead, light is created at (around) one energy, this depletes within 150µm/c the corresponding available stocks of carriers, then lasing goes on at a different energy, and so on until the echo of a pulse around one energy comes back and finds again enough carriers to lase. This is compatible with a constant emission power. Only the emitted wavelength hops every 150µm/c. And if carriers are replenished or thermalized quickly enough, more pulses can happen around one energy within a ping-pong delay in the material. As opposed, He-Ne offer a more uniform transition energy, enabling a quiet operation with long coherence.

Separating 3He from all produced natural helium would supply a bit more than the present demand, a handful of kg a year. Though, I begin to guess a different need for 3He, in much bigger quantity... Because tokamaks such as Iter need tritium to run, and must regenerate tritium to be credible, which is a very strong obstacle: sheer feasibility and pollution, as one consumed tritium makes a single neutron needed to regenerate a tritium from a 6Li, so dirty neutron multipliers must offset the losses. I tell it for years, maybe tokamak proponents get slowly forced to look into the problem. One other way would let 7Li absorb the 14MeV neutron produced by the D-T fusion. 7Li splits into 4He and T, and releases a neutron - but with energy too low to react with 6Li nor 7Li. This secondary neutron could react with 3He to make a second T. Well thought guys, but this needs 3He in big amounts, not available from terrestrial helium deposits, and of course not obtained from T decay. This can be a reason for the query at 3He artificial production. It can also be the reason why some people and agencies consider mining 3He "for nuclear fusion" where it is less scarce: on the Moon or from gas giant planets. It's not for D-3He fusion, which is hugely more difficult than the already remote D-T fusion: it's for regeneration of tritium in D-T reactors. Well, sorry folks, but isolation from normal terrestrial helium just makes a few kg a year. And despite being a space visionary, I don't really imagine affordable Selene mines. But for a few millions instead of many billions, we would have cheap electricity storage that enables wind and Solar energy on the big scale.

The terrestrial proportion of 1.37ppm means: process 1,000 tons to get 1.37kg... So one better puts the separation plant at the helium well directly. Liquefying helium in big amounts isn't easy nor cheap. An alternative would be a molecular pump or a gaseous diffusion plant. Better, do both simultaneously: have a many-stage pump (for instance as an axial turbomolecular pump) and put a gaseous diffusion shunt around each stage, if natural leakage doesn't suffice. Each stage is more efficient at 4He while each shunt lets 3He leak better. Because the masses differ significantly, each step has an interesting yield. This can work at room temperature, with the speed attained by an optimized metallic axial pump. Cold improves. Natural helium would be introduced near the "depleted" (purer 4He) exit, so most flow absorbs little pumping power. Enriched 3He would exit at the pump's extreme inlet, after many steps that most helium does not pass. Many other methods must work. Gas chromatography can use many (many) short fibres in parallel, I suppose with oscillating pressure at the natural helium side so that 4He has no time to diffuse to the end of the fibres. I wanted to use a superconducting leaking ceramic to repel 3He as is done with O2, or a ferromagnet to alter the diffusion of 3He versus 4He, but since a nucleon is a weaker magnet than an electron, the effect is tiny at room temperature - much smaller than the contrast in diffusion speed. 3He seems 10 times better than Li to produce 3H from a neutron flux, that is at a nuclear reactor. That can explain the query for 3He.

Your conception of wave and particle is wrong, and this misleads you. A wave is a particle (or several particles) because it can interact very locally if needed, for instance at one pixel of a camera, and concentrate there everything it has. But nothing beyond.

It's nothing technological. Science-fiction if you wish.

All centers in the same direction is impossible, because if one aligns up, its side neighbours will naturally align down. But in ferromagnetic materials like soft iron or hard magnets, the neighbours have different strengths, so that one over two aligned up and the next down results in a net global field. It can also be local electrons up and shared electrons down.

Semiconductors have indices like 12, so phase coherence might - could have - recover within the diode if it's some 0.4mm long. Though, this would hint at a bizarre lasing mode, something like superradiance, where coherence lasts for 150µm but is recovered cleanly after a full ping-pong path in the lasing medium. I'd search at other parts of the experiment, like the splitter, the mirrors... You can decide between the diode and the rest by building a simpler experiment: a resonating cavity without the interferometer. Take semireflective mirrors with antireflection coating, check if you obtain resonances when the mirrors are 4.4mm apart. Or more convincing: make two-slit interferences, offset the diode, observe if you get fringes again when the paths differ by 8.8mm.

A very reflective mirror shouldn't emit IR by itself: these two are antagonist. That is, at one single wavelength, if a mirror reflects 99%, it can emit only 1%. At long IR, most metals tend to be very reflective. Hence I wonder if maybe you aluminium is behind the glass, or if the aluminium has grown a too thick oxide (heat worsens that), or has an added protective layer that radiates IR. If aluminium, blank and at the first glass' side, doesn't reflect enough, you can try gold, which oxidizes less when warm.

The propagation speed tells how far the ripple reaches at a given time, and this makes a circle. The stone's shape has some influence by telling where the ripple starts, but once the ripple is much wider than the stone, this influence is small.

Thanks for the links! The first, called ATAC design, uses wavelength division multiplexing to send messages on a single light carrier that joins all nodes. - It is limited to ~100 nodes by the carrier separation - Good throughput thanks to the fast modulators - The authors give a chip area for the modulators and the photodiodes but NOT the filters, which I feel just dishonest. WDM filters are typically "racetrack" loops which must be long to separate close wavelength. Just 64 nodes need 64*64 big filters, and the authors hide this. - Over 100 nodes, the connection matrix cannot be complete. In contrast, silicon switches make 1000 nodes with one chip and are expandable. I can't access the payfor second paper. The abstract tells "network partitioning options" which sounds opposite to "full connection matrix". The third, again the ATAC design, describes a 100 cores paperwork but grouped as 64 clusters of 16 cores, that is, not a flexible full matrix. It's a hierarchical design with all associated bottlenecks. As compared with these designs, a simple silicon chip with N2 switches or gates to route the data is better: uniform routing, throughput, absence of interlocks and constraints...

With electric signals on silicon, we can make the interconnection matrix, full, and reacting at the speed of individual data packets. The power consumption for a 500*500 switch matrix fits a single silicon chip - only the number of in-out pins limits the chip. Do you know a routing method for light that is as good?

75 ohm give the minimum loss for a coaxial cable with full polyethylene dielectric. The same "e" optimum ratio of conductor diameters gives 93 ohm for coaxial cables containing little dielectric. This impedance was not chosen to match any particular antenna, and without proper adapters generally it doesn't. Fet preamplifiers have an input matching network, but their best match for signal-to-noise ratio does not match source and load, which only gives the strongest signal but not the minimum noise. Bipolar preamplifiers are a bit mismatched, Fet more so. A different cable impedance wouldn't improve: the optimum matching network would always give the same mismatch ratio, because this mismatch equals the mismatch between the best impedance seen by the Fet for signal-to-noise and the input impedance of the Fet. The power lost at mismatch if using 50 ohm instead of 75 ohm is miniscule, and certainly not competing different losses per unit length in the cable. Yes, the feed impedance of an antenna varies a lot over a wide frequency band - except for very special designs like logperiodic, but these are bad in every other aspect. An Uda-Yagi performs well over one octave, say 470-860MHz, which is already quite good and unexpected if one considers it as a group of resonating dipoles. Obviously, the operation of an Uda-Yagi needs a different explanation compatible with its wide band, and a guided wave is a better explanation. Over a reasonable bandwidth, a director gives a more constant radiation impedance than a reflector, and also a bigger one that is easier to match with the cable.

Or the planets with circular orbits are more difficult to detect because of their physical caracteristics. But I don't see why. In our Solar system, we have Mercury whose orbit is seriously eccentric. Mars isn't very circular neither. Most minor planets have eccentric orbits, with Pluto being an extreme case.

Some celestial bodies can be synchronized. For instance, some asteroids share Jupiter's orbit and are 60° before or after it. I've not heard of 180° positions. Some moons of Saturn have synchronized their periods (1:2:4 ratio from memory). But I don't expect comets to do it: - They are erratic by nature and won't stay on one orbit long enough - They are too light to synchronize an other.

I can't imagine the light as emitted by the diode regaining coherence every 8.8mm. But if you have a resonator on the path - with many echoes in the resonator, since you see several coherences - it could be an explanation. Something like a mirror that is metallized at its rear face, or a dichroic mirror reflecting from the second face, or maybe a filter, a beamsplitter... Look for something as thick as 4.4mm air, for instance 3mm glass.

By definition, rays produced by ionization are X-rays at most, not gammas.

My proposal is to have an interconnect network less ridiculous as compared with the processing power: - Full throughput of the processor's links, not Ethernet. But serial links, alas. - Full interconnection matrix! Not some hierarchical network, hypercube or other graph with bottlenecks. - I claim that a chip can do that for a few 100 processors. No optics needed. And as a demonstrator, small processors can use a programmable logic chip for the interconnection matrix.

Finally one option I like. Here we can buy cheap pellets made from biomass, which can be chipped shrub, pressed sawdust... They burn gently with small flames. The standard use is in an expensive special boiler that has an automatic feed. Alternately, an automatic feed and burner can be added to existing boilers. My proposal is to build or buy such an automatic feed, and burner if any useful, and fit it in your fireplace. - Renewable and affordable fuel - Decently safe, especially if the pellet silo is outside. - Automatic operation. To be checked, among more things: - Pellets for sale where you live? Or sawdust available? - Chimney dirtier than with good firewood - More ash to be removed - automatically? Here the pellets seller collects it. In case you don't find the automatic feed ready to buy, make your own and sell it to others... Marc Schaefer, aka Enthalpy

Leaving Encelade's orbit, the probe keeps six engines and discards six to weigh 1045kg and join Pallene (212Mm at 13375m/s in 1.15 days). The 752m/s spiral transfer takes 52 days and leaves 983kg. Search for unknown moons on the path - the spiral transfer eases it. The probe has probably optical and radar imagers to inspect the known moons; a lidar may also help discover new ones. Occasionally, a concentrator for an engine, for electricity production or for radio transmissions can pump a properly doped Yag with 200W Sunlight. The R=2.5km Pallene costs nothing to approach and leave, but is too big to land. ---------- 983kg leave Pallene for Methone (194Mm at 13982m/s in 1.01 days). The 607m/s spiral transfer takes 39 days and leaves 936kg. Methone has R~1.6km so I'd rather land elsewhere. But near passes permit to fire bullets (hydrogen cannon?) to help analyze the surface, and maybe evaporate some material with the pulsed Yag. ---------- 936kg leave Methone for Mimas' orbit (185Mm at 14301m/s in 0.942 days). No trojans are reported, so it's time to check the places. The 319m/s spiral transfer takes 20 days and leaves 912kg at the first Trojan place. 60° in 60 days cost 2*12.5m/s or 2kg. Mimas' (R=198km) influence extends only 520km against Saturn, so the probe can't properly orbit the moon; I consider again a slightly tilted elliptic Saturn orbit that circles Mimas. For that, I budget 50m/s or 4kg at arrival, and as much at departure. One more hop to visit the other Trojan place, and the probe weighs 900kg. ---------- The 746m/s spiral transfer to Aegaeon (168Mm at 15047m/s in 0.808 days) needs 44 days and leave 847kg. Detect small moons and rocks meanwhile, since a ring begins there, and land to analyze them. With R~250m, ~20µm/s2 gravity (1/10th the probe's thrust) and 0.1m/s escape speed (spring), approaching Aegeon and landing costs no propellant. 200W concentrated Sunlight can evaporate snow at some distance for analysis, including without landing, but the hotter Yag is better; Sunlight may be good for an oven. Hydrogen at engine exhaust has also enough enthalpy per mole, as kinetic energy, to evaporate many materials at a moon from a limited distance; replacing temporarily hydrogen with, say, rubidium or xenon, enhances sputtering at the target, and special chamber and nozzle can keep the jet more concentrated. Three engines are kept and three discarded to weigh 772kg. The big hydrogen tank should already have been discarded, keeping a small one, but I've forgotten it. ---------- The 1411m/s spiral transfer to the main A ring (140Mm at 16458m/s) needs 138 days and leaves 689kg. As usual, detect small moons and rocks meanwhile, and land to analyze them. Venus and Earth flybys, plus smarter transfers and captures at Saturn's moons, would leave much more probe mass. The Solar thermal engine enables the described dumb scenario, but smart methods developed for chemical and ion engines apply here. At the ring, capture the smaller rocks and land on the bigger to analyze them. Continue to dive as long as hydrogen suffices. ---------- Saturn's main rings can be a separate mission. The probe passes nearer to Saturn to save capture propellant, brakes at Titan and changes the orbit inclination there for flexible launch opportunities; orbiting no moon save further propellant. Ejecting vapour instead of hydrogen, the probe finds propellant in the rings and can navigate everywhere without limit. Marc Schaefer, aka Enthalpy

OK, I understand the original poster has no interest for hardware, despite the orgiinal query was about optical computers. For those people interested in hardware: I strongly believe a full interconnection matrix for many processors with serial links is easily achieved and interesting for servers and supercomputers. A special chip connecting 500 processors with very fast links is a big and expensive project. Though, a university can make a demonstrator for little money and reasonable time. Instead of the biggest processors, just connect microcontrollers. Their links have a modest speed, fewer signals, and some are bidirectional from the beginning. One example is the I2C bus http://en.wikipedia.org/wiki/I%C2%B2C The interconnection matrix can then be programmable logic: cheap. The hundreds of processing elements can be affordable: Arduino or some other. The reasonable project demonstrates all mechanisms for a full-speed chip: full matrix, collision avoidance at the processing elements, intermediate data storage... Marc Schaefer, aka Enthalpy

Willkommen! I feel that "heat treatment technology" is a bit vague. Apparently (sorry if I misinterpret) you include varied materials, varied processes, for varied applications - carried out by very different companies for very different customers. Such a large study would, to my opinion, put artificially together some fields that I'd keep distinct. Not the same people, or not at the same time, are interested by the heat treatment of alloys and by the destruction of toxic substances - so a study including both (or more) would be less well targeted (to my taste). I suppose you'll get more answers here if you define the class of materials you consider. Alloys? Waste? Food? "The treatment of any kind of material is conceivable" not always. Most ceramics, some alloys have zero response to heat treatment. Other materials are destroyed before their properties change usefully. Wearout and breaks... It does happen. Wearout is difficult to predict, so if it arises, the developer more often changes the material of the shape than the heat treatment. Treatment would be an answer, but other methods have a bigger influence - and when the result is difficult to predict accurately, one likes strong means of corrective action. ---------- One example I can think of is not related with wear but just with plain strength. Few titanium alloys are readily available, the only one bringing arguable advantages over steel is Ti-Al6V4, and this one is always delivered annealed (German weichgeglüht). I'd like to get it heat treated (precipitation hardened, German warmausgelagert). Other alloys that were traditionally delivered annealed are slowly becoming available hardened, fortunately: 17-4PH, Cu-Be2... to better fit the user's needs. ---------- For alloys, heat treatment is defined by standards which also define the resulting minimum guaranteed properties. Unless a customer has huge needs, he and the supplier will probably stick to these standards. If defining more accurate treatment conditions, work and trade outside these standards is a (moderate) difficulty.

Thank you! And lithium belongs to the elements interesting to separate into isotopes, so a simple method is useful.

Sure! But through ionization, they produce only X-rays, not gammas.

Dozens die a year in each European country exactly from such heating. CO, CO2, fire. Smoke AND monoxide detectors are the very least. You won't die from lack of oxygen, because excess of CO2 will kill you before. Not a poison... but our lungs accept a limited amount in air until they can't remove it from blood any more - and this happens well before the oxygen is exhausted, and well before a flame shuts. Kerosene doesn't catch fire easily. But its fumes are more toxic. Kerosene itself isn't perfectly sound neither: aromatics. Gas bottles are forbidden within homes in many countries. Sound idea. Even heaters well hold at a wall, supplied by a metal pipe from outside gas, directly connected by a metal vent to a chimney, produce explosions and fires all too often. Some heaters can be put in place of the wood directly in an existing chimney. Possibly the least scary.

Hohmann from Iapetus to Hyperion (orbit R=1501Mm at 5027m/s in 21.3 days) would take years, so the transfer shall push continuously. The probe brakes by 1763m/s and loses the orbit inclination of 8.31° or 601m/s as a mean. For this, the engines push up or down by 41.5° for 2*90° in each turn around Saturn, and flat for 2*90°. The perfectible de-tilting costs 253m/s, bringing the spiral transfer to 2016m/s. Beginning the push in elliptic orbit around Iapetus gives free 8m/s. Hyperion's influence reaches 2.2Mm against Saturn. The target orbit is polar 0.5Mm * 1.1Mm covered in 2.6 days, with periapsis in Hyperion's orbital plane at the daylight side. Capture computed from 50m/s above Hyperion's gravity passes with 34m/s at 0.60Mm from the moon's center, and reaches the final orbit in a single 2.0+0.4 days brake consuming 34+6m/s: 10m/s are for free. HyperionWeakBrake.zip From 3116kg in Iapetus orbit, the 2016-8-10m/s transfer puts 2653kg in Hyperion orbit in 129 days. ---------- Around the following moons, orbits computed without Saturn's influence have a period similar to the moon's orbit around the planet, and their apoapsis exceed the moon's reach against Saturn's gravity gradient, so my capture costs are inaccurate. To obtain an estimate nevertheless, which should be pessimistic, I separate the transfer between the moons' orbits from the capture by the moon, computed with the weak thrust, from zero speed above the moon's gravity, and neglecting Saturn. This would nearly be the case if the major axis were polar; though undesireable for the exploration, it protects the apoapsis from Saturn's tides. It is my hope that specialists find better scenarios. Maybe the probe can enter the Moon's well near the outer Lagrange point being temporarily the apoapsis; the first orbit around the moon would then have its apoapsis perpendicular to Saturn hence protected, giving time to lower it enough. A polar orbit with periapsis in sunlight over the moon's equator should be possible. ---------- From Hyperion to Titan (orbit 1222Mm at 5571m/s in 16 days), the spiral transfer takes 544m/s in 31 days, as Hohmann saves nothing for near orbits. But gravity assists at Titan? While Hyperion offers 10m/s, capture at Titan (3.1Mm * 25Mm) by weak thrust is counted as 298m/s. From 2653kg in Hyperion orbit, the 544-10+298m/s transfer puts 2481kg in Titan orbit in some 100 days. ---------- 2421kg leave Titan after 298m/s expenses. Big jump from Titan to Rhea (orbit 527Mm at 8483m/s in 4.5 days). Six engines are discarded to save 150kg; the 12 remaining push 0.35N and eject 3.4kg/day. The spiral transfer begins with 2271kg, costs 2912m/s and 198 days, ends with 1796kg. Hohmann would save 120m/s but is too long. Slingshots at Titan, maybe Rhea? Capture (1.6Mm * 11.9Mm in 3.3 days) is to cost 107m/s and puts 1780kg in Rhea orbit. ---------- 1764kg leave Rhea for Dione's orbit (377Mm at 10030m/s in 2.7 days), but of course the probe visits Dione's trojans: Helene before and Polydeuces after - same orbit, just separated by 60° from Dione. The 1547m/s spiral transfer takes 86 days and leaves 1557kg. Helene's (R~18km) Lagrange points are 71km away, not bad for observation if the instruments see that near. Or a minimally elliptic and tilted Saturn orbit might make a single-turn helix around Helene's orbit. Negligible fuel expense. Covering the 60° to Dione in 60 days costs 2*25m/s (would be more at Titan's orbit). Capture at Dione (1.1Mm * 7.3Mm in 2.3 days) costs 86m/s and puts 1540kg in Dione orbit. Leaving Dione and joining Polydeuces puts 1523kg there. The R~1.3km object can be sniffed from all directions at negligible cost. Its mass and density are still unknown; to my taste, estimated 200µm/s2 gravity and 0.7m/s escape speed are slightly too much to land there and analyse probes, but unknown smaller moons may float near Helene and Polydeuces. ---------- 1523kg leave Polydeuces for Tethys' orbit (295Mm at 11339m/s in 1.9 days), and the probe visits also Telesto and Calypso, Tethys' trojans. The 1309m/s spiral transfer takes 63 days and leaves 1371kg. The Lagrange points are ~38km from the trojans, no fuel expended there. 60° in 60 days cost 2*20m/s. Capture at Tethys (1.1Mm * 4.8Mm in 1.8 day) costs 77m/s and puts 1357kg in Tethys orbit. The probe weighs 1343kg at Calypso. ---------- 1343kg leave Calypso for Encelade's orbit (238Mm at 12623m/s in 1.4 days). No trojans are reported, so it's time to check the places. The 1284m/s spiral transfer takes 55 days and leaves 1211kg. Observe new moons there - maybe. 60° in 60 days cost 2*16m/s or 3kg to join Encelade. Encelade's Lagrange point is only 950km from its center, and the surface 252km... So the probe should not orbit Encelade, but rather follow a minimally elliptic and tilted Saturn orbit that makes a single-turn helix around Encelade's orbit. I budget 50m/s or 5kg to stabilize around Encelade, putting 1203kg there. I would not separate a lander nor diver there. This is worth an own mission. But a penetrator maybe. And spend all due time there, for sure. The probe weighs 1195kg at the other Trojan place. Find new moons maybe, possibly land on them to analyse probes. A spring lets take off. ---------- More to come. Marc Schaefer, aka Enthalpy