Enthalpy

Senior Members
  • Content Count

    3423
  • Joined

  • Last visited

Everything posted by Enthalpy

  1. I had said that the yellow smoke was lead oxide, it is known now that hundreds of metres around the cathedral are polluted by lead, even if newspapers still put "molten". I had said that the dense smoke had fallen down as it cooled, and you can see it on the pictures. The first 100m are not the worse location. I would have preferred to be wrong. Tons of lead powder spread over a city are a catastrophe. Newspapers begin to discover the cover-up by officials. en.rfi.fr - france24.com - medicalxpress.com and the cited "investigation site" is nearly the only one in France that does more than pasting the AFP news mediapart.fr (in French, and pay) some sites relay the information nouvelobs.com (in French) - tellerreport.com Between "10 times some alert limit", "800 times the legal limit", "child with lead levels in blood much higher than acceptable limits" and "astronomically high lead levels on adjacent roads, blood tests for children under 7 and pregnant women", the discrepancy is vast. I underline that no "lead safe limit" exists.
  2. More tinder for the ancient versus modern discussion, in this case, rebuild exactly as it was last modified, or modify again (...to what I add: complete eventually the cathedral, even if 8 centuries late, by building the spires originally planned on the towers). The French president had wished to rebuild "even more beautiful in five years" and asked for architect proposals to replace the 19th century spire on the transept. The Senate, which has no final say but where the president has no majority, voted to "rebuild according to the last known visual state." The National Assembly voted the final text, removing this constraint, and apparently without a time frame, but organizing the use of the donations. eurasiareview.com france24.com The Vatican had expressed their view that the cathedral has a mission for faith, and the reconstruction should observe that. At least, that's what I read, make your opinion: aleteia.org and admire how French newspapers transform that into "the Vatican wants an identical reconstruction" in title and comment despite their citation doesn't tell it: bfmtv.com (in French) Lead pollution is a damned good reason not to rebuild the roof as it was.
  3. Why firefighters were called very late is better known now nytimes.com Notre-Dame had a very complicated fire detection setup, à la French engineering. When a first alarm rang, the look-out misinterpreted the detector location and sent the watchman to a wrong attic. Only a second alarm, half an hour later, let discover the fire location. Why the firefighters made big efforts to extinguish the towers got a credible explanation too. Huge bells hang there at wood beams that were threatened by the flames. Falling bells could have collapsed the towers and more. ========== I still haven't read what intermediate steps could let a cigarette or short-circuit light thick oak beams. Dust? The investigators reason on technology from a century ago, as usual. No highly flammable substance was found "hence" the fire was not arson. Fabulous. As if power lasers didn't exist for decades. They cut missiles from 200km range, so maybe someone could perhaps imagine they can light wood? As if a monument known worldwide, symbol of a capital, bringing tourist flux, susceptible to damage worth billions, weren't an obvious target justifying high-tech. No, not a century ago after all. Two millennia, hi Archimedes wikipedia
  4. Here's a string chart for a hammered dulcimer, hackbrett or small cimbalom with octave intervals between the sections, as suggested on July 13, 2019 11:50 PM here. 14+14 choirs give it three octaves and a seventh, reaching higher notes than the grand cimbalom. The sections overlap by two semitones, more at the bass bridge because room is available. Additional bass strings are possible. The chosen shape and sizes on the diagram give the plain steel strings a consistent fast propagation, from 1.17 to 1.28* the velocity in air, and the sections join smoothly. The instrument is 1.12m wide between the saddles at G, and 0.34m long at the strings, so including the soundbox and transport box, it fits on one's back. Marc Schaefer, aka Enthalpy
  5. Could manganese steel make musical strings? Also called Hadfield steel or Mangalloy, wikipedia this steel containing 11 to 15% Mn and other condiments hardens quickly by cold-work to UTS>2000MPa while staying tough, good start for a string. Marc Schaefer, aka Enthalpy
  6. At hammered string instruments, the straight strings I proposed arrive with alternating heights at the hitch and tuning pins, as noted here on July 08, 2019. Here's a design idea after all, with metal pinblocks. For minimum deflection at the saddles, I propose to bring the lower ends below the pinblock and the higher ones above. The strings being straight, the ones ending high at right are low at left and reciprocally. So the low ends can hang at hitch pins below the pinblock while the other end have tuning pins above the pinblock. For more thickness at reasonable weight, the pinblock can be cast or milled to leave height at the pins, and if possible at the saddles to improve the stiffness. The sketch isn't exact to the pixel, but 4 or 5 strings per choir look easy. Here the tuning pins limit to 12mm choir period, and a wider pinblock would accommodate them behind the hitch pins for compacity. Conical tuning pins are simpler; the string exits threaded pins at constant height, as is known. The soundboard can extend below the pinblocks to the box, which is said to improve pianos. Pianos have wide holes in the iron frame above the soundboard. Some felt at the hitch pins can reduce noise. Hanging the strings is less easy below the pinblock. I suggest hollow hitch pins to peep through, optionally holes in the pinblock. Springs of durable material and proper strength shall hold the strings while not stretched. They can be offset, so the string passes at the right of the spring. While operating there, one would wisely protect the soundboard, for instance with a mirror. Marc Schaefer, aka Enthalpy
  7. Instruments smaller than the grand cimbalom are more common: hammered dulcimer, small cimbalom, tsambal, Hackbrett in the Alps, and many more names. Their range starts rarely below G below the treble clef and exceeds sometimes the E above. These small cimbaloms could get a subset of my strings chart described on June 16, 2019 here. This example ranges from G to A, exceeding most small cimbaloms, with only 12+11 choirs, including two notes overlap between the string sections, and offers the logical arrangement of the notes. G strings propagating the sound 1.20* faster than the air have 1.042m speaking length, 1.146m between the saddles, approximately 1.26m between the outer pins, so the instrument could be about 1.3m wide. This design option without wound strings will sound better and consistently, as all strings propagate between 1.20 and 1.26*C with smooth transitions between the sections. 24mm spacing between note pairs let the strings occupy 0.3m only, for an instrument <0.4m. Then 1.3m aren't so bulky, similar to a bass guitar, and the instrument with case is easily carried on one's back. ========== Other intervals are possible between the sections, for the big cimbalom too. One example: The lowest section has 11 choirs, say from G to F The three others share 14 choirs starting just above They range from E to F at octave intervals. The instrument spans then 3 octaves and a seventh, more with optional bass strings. It needs 25 choirs, making it 60mm longer, versus 84mm if the usual bass string layout adds that range. The string's propagation speed is almost as consistent. Learning seems easier. Marc Schaefer, aka Enthalpy
  8. Users and luthiers confirm that strings rubbing at bridges hamper the tuning of cimbaloms and similar instruments. A string deflected by 25mm over 250mm pushes 0.1* its tension on the bridge. If µ=0.3 at the narrow metal contact, 3% tension mismatch let the string move, so both sides of the string can have 1.5% detuning, or 1/4 of a semitone. High strings are deflected 5 to 7 times, adding the rubbing force. If one section of a choir is well tuned, the others may be off. Sites recommend to pull the strings from the bridges the equalize the tension, or to strike the strings forte, or to tune first the section farthest from the tuning pin, then nearer and nearer. Dutchak cimbalom puts tuning pins at both ends of the high strings. ========== My strings chart of June 16, 2019 scienceforums has at most 2 bridges per string, easing the problem. Violinists pencil their ebony nut (saddle) so the strings glide more easily. Improves metal pairs too? I suggested here on June 17, 2019 scienceforums to cover the bridges' metal and saddles with catalytic nickel that embeds Ptfe particles. At high pressure, it cuts µ by nearly 10. The piano's pins let the string enter and leave the bridge parallel to inject no force, but their strong zigzag and small bias lets the strings rub much and press little at the bridge. It's the best existing design, but maybe a smaller zigzag and more pin bias is acoustically as good and eases the tuning. Nickel with Ptfe should be tried too. Adopt the design at the cimbalom? My springs at the bridge thrive to push hence rub as little as the vibration needs scienceforums their force is repeatable and adjusted by design, or even by choosing a fastening position. Let's imagine (beware I didn't try!) that a 10mm deflection over 300mm, or rather the equivalent force, suffices to apply the vibrating string on a bridge. µ~0.04 from nickel+Ptfe limit the tension mismatch to 0.1% and the frequency to 0.07%, imperceptible even with two bridges. Marc Schaefer, aka Enthalpy
  9. Between the pedal and the dampers of a cimbalom, dulcimer and similar, I wanted to propose a steel cable in a housing, like for bicycle brakes, but this exists already. One example: saitenart.ch At bicycles they rub, but at cameras they don't. Most cimbaloms and hammered dulcimers have 4 strings for most notes. Maybe they have an excellent reason I didn't grasp. Standard acousticians will answer "inharmonicity" but I'm not convinced.
  10. Hammers heavier than the string is long known, Benade, Rossing and more known names. Treble hammers are reportedly even heavier than my estimate. BUT they strike the treble strings near an end, not at the middle as I imagined, so the string much stiffer there rejects the hammer faster than what I computed. Though, I can't really believe hammer heads striking at 8% of a 48mm long string, so I'll give it a look next time I access a piano. Apparently the ancestors concentrated on the effect on the timbre, not on the (perceived) sound height.
  11. Cimbalom too, but with different styles, varied pieces and composers there: Daniel Skála on Youtube
  12. Manufacturer sites suggest that cimbalists rarely tune their 133 strings. A street musician can't often enough. But humidity and temperature let the wooden frame and soundboard detune the instrument. A metal frame shall stabilize the cimbalum's tuning, once its temperature matches the strings. Some manufacturers have steel or graphite between the wooden pinblocks, but metal everywhere would improve. Someone wanting the typical detuned cimbalom sound can still achieve it. So here it goes, and much applies to other instruments. ========== When humidity and temperature change the soundboard's stiffness, string passing the bridge straight keep their tension at the piano. Avoiding the push enables also a thinner, louder soundboard. The same at a cimbalom: Every second strand of strings arrives higher at the saddles and optionally the tuning pins and hitch (hanger) pins. The dampers, saddles, optionally pinblock must adapt to this. I happily leave this design to someone else. At most pianos' bridge, the strings make a zigzag around two inclined nails agraffes exist too, by phoenix phoenixpianos.co.uk and since the strands are spaced at the bridge of a cimbalom, dulcimer or zither (and this could apply to more instruments), we might pinch the strings delicately so they glide more easily: Spring manufactures bend spring wire to any shape, nonrecurring costs are reasonable. Many designs are possible, including one spring for many notes. The top shape must push on all strings. Here the ends centre the top on the bridge. One end hard to detach would prevent losing the spring. Springs give a force more predictable than screws or others. A layer of catalytic nickel loaded with Ptfe would ease the friction where possible, including at the saddles, but on the springs it's uneasy. When redesigning a cimbalom bridge, one might try strings of equal length within a note. At high notes, when the fundamentals of the short steel strings are in unison, different lengths prevent it for the partials, and possibly we perceive it. Pianos make this effort. ========== To match steel strings, I propose a frame of duplex stainless steel, but normal ferritic or martensitic steel, or cast iron, or austenitic stainless, would be decent to: scienceforums A grand cimbalom has 20 strands of 4 plain steel strings. I take D=0.7mm and D=0.8mm stretched to 1.1*C, or 1111MPa, so they cumulate 40kN roughly, spread over 0.48m. To simplify, I take uniform 0.8m from left to right pins, and the frame must preserve accurately the 4200µm string stretch. I can't estimate decently the bass strings. They will need some more frame, extrapolated from the plain steel strings section. Beams 80mm and 300mm below the strings receive 54.6kN and 14.6kN. 10.9cm2 and 2.9cm2 duplex there cumulate 8.8kg (more for the bass strings) and deform by 200µm, acting together 346µm on the strings. So tuning needs 3 passes after re-stringing, but in normal use, even if all strings had been a quartertone wrong, the first string drifts by 0.24% as all others get tuned, and this is inaudible to most people. The 50MPa stress doesn't determine the metal amount. If spread at two locations, 1/4 and 3/4 of the 0.48m, it leaves 5.5cm2 at the beams compressed by 27.3kN each. Even a 15mm*37mm plain section buckles at 32kN, and much more if supported laterally by skewed beams. Tubes reach more quickly the air temperature, but welded struts, cast parts, or a plate milled or laser-cut, are resistant enough. The pinblock of (duplex stainless) steel scienceforums can be welded on the beams and deforms little between the beams. If it's 80mm wide and keeps 2*140mm2 steel at the rims, EI=83000 SI, and over 4*0.12m with two support, 4*10kN bend it by 9µm only. 20mm thick plates would weigh 2*6.0kg but they can be cast or milled to leave steel just at the top, around the pins and at ribs between the pins, for maybe 2*2kg. Intuitively, limited torsion needs a closed section, obtained by welding plates or a truss. ========== The drawing shows strings at one single height, hiding complexity around the pinblocks. The soundboard holds at the metal frame but must be free enough to expand. Some flexible metal parts could contribute its movements, as on some Camac harps. This first attempt claims a metal frame is feasible, stable and light. Simpler and cheaper must be possible. Marc Schaefer, aka Enthalpy
  13. Small corporate file servers, web servers, database servers and number crunchers are commonly built of few dozen blades holding each a pair of big recent expensive processors, with a rather loose network between the blades, and a Dram throughput that follows less and less the Cpu needs. I propose to use the better networks already described here and assemble more of used outfashioned cheap processors. This shall cumulate more throughput from the Dram, the network, optionally the disks. ========== Here I compare old and recent processors. All from Intel as they made most servers recently, but I have nothing against Amd, Arm and the others. I checked few among the 1001 variants and picked subjectively some for the table. North- and southbridges are soldered on existing mobos, hence not available second-hand nor usable, and new ones should remain expensive when outfashioned. So I checked only Cpu that directly access the Dram and create many Pci-E links to make the network. Chips on Pci-E links shall make ports for Ethernet and the disks like on add-on cards; I like Pci-E disks but they rob too much mobo area here. I didn't check what the Bios and monitoring need. One special card shall connect a screen, keyboard etc. I excluded Pci-E 2 Cpu for throughput and Pci-E 4 Cpu for price in 2019. Pci-E 3 offers 8GT/s per link and direction, so a good x16 network provides ~16GB/s to each Cpu. Line e is a desktop Cpu, less cheap. Lines f g are modern big Cpu that make a server expensive for a medium company. Lines a b c d are candidates for my proposal, in 2019 these have Avx-256 and Ddr3. None integrates a Gpu that draws 15W. | # GHz 64b | # MT/s | W | W/GHz | Cy/T | =========================================================== a | 4 2.8 (4) | 3 1333 | 80 | 7.1 (1.8) | 2.8 (11) | b | 6 2.9 (4) | 4 1600 | 130 | 7.5 (1.9) | 2.7 (11) | c | 8 2.4 (4) | 4 1600 | 95 | 4.9 (1.2) | 3.0 (12) | d | 6 2.4 (4) | 4 1600 | 60 | 4.2 (1.0) | 2.3 (9.0) | <<== =========================================================== e | 6 3.3 (4) | 4 2133 | 140 | 7.1 (1.8) | 2.3 (9.3) | =========================================================== f | 72 1.5 (8) | 6 2400 | 245 | 2.3 (0.3) | 7.5 (60) | g | 24 2.7 (8) | 6 2933 | 205 | 3.2 (0.4) | 3.7 (30) | =================================================================== a = Sandy Bridge-EN Xeon E5-1410. LGA1356, DDR3, 24 lanes, 20€. b = Sandy Bridge-EP Xeon E5-2667. LGA2011, DDR3, 40 lanes, 40€. c = Sandy Bridge-EP Xeon E5-4640. LGA2011, DDR3, 40 lanes, 50€. d = Ivy Bridge-EP Xeon E5-2630L v2. LGA2011, DDR3, 40 lanes, 40€. =================================================================== e = Haswell-E Core i7-5820K. LGA2011-3, DDR4, 28 lanes, 130€. =================================================================== f = Nights Landing Xeon Phi 7290. LGA3647, DDR4, 36 lanes, ++++€ g = Cascade Lake-W Xeon W-3265. LGA3647, DDR4, 64 lanes, 3000€ =================================================================== First # is the number of cores, GHz the base clock, 64b is 4 for Avx-256 and 8 for Avx-512. Next # is the number of Dram channels, MT/s how many 64b words a channel transfers in a µs. W is the maximum design consumption of a socket, called TDP by Intel. W/GHz deduces an energy per scalar or (Simd) computation. It neglects the small gains in cycle efficiency of newer Core architectures. Cy/T compares the Cpu and Dram throughputs in scalar and (Simd) mode, ouch! It's the number of cycles the cores shall wait to obtain one 64b or (Avx) word from the Dram running ideally. The price is for second hand, observed on eBay for reasonable bargains on small amounts in 2019. While an OS runs well from cache memory, most scientific programmes demand Dram throughput or are difficult to write for a cache. For databases, the Dram latency decides and it didn't improve for years. An Dram easy to use would read two data and write one per core and cycle, or Cy/T=0.33, which no present computer achieves. I favoured this ratio when picking the list. More cores make recent Cpu worse, wider Simd even worse. Assembling many old Cpu cumulates more Dram channels. Process shrinks improve the consumption per computation. If an oldfashioned Cpu draws 60W and a recent one saves the half by finishing faster, over 1/3 activity and 5 years the gain is 438kWh <100€, which doesn't buy a fashionable Cpu. If the Dram stalls the recent Cpu more often, the gain vanishes. ========== Each Cpu with its Dram, Ethernet and disk ports shall fit on a daughter board plugged by Pci-E x16 (or two x16 if any possible) on a big mobo that makes the network. But if Pci-E 3 signals can cross a Ddr4 connector, then carry 32+ lanes. The network comprises 16 independent planes where chips make a full crossbar, or if needed a multidimensional crossbar. Can a Cpu on the mobo make a 40*40 crossbar? It takes many Cpu there, and software makes slow communications. At least one crossbar Asic exists for Pci-E connections among Cpu. If that Asic isn't available, make an other. A 32*32 full matrix chip fits in a Dram chip package and can connect 1024 Cpu in a 2D crossbar. 15*16=240 Cpu and 8+8 lanes each take (15*16)*8=248 matrix chips. A chip can serve several smaller planes. The routing density needs a mobo with many layers. Repeaters may be necessary. A big machine with 480 Cpu connects any two Cpu in two hops and transfers 2TB/s through any equator in any direction. Better than a few fibres as a hypertorus. Many small Cpu outperform again few big ones. ========== Liquid cooling takes few mm over the Cpu. Some alkanes are insulators, good coolants, and hard to light Low-freezing rocket fuels New Dram chips soldered directly on the daughter boards, like on graphics cards, would enable 12.7mm spacing. Few 10€ pay 16GB presently. Used Dram modules would be bigger and more flexible, tilted connectors exist for Ddr4 at least au.rs-online.com and horizontal connectors for So-dimm. Or have a minimum Pcb to hold a second Dimm connector making the angle. Daughter boards need local regulators for the Cpu, Dram etc. Like graphics cards do, they can receive 12V from (many) Pc power supplies with minimum recabling. As the cabinet's sides are usefully reserved to Ethernet, the Sata disks and power supplies could reside in the doors. Using an Ivy Bridge-EP Xeon E5-2630L v2 or similar, each daughter board might sell for 250€. A small cabinet with 30 daughter boards would sell for 10k€, cumulate 3.5TFlops on doubles, Dram 1.5TB/s, network 240GB/s through the Equator. A big cabinet with 480 daughter boards would sell for 160k€. 56TFlops, 240TB/s, 1.9TB/s. 400MB/s disks, one per board, would cumulate 190GB/s. Drawings may come. Perhaps. ========== While not a competitor for the clean-sheet architectures I proposed previously, such machines assemble existing hardware. As Pci-E is fully compatible, the number and nature of the daughter boards can evolve, the size of the mobo too, and the boards can serve in successive machines at different customers. As they depend on available second-hand processors, the daughter boards would be diverse within a machine, and the software must cope with small variations in the instruction set. With reasonable capital, a startup company can buy used Cpu on eBay-Alibaba-etc, or rather complete old-fashioned servers with Dram and Ssd, and reorganize more components better around the superior network. Marc Schaefer, aka Enthalpy
  14. And a (not favourable) opinion about a plastic trombone, with sound SITNi-5uVR0 it sounds terrible even with the metal mouthpiece. The musician gives more reasons against the instrument.
  15. More opportunities to hear the baroque trumpet, played by Justin Bland here, wow: rRC1uN7r1mI and oeX8P0cqbUc on Youtube The modern trumpet can't imitate this sound desired by the composer and desirable. Incentive to build the semibaroque trumpet.
  16. At instruments with many strings, the tuning pins and wrestplank (or pinblock) are uneasy to design, due to the big force, the many tuning pins fitting in limited space and budget, the quest for easy and stable tuning. At hitch pins (or hanger pins), the design without tuning is easier. Usually as depicted below, cylindrical accurate steel tuning pins hold by force in tight holes in a wooden wrest plank. Wood provides strong friction and, thanks to its elasticity, it demands a lesser accuracy in the hole tightness. Pins are of hardened steel at most zithers, dulcimers and cimbaloms but a nickel layer would prevent corrosion and bring smooth strong friction as at many pianos; maybe nickel adheres better on medium-carbon steel like 30CrMoV9 tempered at 500°C for hardness. At an example cimbalom, a D=0.8mm string that propagates at 1.3*C pulls almost 800N, and if it leaves the pin just 10mm above a stiff wrestplank, the bending moment is 8N*m. This induces 380MPa in a D=6mm pin, and wood's elasticity worsens the lever length. ========== If wood shall provide the pin 1600N friction from 0.4 coefficient on 25mm height, it needs 8MPa contact pressure at many close locations, too strong for the cross direction, demanding a plywood construction. Banal plywood may fail at hammered dulcimers; Steinway pianos have 7 plies of rock maple at 45° directions. Humidity and also heat let wood expand. Wood creeps also, letting the tuning drift within weeks and months at a new or restringed dulcimer. Here are suggestions for a steel pinblock that shall stabilize the tuning of steel strings. The steel isn't critical, it could be cast iron too, or match a frame steel for easier welding, or be duplex or austenitic stainless steel for best temperature stability of steel strings. Prior experience tells me that electroless nickel doesn't gall against stainless steel, but the pinblock may get electroless nickel too. On figure A1, the hole guides the pin on most height but a ring of metal holds it firmly. The stronger deformation is more easily adjusted. The coefficient of friction must be experimented. Taking 0.3 with nickel, 1600N friction result from elastic 300MPa in a h=2mm dR=1.4mm ring whose D=6mm is expanded by 9µm. Or plastic 550MPa in a h=1.5mm dR=1mm ring expanded by 0.1 to 0.3mm, but the initial force may drift due to creep. Reamers and grinding machines achieve the 9µm difference accurately. Reamers can be customized, here to several diameters. Without ribs, a customized milling tool guided by the hole can make the outer shape. A Cnc milling machine does it too and can leave ribs in the pinblock, and so does casting. The tight parts could have slits or other shapes that ease the deformations. I wouldn't use polymer nor elastomer rings as they creep badly, but separate metal parts could bring more elasticity, like coiled spring wire. They must hold without play at the pinblock. A second ring at the top would exclude dirt and guide even better the tuning pin, figure A2. On figure B1, the toppling moment creates 2*F+F on a shallow gliding fitting, so µ=0.3 would provide only 0.9*F, but the shaft widens. D=6mm to 9mm lets rub 1.35*F at the string, supposed to suffice even if µ=0.23. Maybe. On figure B2, the shaft and the pinblock have a thread whose 30° slope multiplies the rub force by 2 and diameter by 1.23, so 2*F with the deeper fitting and µ=0.3 bring 1.48*F at the string. On figure C, the tuning pin and the pinblock have fitting cones pressed in an other to rub enough. The user turns the pin as he presses, so little force suffices. Violinists do it with two fingers while holding the pegbox with the others, so the tuning hammer (=wrench) will suffice at a cimbalom or piano as it does at a harp. Reamers exist for the euronorm with dD/dh=0.02, but more slope might define the pin's height better. Rubbing 1600N at D=6mm H=25mm means some 11MPa compression or 57ppm deformation, so 12.6-10.7=1.9ppm/K mismatch between X2CrNiMoN22-5-3 and 30CrMoV9 change little over 10K variation. Or make the tuning pins of the same steel, cold-drawn if needed. Marc Schaefer, aka Enthalpy
  17. Steel strings rust, slowly at a piano, faster at mallet instruments that play sometimes outside and are closer to the musician's hands, and more so at plucked instruments. While six strings are easily replaced, zithers and dulcimers can have 30 strands of 3 strings. But how stable can stainless strings be? Experiments shall decide. Stainless steel is abandoned at the piano; I suppose only austenitic steel was tried. Creep and losses may be worse than with carbon steel. All must be hardened by deep cold-work, but I don't have good data for this condition, so the following is unreliable. Martensitic stainless behaves much like carbon steel. Tempered below 300°C, the X20Cr13 stays tough and offers YTS>=1400MPa before cold-work improves it, as is expected but not documented. More alloying elements and less C, like Cr17Ni2Mo, resist corrosion better but offer less hardness and toughness. Variants of X20Cr13 with more carbon exist in some countries, others contain V and similar to form hardening carbides. Among them, X11CrNiMo12 (Böhler T552 and elsewhere) is a turbine alloy with known behaviour at 500-600°C, including creep. Cold-work and under-tempering aren't documented, logically. Tempered at 570°C and without cold work, it offers YTS>900MPa. Expansion 10.3ppm/K and E-modulus drift -174ppm/K (acting -87ppm/K on the frequency) suit a cast iron frame better than high-carbon steel does. Precipitation-hardening martensitic steels harden by ageing after easier cold-working and they resist corrosion far better than high carbon steel does. I have no modulus drift data about the Maraging Ni18Co12Mo5Ti (tough YTS~2360MPa without cold work) but expansion like 9.9ppm/K could fit cast iron. The stainless X3Cr13Ni8Mo2Al aka PH13-8 (Böhler N709 and elsewhere) offers tough YTS~1400MPa which cold work supposedly improves, expansion is 10.3ppm/K like carbon steel. The stainless PH15-7Mo becomes martensite by cold-work prior to ageing (YTS~1800MPa, can improve?), it expands by <9ppm/K to suit a cast iron frame better. Ledeburitic stainless resembles high-carbon steel. Hrc=57 to 60 as tempered nearly suffices, but can it be drawn, can wires be bent? Expansion 10.1ppm/K and E-modulus drift -174ppm/K for X90CrMoV18 suit a cast iron frame better than high-carbon steel does. Austenitic stainless harden by cold work and stay tougher than carbon steel. X12Cr17Ni7 achieves quickly 2000MPa and more, X2Cr17Ni12Mo2 needs deeper area reduction but resists finger corrosion better. Undrawn 15.6ppm/K would fit a frame of austenitic stainless steel or copper alloy. Precipitation-hardening austenitic steels expand even more: 16.2ppm/K for X5Ni26Cr15Ti, whose response to cold drawing isn't documented. Duplex stainless strings would excel against corrosion. YTS and toughness respond to cold reduction similarly to X12Cr17Ni7. 12.5ppm/K for X2CrNiMoN22-5-3 suggests a frame of duplex or austenitic stainless steel. CoCr20Ni16Mo7 resists corrosion better than all steels. It's known to exceed 2050MPa by cold-work plus ageing. 12.3ppm/K would match a duplex or austenitic frame. Nickel alloys for turbines are optimized against creep and known to harden by deformation. Expansion of 12.3ppm/K and E drift of -313ppm/K would let the alloy 718 fit an austenitic frame. Marc Schaefer, aka Enthalpy
  18. What lets the tuning of a metal string drift? Humidity has no expected quick effect on steel. Creep acts very slowly at a piano, where margin below the proof strength might matter, knots quality too. Cold-drawn high-carbon steel expands by 10.4ppm/K. This compares with the string's stretch: 1111MPa/210GPa = 0.53% for 1.1*C. The sqrt drifts the frequency by -980ppm/K. Young's modulus drops by 300ppm/K or less: 210GPa to 205GPa from +20°C to +100°C, accelerates above. At constant length, the sqrt drifts the frequency by -150ppm/K. The change of the string's speaking length, for instance 12ppm/K, is negligible. The frame's expansion and deformation matters much. Stretching the string by 0.74% for 1.3*C reduces to -704ppm/K the thermal expansion effect. But if a string used at 0.8*C isn't overspun, the thermal expansion effect climbs to -1860ppm/K. The relative importance of Young's modulus drift goes the opposite way. Strings of cold-drawn titanium alloy, if practical, would expand less: 9.3ppm/K for Ti-Al6V4 vs 0.57% stretch at 1.1*C, while Young's modulus drops by 450ppm/K. Cold-drawn austenitic stainless steel seems to reduce its Young's modulus as quickly as carbon steel, but the X2CrNiMo17-12-2 expands by 16.2ppm/K (at least when annealed!) and the PH 15-7 Mo by 9ppm/K in condition RH950. Prior to cold-working, duplex X2CrNiMoN22-5-3 reduces its Young's modulus as quickly, but expands by 12.5ppm/K. Gut, polyamide and fluorocarbon polymers behave differently. ========== A perfect steel or cast iron frame expands by 10.4ppm/K too, leaving -150ppm/K due to the drift of Young's modulus. So if you tune at +20°C a cimbalom with hypothetic perfect steel frame and play it outdoors at +10°C, it goes sharp by 0.15%. Inaudible to most people, more so if all strings drift equally. In contrast, a wooden frame does drift, over temperature with some woods, and by humidity always. If the temperature changes by 2K in your room or concert hall, the piano with iron frame drifts by 0.03%, inaudible. ========== The strings' tension deforms the frame, whose Young's modulus drops with temperature too. So should its metal expand faster as a compensation? I don't believe so. The frame must deform far less than the strings to make the tunings independent. Also, the frame's deformation varies among the strings, so thermal expansion couldn't compensate it everywhere. Better a stiff frame whose expansion compensates only the strings. The frame is naturally bulkier than the strings anyway, and it must vibrate less, but its shape too must be stiff. ========== The frame can compensate the strings' Young's modulus drift too. At cold-drawn high-carbon steel stretched for 1.1*C, it acts as 0.15* the thermal expansion, so 12.0ppm/K at the frame would let play from 0°C to +40°C without the 0.3% frequency drift. For instance the stainless duplex X2CrNiMoN22-5-3 offers 12.5ppm/K, is strong and has nice fabrication capabilities. The martensitic X20Cr13 would be less perfect with 10.2ppm/K and the usual austenitic alloys less good with 15.8ppm/K. Aluminium expands more: AA2014 22.7ppm/K, AA5083 23.8ppm/K. 10K variation would detune steel strings by bad 1.1% and 2K by not good 0.2%. For Ti-Al6V4 strings too (harder alloys exist), a frame expansion of 11.9ppm/K would be good. That is, titanium strings could coexist with steel ones, overspun or not. Marc Schaefer, aka Enthalpy
  19. More and more aeroplanes go electric, as the Beeb reports from Le Bourget bbc.com some claiming far better performance from battery-powered craft than my estimates here despite their wing isn't as wide. Hydrogen is missing in the report. Because of that, "electric" aeroplanes are said not to fly far, but the limit results from batteries. Hydrogen tanks and fuel cells give aeroplanes much more range than kerosene does.
  20. Catalytic nickel protects against corrosion and is excellent against galling. I used some, with embedded Ptfe particles, at about 600MPa pressure and nearly no speed, against a martensitic stainless steel that galls horribly. The friction was tiny and very smooth without the stick-slip felt with zinc or phosphate layers against tempered steel. Embedded particles of MoS2 or graphite may be good too. Easy and smooth gliding would improve the bridges and saddles under the strings of some instruments. The piano uses steel nails to deflect the strings at the bridge. Wood receives already a gliding surface, the nails not, despite the force is bigger on them. Catalytic nickel with Ptfe should stabilize the tuning earlier. Some pianos have agraffes on the bridge instead. Same advantage. At least the cimbalum bends the strings over a metal rod at the many bridges. Easier gliding would equalize the tension among the sections between the bridges to improve the intonation. Especially important as the cimbalum has several strings per note. Many instruments have metal saddles or pins near the ends: harp, piano, cimbalom... where the deflection can be big. Better gliding would stabilize the tuning here too, just like violinists put graphite on the wood there. Tuning pegs would better rub smoothly too. They exist of hard wood or metal presently. At the violin, stick-slip of ebony pegs in the maple pegbox is a pain. But to replace ebony, nickel should rub strongly (no Ptfe), be black (graphite glides too easily), leave the fingers clean (embedded Ptfe doesn't). My gut feeling is that a hard polymer like LCP, possibly with a filler, has better chances than a metal. The piano, harp, cimbalom and others have metal tuning pins. Catalytic nickel protects against corrosion, rubs strongly without galling, and hopefully moves smoothly. Steinway pianos have already nickel-plated steel there. At string hooks, especially where piano strings make a U-turn without a knot, the strong friction of nickel might help tuning. Marc Schaefer, aka Enthalpy
  21. Here I propose a simpler notes chart for the cimbalom. The bass strands keep the usual positions up to B=234Hz, the rest differs: Semitones progress smoothly but for three jumps. At the jumps, the sections overlap by three semitones, similarly to trill keys at woodwinds. The sections have a constant interval. At the violin it helps. I imagine this chart makes the cimbalum easier to learn and play, but again I don't play it. Example of a usual (but incomplete) chart: beyondkarpaty.mutiny.net Hammered dulcimers would resemble more, but the intervals and bass strings differ. A pair of straight dampers can reach C=1051Hz while most Schunda-like models stop three semitones earlier. They are far from the struck positions. Schunda had achieved a nearly rectangular instrument shape at the cost of complicated notes chart and very inconsistent string sound speed even among consecutive notes. In my chart, the trapezoidal shape keeps the string sound speed between 1.10*C and 1.16*C at the treble and medium, varying very smoothly even at the section jumps, and decreasing gently to 0.54*C at the bass. Other string lengths would adjust these example figures, say between 1.20* and 1.27*. All bridges and saddles are straight on the drawing, but curves as at the piano could further equalize the string sound speed or limit the instrument's width. The outer bridges leave 10% non-speaking length in the corresponding strings, less than at a piano, but this can increase if accepting a slower propagation, at all these notes or only the lowest ones. Schunda's design has 39 strands, my chart has 43 with fewer spun strings. I suppose three strings per strand suffice. Marc Schaefer, aka Enthalpy
  22. And oops. All pictures show one hook and one knot per string. Bad reason.
  23. Here's a sketch of a big cimbalum that widens enough at the near strings to keep a good tension in the medium register, as suggested here yesterday. Its slope resembles much a small cimbalom. I fully ignore whether a wider instrument is difficult to play. If it helps to play near the centerline, the strings' tilt can be kept using taller bridges. For nicer and more uniform timbre, sound is here consistently 1.22* to 1.35* as fast in the medium strings as in the air. It remains faster than in air in half of the bass strings, whose plain steel saves money, then decreases in the spun strings to 0.58* at the lowest note. I'd keep the mass of the strings, that is, thinner if longer. Less stress is also welcome at the bigger frame. I've kept the distance from the bass bridges to the outer rails. Bringing less stiffness, strings passing the bridges straight might let shorten this distance. It is very short at a piano. Thanks to its metal frame, the piano also extends its soundboard very far under the agraffes and tuning pegs: to be copied if possible. Marc Schaefer, aka Enthalpy
  24. Here are some thoughts about the cimbalom. en.wiki - fr.wiki - cimbalombohak.sk - cimbalom.hu Big warning: I don't play the cimbalom nor any related instrument, so much here is probably b**ocks. If at the end one detail or an other makes sense, fine. Many instruments are called cimbalom, the name varies also much, and other instruments can be very similar. I consider the grand cimbalom, of Hungarian style, developed by Schunda around 1870. ========== If I see properly, the two sets of dampers are pushed down directly by long beams, possibly less stiff than needed. Hoping to make settings easier and more stable, I suggest: Individual movements for the dampers hold at a fixed beam; Individual springs to push each damper against a string course; One common action on each side, moved by the pedal, to pull all dampers from the strings; Optionally, the contact between the action and the dampers can be adjusted individually. The dampers for the central portions of treble strings still need some transmission. I wish the strings would sound for longer with the dampers. Enabling fine adjustments must help. ========== The medium alternates long string courses with others split in two by a bridge (bridges have voids for the uninterrupted strings). Angles by the bridges put courses higher at one end or the other to help the musician hit the desired note. If no bridge shortened these longer courses, an identical ratio between the full length, the longer part and the shorter part would be the golden number, (1+sqrt(5))/2 ~ 1.618. At identical sound speed in the strings, the intervals would be 8 or better 9 semitones (minor or major sixth) between full length, long part and short part, so 2*9 courses would span 27 semitones at uniform sound speed. The bridge loses about 2 semitones. That's still 4 semitones more than presently, with notes arranged more logically and with uniform sound speed. Whether this is advantageous, and enough so to learn a new string chart? ========== Why 4 strings per note? For a strong attack but longer sustain at medium and treble notes, 2 strings offer eigenmodes with no net force on the bridge and soundboard, and 3 strings suppress the roll moments too. 4 bring no further advantage here, and the piano has only 3. String inharmonicity improves with finer strings, and then more strings keep some moving mass. Though, I believe inharmonicity has been hugely overstated; it's not even a drawback with reasonable diameters like here. Maybe 4 strings cost less than 3. They replace a time-consuming knot on the instrument's left by a turn, plus one cheap tuning peg and its hole at the right. Pianos share some wires before and beyond the turn among adjacent notes, but their tension is very nearly the same, as opposed to the cimbalom. I suppose there is some design flexibility here. ========== Existing instruments widen very slowly at the low notes. Did Schunda consider his design already bulky and heavy enough? Consequently, the nearer strings are far too short, see the drawing: the farther medium strings are healthy 1.23*C or even 1.35*C long (as compared to the sound speed in air) but the nearer medium strings drop to 1.03*C or 0.83*C, and the nearer bass strings to meager 0.37*C, usually a receipe for bad sound. Small cimbaloms widen much more strongly at the nearer strings. Would it hamper playing the big instruments? At least, longer near strings would keep a decent sound speed. The farther medium strings could keep their length and the nearer be 1.3* as long. With the present notes chart, the nearer medium strings would be 1.08*C and 1.32*C long, perfect for plain steel strings and for the transition to spun bass strings. If keeping straight bridges, the lowest bass string would be 1.63* longer, or more decent 0.60*C. Drawing later and maybe. The instrument then widens from 1.5m to 2.1m. Many cimbaloms still have a wooden frame. I hope a metal frame would stabilize the tuning and let the instrument weigh less than Schunda's 1870 design, 100kg. Later and maybe. ========== The angles in the strings hamper the movements of the bridges and put also much pressure on the soundboard, which must be sturdy and is even supported by pillars under the bridges. Efficient reasons for the lack of sonority. Zig-zags at the bridge, like at the piano, would solve all. They are not possible at the highest strings. Elsewhere, they need an instrument higher at the sides, which a metal frame should enable. Later and maybe. Marc Schaefer, aka Enthalpy
  25. From fresh pixel counting on grand pianos pictures, their treble and medium strings propagate the sound 1.2* as fast as air. The factor drops smoothly at the bass, which are spun with copper as soon as the factor is 1. Hungarian style grand cimbaloms have some strings long as 1.2* air half-waves, for few notes even 1.35*, but this drops to 0.83* at some plain steel strings, and to meagre 0.37* at the lowest spun strings. The puzzling arrangement of the strings gives very different string lengths to neighbour notes, jumping from 0.83 to 1.35 and back within semitones. So if the factor 0.83 contributes to the sound of some cimbalom notes, an imitating grand piano could be lowered by 6 semitones from 1.2. Or by 7, a fifth, for easier transposition. The cimbalom has also 3 or 4 strings per note except the lowest ones. At a piano, their tunes must match exactly to sound good. At a cimbalom, which has usually a wooden frame and produces from most strands several notes separated by bridges, the perfect match must be rare and brief. Unmatched tunes in strands may contribute to the cimbalom sound with its typical lisp. The prepared grand piano can imitate this easily. And of course, use special hard hammer heads or mallets. Marc Schaefer, aka Enthalpy