--- Log opened Sun Nov 07 00:00:25 2021
00:41 -!- spaceangel [~spaceange@ip-89-176-181-220.net.upcbroadband.cz] has joined #hplusroadmap
01:37 -!- spaceangel [~spaceange@ip-89-176-181-220.net.upcbroadband.cz] has quit [Remote host closed the connection]
06:42 < docl> I'm assuming everyone has seen this https://dberard.com/home-built-stm/
06:42 < docl> .title
06:42 < saxo> Home-Built STM | Dan Berard
06:50 < docl> thinking about the combinatorial thing -- things depend on multiple other things and this causes an explosion of combos, but this collapses at some point because things can be multipurpose, giving you more of a diamond shape instead of a pyramid
06:52 < docl> often there's efficiency tradeoffs in adapting a multipurpose thing instead of specific purpose, but we aren't as concerned about efficiency as much given the automated expansion you get with self rep
06:54 < docl> anyway it seems like there are sweet spot things that are both multipurpose and simple. I think piezoceramics might be one of those. you can move things without wrapping coils, just mixing and baking the right kind of clay. and it permits ultrahigh precision without stepper gears and stuff
07:14 < docl> if you are content to go slow, thermal expansion might be exploitable... bimetallic strips are old school. problem is if you change temperature on a thing it tends to affect everything around it. I could see this being used to wind clockwork as a crude solar collector but in a precision context thermal effects are usually noise :/
08:07 < lsneff> yeah you essentially need piezoelectrics to get submicron precision actuation on anything, but good look getting piezoelectric actuators to move anything a substantial distance, even with leverage
08:08 < lsneff> One can potentially use motor coils printed into a pcb
08:08 < docl> actually they seem to work fine, at least there seem to be ways to do it https://www.youtube.com/watch?v=uFZsH62ewYo
08:08 < docl> .title
08:08 < saxo> Piezoelectric motor - YouTube
08:09 < lsneff> Ah an inchworm drive
08:10 < lsneff> Are there any examples of an enclosed motor that uses that technique?
08:11 < docl> here's another one that works a little differently https://www.youtube.com/watch?v=-5jf_wwB_MI
08:11 < docl> .title
08:11 < saxo> PiezoWave - Miniature piezo motor - YouTube
08:11 < docl> plus there are those piezo actuated membrane water pumps
08:15 < docl> .title https://www.youtube.com/watch?v=u-h8dhvmB1E
08:15 < saxo> Piezoelectric Micro Pump -Takasago Fluidic Systems- - YouTube
08:16 < docl> you could use them with hydraulics
08:23 < docl> hmm, wonder if you can have hydraulic oil that expands under an electric charge. piezohydraulic fluid.
08:25 < docl> piezoceramic + hydraulic fluid is the more obvious option, but if you can do it directly with the fluid that might be easier
08:27 < docl> (in terms of narrowing the combinatorial diamond)
08:29 < docl> narrowing the diamond happens 2 ways: making something more flexible-use (so it the dependency tree collapses sooner) or making stuff that depends on fewer things (so it doesn't expand as much) 
08:31 -!- srk_ [~sorki@user/srk] has joined #hplusroadmap
08:32 < docl> one approach might be to suspend pzt nanoparticles in hydraulic oil. increase charge on the reservior and those particles grow to occupy more volume. I'm guessing not enough of an effect without an absurdly big reservior though.
08:32 -!- srk [~sorki@user/srk] has quit [Ping timeout: 260 seconds]
08:34 -!- srk_ is now known as srk
08:37 < docl> seems pretty obvious you could use nanomechanical or micromechanical methods to do something like pump a gas into a chamber to alter size based on applied charge or magnetic field. not useful unless you have a relatively low-tech method of producing the structures without too many dependencies though
08:58 -!- spaceangel [~spaceange@ip-89-176-181-220.net.upcbroadband.cz] has joined #hplusroadmap
09:10 < lsneff> yeah, unfortunately that doesn’t really exist
09:10 < lsneff> an artificial muscle like that would transform several fields
09:12 < docl> hmm. pzt assumes chemical refining sufficient to obtain all three elements. if you're trying to build starting with clay I wonder if that's practical. maybe something like aluminum nitride crystals would work better? seems to be insoluble in water which could be relevant if your hydraulic fluid is water
09:15 < docl> rochelle salt is the usual diy piezo experiment, but that involves organic chem and is soluble in water.
09:16 < docl> .wk https://en.wikipedia.org/wiki/List_of_piezoelectric_materials
09:16 < docl> .wik
09:16 < saxo> "Community portal – Bulletin board, projects, resources and activities covering a wide range of Wikipedia areas. Help desk – Ask questions about using Wikipedia." - https://en.wikipedia.org/wiki/Main_Page
09:16 < docl> .wik List_of_piezoelectric_materials
09:16 < saxo> "This page lists properties of several commonly used piezoelectric materials. /  Piezoelectric materials (PMs) can be broadly classified as either crystalline, ceramic, or polymeric.  The most commonly produced piezoelectric ceramics are lead zirconate titanate (PZT), barium [...]" - https://en.wikipedia.org/wiki/List_of_piezoelectric_materials
09:18 < docl> muurkha: I think the idea of self rep with clay as the base material is interesting. any thoughts on actuators for that?
09:31 < docl> I guess a clay system needs a somewhat high heat source to glaze watertight vessels and permit chemistry, so either you use fire (dependency on wood or coal) or you need glass/shiny stuff for optics for a solar oven. can't quite picture a clay based wind driven friction system doing the trick without significant complexity
09:41 < docl> well maybe there are non-optical ways to make solar ovens hot enough for firing clay
10:02 -!- Hooloovoo [~Hooloovoo@hax0rbana.org] has quit [Ping timeout: 260 seconds]
10:05 -!- Hooloovoo [~Hooloovoo@hax0rbana.org] has joined #hplusroadmap
10:16 -!- Codaraxis_ [~Codaraxis@user/codaraxis] has joined #hplusroadmap
10:19 -!- Codaraxis [~Codaraxis@user/codaraxis] has quit [Ping timeout: 256 seconds]
10:57 < muurkha> fenn: thanks!  I was thinking (a) if the FR4 (or, better, phenolic) is glued firmly to a carbon-fiber backing, the TCE of the whole system will be basically that of the carbon fiber because it's so much stiffer;
10:59 < muurkha> (b) the capacitive rectangles are like a millimeter or bigger in size, so there's an analog problem in measuring the typical 10-μm precision.  maybe not enough to matter due to differential sensing and stuff
10:59 < fenn> wouldn't that just cause it to bend, like a bimetallic strip in a thermostat
11:00 < fenn> maybe if the circuit board were really thin, such that the shear strength of the glue were enough to stretch the circuit board completely
11:01 < fenn> but it would be easier to just measure the temperature of the instrument and do a little math
11:03 < fenn> calipers are not really intended to measure much less than 10 microns
11:03 < muurkha> docl: can you get piezoceramics by baking clay?  I mean I know quartz is piezoelectric, but very slightly compared to things like barium titanate and PZT, and I'm not sure you can pole a polycrystalline quartz ceramic the way you can with PZT?
11:12 < docl> not sure if undifferentiated clay has piezoelectric properties, but my guess is no (or not siginificantly enough). I was just thinking that since pzt is a ceramic it works with ceramic-handling tools.
11:19 < docl> looks like AlN might not be so great since the piezoelectric coefficient is similar to quartz, 2 orders of magnitude lower than pzt
11:21 < docl> hmm, maybe most ceramics are in the 1-10 pC/N range? might actually be possible to use clay if you have a way to magnify it enough
11:22 < muurkha> fenn: yes, the differential TCE could cause it to bend causing cosine error, but the total force you can get from the thermal expansion of phenolic is pretty small
11:23 < muurkha> Mitutoyo does sell carbon-fiber calipers (real ones) that are supposedly good down to submicron
11:26 < muurkha> muscle wires are a common way to get mechanical (linear!) actuation via temperature change, and I agree that there's a lot of attendant difficulty with the thermal drift messing with other random things.  generally in metrology uncontrolled temperature is The Enemy
11:27 < muurkha> lsneff: PI sells some walking piezoelectric actuators that can do submicron precision actuation over tens of millimeters.  voice coils can also do that, as they routinely do in hard disks
11:28 < docl> er pN/C I guess is the unit, 1 N/C being equivalent to 1 m/V if I got that right
11:28 < muurkha> also there are traveling wave piezo motors used for camera focusing and stuff
11:29 < fenn> V/m is dimensionally equivalent but makes no sense
11:30 < docl> if you're using clay, you might get variability in the composition so it might be hard to calculate in advance, but once you make it I'm thinking it wouldn't change so you can calibrate it
11:30 < muurkha> generally the volume change of piezoelectrics under applied fields is tiny or zero; you're just deforming the unit cells, not expanding them.  that's why you have to pole PZT
11:33 < muurkha> I think organic chem for selfrep is important to avoid for three reasons: the political unpalatability of "flesh-eating robots"; the limited availability of organic feedstocks in many environments of interest, including in the terrestrial biosphere, which has orders of magnitude more aluminum, iron, and silicon available; and the potential perverse incentives that would be set up for selfrep owners, 
11:33 < muurkha> like the biodiesel problems times a billion
11:33 < docl> hmm. if you could make sphereoids of polled matter the volume might change based on their deformation
11:34 < muurkha> clay is pretty bad at actuators and gears and flexures and stuff, at least at everyday temperatures
11:37 < muurkha> fenn: if you apply 1000 volts over a millimeter, your electric field is 10 times as strong as if you apply 1000 volts over a centimeter.  if the piezoelectric effect is linear over that range of field strengths, you'd expect 10 times as much strain (say, 10⁻⁶ instead of 10⁻⁷) but over a ten times shorter distance that works out to the same displacement
11:37 < muurkha> one nanometer in both cases
11:37 < muurkha> so V/m might actually make sense
11:37 < fenn> but N/C is force per charge
11:37 < fenn> says nothing about displacement
11:39 < muurkha> yeah, I'm having a hard time figuring out how N/C applies, though as you say it's dimensionally equivalent
11:39 < docl> here's my source on them being equivalent units based on a quick google search https://www.unitconverters.net/electric-field-strength/newton-coulomb-to-millivolt-meter.htm
11:40 < muurkha> I just used units(1)
11:40 < fenn> i also used units, but you can multiply out all the kilograms meters and seconds if you need to verify
11:40 < fenn> and amps
11:41 < muurkha> right, definitionally a volt is a joule per coulomb, and a joule is a newton meter
11:41 < muurkha> defining a coulomb as an amp second is going the long way around
11:44 < muurkha> Lukas's Wiki is primarily concerned with atomically precise manufacturing, but we don't need that for selfrep
11:49 < muurkha> let's see.  suppose you put two such capacitors electrically in series so they have the same charge.  the 1mm-thick one will have ten times the capacitance (assuming the same plate area) and thus one tenth the voltage of the 10mm one
11:51 < fenn> i think if you squeeze it until it compresses by 1 meter, you get some number of volts out
11:51 < muurkha> so you'll have the same field strength and thus the same strain in both dielectrics.  multiply by the area and the young's modulus and you get a force
11:51 < fenn> right, you'd have to know the modulus
11:52 < muurkha> that's the weird part, yeah
11:52 < fenn> but if these units are dimensionally equivalent, you wouldn't have to know the modulus
11:53 < fenn> anyway it's past my bedtime
11:53 < muurkha> the area also figures into the voltage/charge relation
11:53 < muurkha> goodnight!
11:55 < muurkha> I wonder if there's a hidden unit in there somewhere like the radians in torque
11:56 < docl> some of Lukas other work might be relevant to clanking replicators https://mechadense.github.io/00.Home-en.html
11:59 < muurkha> thanks!
12:01 < muurkha> oh, I forgot to mention: in addition to piezo and voicecoil actuators, flexures can be used to get submicron positioning out of coarser actuators
12:03 < muurkha> with those plus the more finicky thermal actuators, have we covered all the possibilities?
12:09 < docl> hmm... optical? or ion? those might be more relevant in a vacuum chamber kind of setup
12:10 < docl> hadn't thought about flexures, but that makes sense
12:12 < fenn> my first thought for clay actuators was pneumatics
12:13 < fenn> your options increase a lot with abrasive machining processes
12:41 < docl> I wonder if rapidly spinning bits is possible without a dependency on lubricants and/or hard metals and/or magnetic coils. maybe something that uses airjets like an air hockey table and hangs downward drillpress style?
13:08 -!- Codaraxis__ [~Codaraxis@user/codaraxis] has joined #hplusroadmap
13:12 -!- Codaraxis_ [~Codaraxis@user/codaraxis] has quit [Ping timeout: 244 seconds]
14:45 -!- Codaraxis__ [~Codaraxis@user/codaraxis] has quit [Quit: Leaving]
15:08 -!- spaceangel [~spaceange@ip-89-176-181-220.net.upcbroadband.cz] has quit [Remote host closed the connection]
15:12 < muurkha> docl: are you thinking about focused ion beam machining?  optical patterning (e.g., photolithography) can get down into the submicron range, but maybe not very far.  you could try doing near-field light stuff but that just gets you back to the need to position the light sources precisely
15:12 < muurkha> fenn: what do you mean?
15:14 < muurkha> air bearings are definitely a way to do rapidly spinning bits; I hadn't thought about trying to combine them with fired-clay parts.  they have very high stiffness and, while they're operating, no wear; this is how dentists' drills work
15:14 < muurkha> no lubricants needed.  spinning things rapidly probably requires that they either be extremely small or made of strong materials
15:20 < docl> muurkha: that's the main case, but I also had the thought of maybe using ions as reaction mass to steer tiny drones around (in a space based factory, say). unlikely to be practical for earth based projects though
15:31 < docl> would it be possible to make a drill of 100% clay that lasts a while, I wonder?
17:03 < L29Ah> depends on whether you can restore a portion of clay to aluminum metal and then shape it into a drill bit
17:24 < kanzure> nmz787 has a focused ion beam
17:56 < lsneff> its too bad that reprap ion beam spin-off never happened
17:56 < lsneff> actually, that’s something that could play a huge role in a self-replicating factory
17:58 < lsneff> printers with hundreds or thousands of electron beam welders all aimed independently could probably print big objects pretty fast with several microns of tolerance
17:59 < muurkha> a drill might be the most unsuitable possible thing to make out of clay :)
18:00 < muurkha> if you heat up corundum with electron beams in a vacuum you should be able to reduce it
18:00 < lsneff> in fact, if I wanted to make von neumanns happen asap, I’d start a company and convince VCs to find me to design open-source massively parallel ion beam welders
18:01 < muurkha> if you try that with clay you'll end up with aluminum with a lot of silicon in it
18:02 < kanzure> what would it take to actually solve this problem
18:02 < kanzure> of mechanical self-replication
18:03 < kanzure> if a billionaire was going to open his pocket book and do whatever you say in the interest of self-replication, what would it be precisely?
18:09 < lsneff> you happen to know one interested in solving it?
18:19 < muurkha> lsneff: what's preventing you from starting such a company?  more important things to do with your time?
18:20 < lsneff> because im not confident that I could actually make it happen
18:21 < muurkha> well, that's a reasonable reason to derate it a little, but not to abstain entirely
18:32 < lsneff> im also confident that I have no idea what I’m talking about
18:33 < L29Ah> it will probably be made one way or another in a few decades; the equipment prerequisites are a bit out of the reach of an average tinkerer now tho, so isn't going to happen as smoothly as FDM
18:36 < L29Ah> and the original design suggests terrible speed and large power consumption
18:39 < lsneff> that’s how it always is I guess. open-source hardware is consistently 50+ years behind commercial hardware
18:40 < L29Ah> are there commercial implementations?
18:41 -!- strages [uid11297@id-11297.helmsley.irccloud.com] has quit [Ping timeout: 264 seconds]
18:41 < lsneff> the metalicarap says ion beam welding has been used for 50 years in industry
18:41 < lsneff> perhaps not in 3D printer form though
18:43 -!- strages [uid11297@helmsley.irccloud.com] has joined #hplusroadmap
18:51 -!- redlegion [sid429547@ilkley.irccloud.com] has quit [Read error: Connection reset by peer]
19:04 -!- redlegion [sid429547@id-429547.ilkley.irccloud.com] has joined #hplusroadmap
19:18 -!- redlegion [sid429547@id-429547.ilkley.irccloud.com] has quit [Ping timeout: 268 seconds]
19:21 -!- redlegion [sid429547@ilkley.irccloud.com] has joined #hplusroadmap
19:43 < docl> ok so I'm on an elevator with a billionaire who has stars in their eyes about self rep, what does my pitch look like? something like: self replication runs into the problem of multiple dependencies stacking on each other exponentially, and we need to collapse the combinatorial explosion by simplifying on the one end while increasing the multipurpose utility of products on the other end causing it to 
19:43 < docl> merge down sooner. it has to be shaped like a "diamond" instead of a "pyramid" in order to self replicate. industry technically does this, but the diamond too fat to fit in a single building or even a single firm, and the knowledge is disorganized because it grows and collapses organically. my plan involves starting 3 projects in parallel which all try to achieve the small diamond (workbench scale 
19:43 < docl> ideally, building scale realistically) using different strategies, and trade each other for small amounts of products. over time we add more projects/modules, but coordinate them as they grow such they can fill each other's needs. so there's a clay 3d printing and machining project, a chemlab to mix solutions, electroplate things, etc., and a vacuum oriented project to do things with lasers, ion beams, 
19:44 < docl> and cvd. as we go we adapt the products to be useful in the construction of new units in the three categories. we also make tentative plans to add nano and genetic tech, metallurgy, and so on. once there's a working system, we try to shrink the diamonds further. we're working on one of the world's most important projects and expect to have no trouble attracting top talent, but it requires 
19:44 < docl> multidisciplinary thinking so we'll be looking for that in people we hire.
19:45 < docl> gotta condense that down further
19:54 -!- keyke [~qi@user/keyke] has joined #hplusroadmap
19:54 < keyke> yo sup 
19:54 < docl> hello
19:55 < keyke> which's your favorite nootropic
19:55 < keyke> ?
19:56 < keyke> thus
19:56 < keyke> is there a site or blog with lastest biohacking news?
19:57 < docl> idk still haven't tried most of them. modafinil was pretty cool. I use nicotine mainly because it's accessible
19:59 < keyke> weed also
19:59 < keyke> lol
20:01 < docl> when I lived in oregon I tried some, not sure I'd call it a noot for me but interesting
20:01 < keyke> whats a noot?
20:01 < docl> a lot of it comes down to individual biochemistry
20:01 < lsneff> definitely wouldn’t count it as a nootropic
20:01 < keyke> ye well
20:01 < lsneff> docl: that seemed like it was on the right path
20:02 < keyke> it is an stimulant
20:02 -!- s0ph1a [sid246387@helmsley.irccloud.com] has quit [Ping timeout: 260 seconds]
20:02 < lsneff> I think the combinatorial diamond terminology is good
20:02 < docl> lsneff: thanks!
20:03 -!- acertain [sid470584@hampstead.irccloud.com] has quit [Ping timeout: 245 seconds]
20:03 -!- cpopell [sid506802@tinside.irccloud.com] has quit [Ping timeout: 264 seconds]
20:03 -!- cpopell [sid506802@2a03:5180:f::7:bbb2] has joined #hplusroadmap
20:03 -!- yuanti [sid16585@tinside.irccloud.com] has quit [Ping timeout: 245 seconds]
20:03 -!- redlegion [sid429547@ilkley.irccloud.com] has quit [Ping timeout: 256 seconds]
20:04 -!- s0ph1a [sid246387@id-246387.helmsley.irccloud.com] has joined #hplusroadmap
20:04 -!- yuanti [sid16585@id-16585.tinside.irccloud.com] has joined #hplusroadmap
20:04 -!- acertain [sid470584@id-470584.hampstead.irccloud.com] has joined #hplusroadmap
20:04 -!- redlegion [sid429547@id-429547.ilkley.irccloud.com] has joined #hplusroadmap
20:04 -!- keyke [~qi@user/keyke] has quit [Quit: keyke]
20:05 < docl> I wonder if we can get all the glassware needs of a chemlab in the system without adding a glassworks? maybe just cvd coat clay vessels with silicon and let it oxidize into quartz
20:05 < lsneff> could do the same with 3D printed plastics
20:06 < docl> I just can't see getting too far without doing some solution chemistry (unless the vacuum chamber stuff turns out really well)
20:06 < docl> and it gives a place to add on biolab stuff
20:06 < docl> or things like spiroligomers
20:08 < docl> but the diamond shape thing gives an overarching vision which I think is important for a business case, especially given how multipronged the approach is likely to be
20:15 < docl> I know muurkha has reservations about including organics... it may simplify matters a lot though, if there's something we can't do in clay/metal, and CO2 is abundant
20:16 < lsneff> why clay again?
20:16 < docl> hmm, google does give results for cvd coated plastic
20:16 < docl> well, refractory properties is one reason
20:16 < docl> and it's abundant in nature, basically fine enough grained sand is clay
20:17 < docl> and it can be shaped without high heat (you use heat to fix its form, but while shaping it's an aqueous slurry)
20:17 < lsneff> any data on abundance of clay off world?
20:18 < docl> if you recycle the water, that's basically what lunar regolith is (I think)
20:20 < docl> the composition of lunar soil is basically like dirt on earth. the particle shapes are different (no water erosion) though so it may have different properties in a solution to some extent
20:21 < docl> bed time for me
20:23 < lsneff> it might be difficult to rely on the particular structural/chemical/etc properties of clay when each time it’s harvested, it’ll have different proportions of minerals
20:23 < lsneff> goodnight
20:44 < maaku> docl: regolith has immensely different properties, due to both size and shape of particles.
20:45 < maaku> it's actually a really hard problem to get good lunar simulant
20:45 < maaku> lsneff: we know various wealthy people who could be convinced to fund such an endeavour
20:45 < maaku> but do you think you could actually achieve kinematic self-replication?
22:05 -!- Netsplit *.net <-> *.split quits: streety
22:05 -!- Netsplit over, joins: streety
22:12 -!- Netsplit *.net <-> *.split quits: greenz1[m], otoburb
22:13 -!- Netsplit over, joins: otoburb
22:17 -!- greenz1[m] [~greenz1@2001:470:69fc:105::ca1a] has joined #hplusroadmap
23:01 -!- Croran [~Croran@71.231.214.173] has joined #hplusroadmap
--- Log closed Mon Nov 08 00:00:26 2021