RepRap or conventional manufacturing cheaper?

This thread is a spinoff of the one at [forums.reprap.org].

Posters in that thread have been assuming that either (1) RepRap will be useful because it allows people to print out exactly what they need when they need it and how they want it, or (2) not only is the RepRap project good for that, but it will also be able to print general and limited-customization commodities (tableware, shoes, etc.) more cheaply than a commercial manufacturing service can produce them and bring them to market.

I think it's very interesting how people assume either (1) or (2), but as far as I know no real studies have been done to determine which is more accurate. Personally, I'd be more inclined to believe (1) but if anyone has hard evidence either way it would be good to know.

Kyle

Edited to bypass automatic emoticons.

Edited 1 time(s). Last edit at 08/27/2007 07:14PM by Kyle Corbitt.

Personally. I believe both.

I believe that RepRap will allow some things, for which there is not a sufficient market in a small area, to be produced on demand. This will allow two steps in the supply chain to be eliminated, transportation and storage of finished goods. Consider convincing your shopkeeper to keep every possible variant of salt shaker in stock, at all times, in a community of less than 50,000 persons. This is much the effect that the internet has had on communities. Just because no one within ten miles of where you live has any interest in collecting such-n-such, or doing such-n-such, doesn't mean you can't join an on-line community that centers around just that type of collection or activity. Certainly the internet has made it possible for people to find music that better suits them than top-40/Country/Classical/Praise&Worship, (the local mix of radio stations here.)

As for the second part. Reprap will be a tool for manufacture. It will be able to produce parts with a high degree of accuracy for which the operator lacks the skills to produce himself. He might, however, possess sufficient skills to 'tweak' the design. Back to shoes, as they're clothing, (thus, would benefit from being fitted to the very non-uniform human frame,) and rigid, (thus being able to benefit from something that produces items in a rigid medium.)
You go in for a new pair of boots. The cobbler measures your feet, and prints up the sole to order, the design altered to the proper dimensions for your foot... not all size-12's. Perhaps much of the other pattern working and cutting is also done by related CNC-type machines, (cutting, and perforating, leather to shape, for instance.) Perhaps many of the parts are still manufactured because they're used in many different designs, without the need or desirably of alteration, (metal snaps, and eyes for laces, for instance.) In this way, for a little more per unit, a professional has the tools to make key components that previously would require competence as a whittler, machinist, or woodworker, while also freeing them from the time to actually manufacture the parts that would have required those skills.

This is all speculation on my part. Sorry, no numbers.

Reprap is definitely good for custom design, so both A and B make a lot of sense. However, custom products are rarely cheaper or faster than manufactured ones. Mass production is very efficient. It can make bulk orders for materials in advance, allowing that to be cheaper, and can utilize machinery specially designed to be the absolute best at what it does: make the same thing over and over. A lot of basic products we have today are only available to the public because of inventions like the assembly line and injection molding.

Cobblers used to exist, and they were very good at what they did. They could make perfectly molded shoes decently fast. But now they are very hard to find, because the attractiveness of good, cheap shoes, outweighed expensive cobbled ones.

I personally think that custom markets like clothes and shoes are very good applications for repraps, which will make customization much cheaper, but I think there will still be a large market for $20 shoes at your local department store.

On the other hand, nuts and bolts will almost never be made by a reprap, because why would you use $0.50 of metal and one hour of time to make one $0.10 bolt? (these figures are random and have no great significance in the real world)

Figures are hard to come by, since only a few repraps actually exist. But we could try and estimate how many repraps it would take to equal the output of an injection molding machine. Just take the average cm^3/hour output of plastic for each. I don't happen to have any of that information myself, though.

And if you're trying to imagine a whole factory filled with repraps, remember, they all require maintenance and raw materials. It might be cheaper to build a small factory filled with repraps, but it would be more efficient to invest in a real piece of manufacturing equipment that can produce thousands of items an hour.

Speaking of clothing, how would you all go about making/shaping cloth in a reprap? That's one issue with clothes, they have to breathe.

-Samuel

Well...

Remember people talking about hacking that cricut? Imagine taking something like Blender, scanning in a 3D model of the subject, painting the clothing onto the wireframe, then using a script to assist in slicing up the wireframe so it lays flat...
then using a giant cricut to chop the resulting design out of a bolt of cloth.

A human, probably, would still have to do the actual sewing, as I have a hard time envisioning a mechanical device sewing the resulting parts together without some significant pick-and-place capability.

Alternately. use the wireframe to map out the proper steps to knit that garment, then have a two-needle robotic knitter start in on it. This is possible now. I've seen some 3-d knitted objects in a magazine, (pre-finished carbon-fiber engine valves.)

I agree, for staple objects, mass-manufacture will be the way to go, (which is why I figured my re-invented cobbler would be using pre-made metal eyeholes and snaps, already available at leather working shops.)

As for the price of shoes...
As reprap makes it possible to manufacture a decent pair of shoes, stores now have to provide manufactured shoes that compete in cost with reprapped shoes. Even if no one uses it regularly, if it's there, it'll force manufactured goods to compete with the cost of a reprapped equivalent. This puts a ceiling on the cost of staple objects. A person might buy a pair of decent shoes...but only if they're in a hurry, or those shoes are sufficiently cheap enough that they don't just go home and replicate a pair.

I figure rep-rap, and other types of micro-manufacturing, will allow manufacture to move back into the community. The cost of being a cobbler, as I see it, are...the skills to make shoes, and the tools to make shoes. Even if you can purchase, or improvise the tools fairly easily, the skills require a time investment. In my concept for cutting clothes, you only need to be able to sew straight, and tweak a model on a screen, (with infinite undos). The tools figure out the pattern for a perfect fit, which makes the skills to be a tailor much easier to come by...they're in the computer. In shoe manufacture, as well, some of the skill is still needed, but other skills are replaced by micro-manufacturing, so the barrier of entry is again lowered. Possibly low enough to encourage some people to take up the craft, and allowing them to be competitive doing it.

Finally to your first point. Custom products are rarely cheaper or faster than mass-manufactured products. With Reprap, this would still be true. However, it might become cheap enough, and fast enough, to be competitive, since custom, in itself, is a draw. And it might be faster than mass-manufactured items, where the mass-manufactured item isn't commonly stocked due to the limited audience interested in it.

One other thing.
It would be cheaper to make a factory of repraps...and it'd be more effective to fill that factory with traditional manufacturing tools...but it might be cheaper to fill that factory with reprapped traditional manufacturing tools.
You can one-off the tools themselves, then get to work making the same item over an over to fill a hopper truck.
Plus with, with reprap, you might not care about the reusability of tools. Make a tool that can't, itself, be retooled. You can replace it when the time comes to change output.
In this way, reprap also lowers the barrier of entry for traditional manufacturing, as well.

Edited 1 time(s). Last edit at 08/27/2007 08:12PM by Sean Roach.

Indeed, repraps would be great for making molds. Presuming you can make decently sized objects in metal, a reprap could indeed make certain parts of the manufacturing equipment. It would probably reduce the cost of the regular machines due to the decreased time of development. By giving the engineers a tool like a reprap, they could make test versions of the assembly line much more quickly than before. And depending on the tolerances, just use the reprapped parts in the final version.

I agree, custom is in itself a draw, but I doubt the reprap will ever get quite down to the $20 shoe level. Modern shoes are quite complex. Still, you could probably buy a shoe model and run it over night, to get your shoe. One thing I've been thinking, is that, though custom shoes could be made by the end user, they would most likely be made by a web rp company. Either you measure your feet yourself, or you go to a nearby representative who does it for you. Then they use their expertise in shoe adjustments and mail you your shoe in four days. I'm not sure exactly how hard it is to adjust your own shoe, but it makes sense that the average person would prefer not to have to do it themselves.

Clothes are even more likely to have a tailor-like individual, and/or photo-booth in the mall measure you, charge your card, and send you the machine-made custom copy later, due to their even greater complexity than shoes. Again, if you felt up to it and had a high enough level reprap, you could buy the latest fashion model, and just tweak and print yourself.

About the actual mechanics of reprapping clothes, they already have computer controlled sewing machines in every sewing shop. They aren't cheap, but they seem to make it really easy to embroider fancy designs on things. It makes sense to have a reprap head that can sew, the only problem being the necessary pieces on the bottom side of the material. If it can handle strange 3D curves and work on certain difficult seams, maybe clothes aren't that far-out. I don't know how jeans and t-shirts are actually made, so maybe I should look that up before I make such sweeping statements.

Interesting thoughts on the knitting tool, sounds useful. Crazy idea: Do you think that a sewing mechanism could be used for wires in circuits? Or the knitting one, for that matter?

-Samuel

>
> Figures are hard to come by, since only a few
> repraps actually exist. But we could try and
> estimate how many repraps it would take to equal
> the output of an injection molding machine. Just
> take the average cm^3/hour output of plastic for
> each. I don't happen to have any of that
> information myself, though.

one of the topics of the presentation i saw adrian give was exactly this question. you may want to check my math, but if a reprap machine made a copy of itself, and one comb per day, then in 17 days it would be on-pace with a commercial comb making machine that makes 10,000 combs per day.

the idea being that exponential growth is very powerful. also, when you're done making combs, you can easily switch to making any number of useful, interesting things. something you cant do as easily with a specialized comb-making machine.

Other questions related to figures, are what the relative costs, time, and upkeep are.

By my figures, 17 generations of repraps at a multiplication rate of double per generation, that would be 131072 repraps. These all require some minimal human assembly. Also, multiplying by the projected target price of $400, that many repraps would cost $52.4 million. I have no similar figure for a comb-making machine. Also, each of those repraps individually requires resource feeding and maintenance. No idea how hard that is, or how much it costs. Also, what about energy? I have no information on that either. The space requirement for the repraps would be approximately 2500mm^2 each, totaling about 32.7 million m^2. That's a big building. The volume of the building would be 163.84 billion m^3. In contrast, the Boeing Everett factory is 13.3 million m^3. And that's with the repraps stacked right against each other.

Presuming about 6 hours assembly time per reprap, that would be a total of 786,432 hours. Or 92 man-years at 24 hours a day.

Of course, they would produce a ton of combs.

All I can say is, it will take way more than 17 days to build that many repraps, and you're going to have an army of assemblers to make them. Thanks for giving me some figures to work with, it gives some sense of what we're dealing with. Also, I'm notoriously bad at arithmetic, so if you see something wrong, let me know.

-Samuel

Edited 3 time(s). Last edit at 08/27/2007 09:28PM by Samuel.

Samuel,
point by point.

one. No further comment.

two. repraps need only get down to the ability to make decent shoes for $20, (materials, wear, and energy,) and the cost of shoes has to be reduced to $19 to compete. Thus, for very basic and cheap items, mass-manufacture will almost certainly survive...but the price is forced to stay competitive or RepRap will simply route around it.
Also, the point I was making is the clerk doesn't need to know how to adjust the shoe. If the software can take the numbers, it can adjust the shoe before cutting the parts.

three and four. The problem I see with giving a sewing machine the job of making clothes is holding two disparate parts together and in place while the sewing machine stitches them together. Embroidery in the style of a two-axis plotter doesn't seem like the same type of challenge to me. This is why I figure humans would stay in the loop there. Someone has to pick up the two pieces, and feed them through the sewing machine, or else the sewing machine has to have the ability to do this for itself, with a high degree of flexibility.
But the computer could cut the pieces perfect, every time.

Sewing wires? I think it's pointless. knitting wires? Again, likewise, unless you're knitting some sort of interactive, or reactive, garment, (a sweater with a billboard knitted in, for instance.) It'd probably be easier to just paint the wires in, where rigid objects are concerned. It helps to know how the lock-stitch mechanism works, but I just don't see it being used to put wires in solid objects.


Your thoughts about a web-based RP company are compelling. It may very well be that manufacture starts out as a mail-order custom shop. There isn't, however, much* reason why it can't also emerge as a local shop. Overhead is nil, as most of the materials are either common to many items, (and thus purchased in bulk,) or have an upper capacity, (such as a reprap, able to print one set volume of objects at a time.) So, if the market develops in a reasonably small geographic area, custom shops could find their way back into the community. You'd know your tailor, cobbler, hatter, bookbinder. I personally hope for this, as it would be easier to bootstrap into competition. If you decide to take up some maker craft, you wouldn't need to compete against million(billion?) dollar factories in china or anywhere else. Just develop what skills you still need, and set up shop. If you can't compete against your neighbor of twenty years experience, move, or specialize. So, yes, much of my argument may very well be wishful thinking. I'm hoping that reprap will undermine the ability of the high-dollar factory to undercut local competition by economy of scale alone.

*I can see one reason why something like this could stay mail-order based. If there is insufficient demand for the work, the work would stay centralized to best utilize the resources. If a neo-tailor can produce a shirt in a half hour of labor, then one guy, able to sell his services a total of, say, once per day, would be busy only about 30 minutes a day. He'd have to get all his cost of living into those five shirts a week. The web-based guy, able to concentrate orders, and able to acquire, and fill, sixteen orders a day, could charge 1/16th as much for his labor, and still support himself at the same wage.

You're willing to postulate a future reprap that can produce clothes, but not one, eventually, that is capable of self assembly?
Also, that $400 target price, as I understand it, are for the parts that rep-rap can't make, as the raw materials for what it can are very cheap...unless you have to go through stratsys to get them.
A future, later generation, reprap, might very well be able to produce its own motors, which cuts into the price significantly, along with its own circuitry, (less a few small, but highly complex, components.)

I suppose the ultimate goal would be a reprap that produces between eight, and twenty seven modules, that when plugged together become a reprap, or one module, that when it unfolds itself, becomes a reprap.

As for energy.
As I see it, there are only two costs in this world. Labor, and (perceived) rarity.
Energy has perceived rarity, as much of the supply comes from OPEC, and they decide how much to sell, in order to control the price.
It also has labor. First in the efforts of men to find sufficient sources, (of oil, for instance,), then of men to dig down to it. Don't forget the labor involved in making the machines to look for it, and dig to it. Or the labor involved in mining the materials to make the machines...
Or the labor AND rarity in making a geologist-prospector out of a reasonably intelligent untrained individual. (How rare is it for an individual to actually be willing to, A. not earn a paycheck while paying someone else to teach them, and B. forgo other immediate pleasures to pay attention to the lesson? For any given subject? That eventual payoff better be worth it.)

If the reprap can eventually work with very basic, very common, ingredients, the cost gets dropped down to about that of flour. If it can produce itself, it can likely also produce energy collectors, (windmills, solar panels, water turbines, etc.) This means the energy costs can also be fought against.

Of course, by the time a reprap can completely reproduce itself, everyone would have one, so maintaining a factory of repraps to produce a large number of one object becomes pointless.

Sean,

Nice thoughts, I can't really disagree or comment on any of them.

It seems that there are several markets that would develop through the conjunction of the internet and the reprap.

1) Custom rp companies. Both web and local. The web companies would try and take advantage of volume and lower the price, but in the end may be unsuccessful at beating the locals, unless they aren't in an area because of low demand. Not sure.

2) Web stores selling models. This could be anything from shoes to forks to the latest fashion in clothes. You go shopping with an interactive 3d tool, inspect and rotate all of the models, decide which you like, and pay $5+ dollars for it. It would make repraps a bit more accessible to those not quite willing or capable to invest time in their own models.

3) People selling model generation tools. These would be software devices taking inputs such as foot measurements and automatically tweaking certain models they came with. Much like advanced web design tools, this would make model generation more user friendly.

I'm sure that there are many more different kinds of businesses that one could run just armed with a reprap, but unfortunately, I accidentally labeled them all under "custom rp." Silly me, making something so generic. For instance, you could set up a "photo booth" at the mall, and instead of producing prints or those acrylic bubble statues, you could give a 3d model of the person. Or send the data to the clothes shop. Or put it online, to be used in some sort of Second Life game, in which your character looks exactly like you. Who could then in turn buy real clothes for you from a rp company with a store on Second Life.

The possibilities are endless, and I don't have enough time or finger stamina to think of them all.

Thoughts?

-Samuel

True, I did kind of overlook the possibility of self assembly, but I had presumed that the figures Zach gave were pertaining to the current gen of reprap. And on that presumption I based my calculations.

Personally though, I think that the reprap, though a very few will get to the high-level of clothes production, will for the most part be marginally limited in capability. Also, I think that the user base will be very vast before any repraps even approach total self assembly. Thus, I think we will have the rp vs. manufacturing scenario long before that. Thus, I think my calculations still hold at least marginally. Obviously, they are hypothetical, and just meant to compare the usage of large scale rp to large scale manufacturing.

I do agree, though, that if total self replication is achieved, it will only be at minimal cost that all those repraps will be produced. But they will still be much slower and take more space than the manufacturing equipment. Besides, if we are to presume such developments as total self replication in repraps at the time of the contest, what kind of developments will have been made in regular manufacturing?

-Samuel

Thoughts.
You don't stand a chance. Finger stamina alone wouldn't be enough.
You have but one brain to conceptualize with. The best way to figure out how many unique and interesting ways a reprap can be put is to make a few (thousand), and check back in a decade or so to see what developed. Six billion people are so much more creative than a mere thirty or so.

As for local craftsmen. Reprap would be but one tool of many. Many would be related to reprap, (such as a cloth CNC machine tied to Blender), while others would be traditional tools used in that trade for decades, even centuries.

What you can do with JUST a reprap, however, is probably a tough enough mental puzzle. What you can do by adding a reprap to any trades existing set of tools...

Oh. And I probably paid $200 for the OS on this laptop. OEM is cheaper than boxed, after all. Some things may very well sell for a bit less than an hours wages at minimum wage, but some things may sell for much more...of course, pirated and reverse engineered versions will be everywhere.
I figure most people would use a reprap to reproduce others designs, actually. RussNelson brought up the argument that basically boils down do, "do what you're best at, and trade for what others are better at." Plenty of people will use reprap, who are self-aware enough to know they have the creative skills, where tangibles are concerned, of a slime mold.
Do you own a MP3 player? When's the last time you produced your own music? You own a computer? When's the last time you sit down and wrote a utility to run on it? I bet you have 2-d art, hanging on your walls, that someone else created, too. Granted, in this crowd, the odds are fairly good you have recently taken part in programming. Point is...eventually everyone will likely end up using RepRap. The creators will be a minority as small as the percentage of musicians, graphic artists, or programmers, on the internet.
Granted, you did say five plus dollars, not five dollars...

Actually, with self assembling rep-raps, capable of initializing the new copies, or at least capable of producing the means to initialize the new copies, (read, reprapped robots,) and with dirt-cheap materials, (I THINK potting soil is probably more expensive than flour,) they would merely take up more space. They would not be slower. I think that was his point. Your 131072 repraps, AS a factory, would keep up with a typical factory. On the following day, it'd be twice as fast as a typical factory. It'd take up a lot more space, however.

As for what developments would be made in regular manufacturing...probably reprap would be such an integral part of it, no one would consider manufacturing without repraps on hand for the customized parts, or for equipment repair.

I suppose it is true, that in this decade of cheap printing, traditional printing still holds a place in the world. Newspapers are printed much as they were centuries ago. More so, as, at least as of about 1982, they weren't using movable type anymore, but rather replaceable aluminum plates reminiscent of woodcuts. Used for one run, and discarded.
Of course, at that time, it was the adaptation of photography into the craft that had altered the nature of newspaper printing. I have no idea what computers have done to the printing press. (Oh the things you remember from a grade school field trip.)

hey guys,

the exponential growth was just an example of how powerful this technology *could* be. i'm not suggesting that its practical for comb manufacturers to do this sort of thing, only that if we're successful with reprap, that this type of growth is possible.

one very interesting application of a technology like this is space exploration. send up a seed-module with a reprap in it, and then it harvests raw materials from the local environment, and then creates whatever it needs on the spot.

pie in the sky? yes. something i believe humanity can achieve someday? definitely.

Its mostly about what MIGHT be possible, this is afterall a research project.

Personally I imagine small scale production facilities popping up around the country, similar to the old town blacksmiths. A lot of country farms had a basic forge onsite but took more difficult jobs to a local blacksmith. In the foundry business these are called job shops and their are generally a lot of them around. I think the existing job shops will be VERY interested in getting rid of the expensive collection of patterns they have to keep in stock currently.

A customer places an order for something and a reprap type printer produces postives for the parts that can be made from plastic, and thermoset molds for those parts that have to be poured in metal. Then the finish parts are packaged with the stock electrical assemblies and the customer can pick them up. The customer can pay to have the parts assembled or just pick them.

I can imagine something like a air-fluidized mold filling system being relativly easy to build with reprap materials and certainly the molds would be easy to make. That would allow me to make replacement parts for a large number of the materials and parts in my everyday live.

So maybe I cant make a thousand combs a second but honestly I only need a handful of the things anyway.

Mike

ZachHoeken Wrote:
-------------------------------------------------------
> one very interesting application of a technology
> like this is space exploration. send up a
> seed-module with a reprap in it, and then it
> harvests raw materials from the local environment,
> and then creates whatever it needs on the spot.

I think this is pretty much what NASA had in mind when they came up with the estimate of approximately Pentium-4 complexity you've mentioned elsewhere. I quickly got lost in the sea of acronyms in that report, but the general tone seemed to be that it was quite doable.

I was just trying to illustrate that it doesn't make sense to use repraps in place of conventional manufacturing. The resources that were spent making the vast amount of repraps would be much more efficiently used in a regular manufacturing assembly line.

I agree entirely with Mike's statements. I don't need that many combs either.

-Samuel

I think that conventional manufacturing of bulk parts will always be cheaper than any replication machine. After all, people are a replication machine, and yet we have resorted to conventional manufacturing.

well yeah, a specialized process is generally always going to be more efficient at a certain task than a generalized device. but it loses big time when the process needs to change, even slightly.

thats the major advantage of reprap, imho. after you get over the self-replicating aspect, is that what you are producing simply becomes a software file, that you can change at whim. rapid development times, extreme customization, and open source collaboration object development all become possible.

also, if reprap gets to the point where it can fabricate specialized devices to do high volume manufacturing, then the whole point becomes moot.

I think that the injector molding machine is going to be the winner in the race for the manufacture of cheep standardized combs.

However, if there were a demand for very-low volume custom-designed combs as a fashion accessory, 3-D printing is likely to win out.

Sounds like an interesting business venture, drvanthorp. Maybe you're on to something.

ZachHoeken Wrote:
-------------------------------------------------------
> well yeah, a specialized process is generally
> always going to be more efficient at a certain
> task than a generalized device. but it loses big
> time when the process needs to change, even
> slightly.

A good example is whats going on at the moment with the 1.3 revision of the PowerComms board. Four release candidates in a couple of days! If we had the reprap v2 which will be able to print circuit boards then you've also got another speedup so you don't have to wait for a manufacturer to create the boards for you.

how about a comb that is custom tailored to your particular grip, hand size, etc.

>
> A good example is whats going on at the moment
> with the 1.3 revision of the PowerComms board.
> Four release candidates in a couple of days! If we
> had the reprap v2 which will be able to print
> circuit boards then you've also got another
> speedup so you don't have to wait for a
> manufacturer to create the boards for you.

exactly. you also dont need to buy a large number of boards to get a decent price. being able to run off one board, at a very reasonable price shortens the development time and makes things so much easier to work with.

imagine how hard writing software would be if instead of compiling you had to ship off your code to some manufacturer to have it assembled into a working program. it would be a nightmare, and that in and of itself would put it out of reach of a large number of people.

the idea of putting this technology in the hands of amateurs also has some very powerful implications.

reece.arnott Wrote:
-------------------------------------------------------
> If we
> had the reprap v2 which will be able to print
> circuit boards then you've also got another
> speedup so you don't have to wait for a
> manufacturer to create the boards for you.


I don't think we need to wait for RepRap v2 to print circuit boards. I'm convinced that with the right software and a willingness to do a bit of manual labor (read: place in etching bath) v1 will be perfectly capable of printing circuit boards. If the extruded plastic for some reason doesn't work satisfactorily as etch-resist, it will be trivial to mount a pen with etch-resistant ink (there are a few leads online I'll be looking to follow up, such pens definitely do exist) on the RepRap print head. PCB creation is, I hope, a lot closer than we all think.

drvanthorp Wrote:
-------------------------------------------------------
> I think that the injector molding machine is going
> to be the winner in the race for the manufacture
> of cheep standardized combs.
>
> However, if there were a demand for very-low
> volume custom-designed combs as a fashion
> accessory, 3-D printing is likely to win out.

Exactly. But don't forget that reprap could be made to print out tooling molds for injection molding machines, bringing that technology closer to the individual too. This way we could have the benefits of both production processes.
The injection machine could be owned by that small scale production center and would be shared by many RepRapers. Molds could be kept or donated to the comunity, some kind of a sharing basis. Mold STLs could be GPLd and would be open to improvement or forking...

basically this is an extension of RepRap phylosophy to other production processes you would pay for as a specialised service.

Personally, I think I'd just grind up the molds, and re-use the plastic, and metal.
I believe greensand is already reusable.

Being able to print them out on demand, (with some small delay,) makes storage moot for anything more than sentimental reasons.

Also, I believe this thread has gone circular.

The issue of clothwork and knitting have come up a couple of times so I will put my bit in. My wife is a seamstress and I am a semi-skilled knitter. I have tried in the past to knit thin gauge wire, the problem is that raw wire doesnt bend nor give very easily. If you were to use a spool on thin braided and insulated copper wire (speaker wire would be a good example of what I am thinking of) in a shuttle design (think weaving not knitting) then you should be able to run a staight line across the work surface. Then you can lay layers on top to tack the wire down and run the shuttle back across to force the wire to move up a layer, tack the higher layer and repeat. The result would be a snake of wire moving up the part. You would likley need at least two shuttles to build different ciruits in the same part but it would be possible. Run the shuttle on a rail and have the rail attached to a tool head and you could move it in and out of the work space as necessary, and then just trail slack out or cut the wire as needed. Automatic looms have been in use for a very long time.

Clothwork is far more complicated, the major issue is that almost nothing is made completely by machine in the textile industry. I am not aware of any machine that can sew sleeves on for example, because it is so much cheaper to have that step done by hand in China or South America. Designing a machine that can "ease" the sleeves, basically build in small imperfections so that two 2-D objects come together and form a 3-D one would be extremely difficult. Such imperfections are easy for an experienced human but incredible difficult for a computer to even comprehend let along execute. The crotch of a your pants and the collar are other points that require "easing" in order to fit.

Knitting machines do exist and they ignore the "easing" requirement because knitted material can be both created and shaped at the same time. That isnt possible with broadcloth. However most knitted garments are still made by cutting and sewing due to the cost advantages of making large sheets of knitted material and then treating it like broadcloth. I have made one "seamless" sweater where by increasing and decreasing I shaped the final result, it takes significantly more planing but not really more skill. The technique is far better suited to a one off reprap type siuation rather than a tradition factory. Triple zero sized knitting needles can produce cloth from thread that is for all intents identical to normal woven material, although it does have more stretch (its what T-shirts are made of).

My suggestion is that we consider the textiles industry, as I have said in the past a good seamstress knows more about creation and control of small scale manufactoring than most engineers. But as a process knitting is our best bet since the stitch by stitch creation process lends itself a lot more to computer control than the "easing" that is such a large part of traditional clothwork.

Mike

The thoughts and ideas expressed in this post do not reflect those of my employer and are intended only as communications between individuals. Any attempts at implement are at your own risk

Edited 1 time(s). Last edit at 09/12/2007 09:03PM by ohiomike.

Hmm,

You may very well be right about "easing". My "experience", if you can call it that, is from tinkering around with unrelated things, and extrapolating Potential alternate uses for them.

I'll list the one I think could, potentially, handle your "easing" problem.

Blender.org

The program that you'll find there, is a highly flexible 3-d modeling program, (with a very steep learning curve.) While playing around in Second Life, I learned they were going to add a new type of prim, and got to looking around for how to implement that in Blender. This is a way. This same method could, possibly, be extended to "easing".

Basically, you start out with the 3-d model, then start figuring out where the seams will go. After you've gotten all the seams figured out, you "unwrap" the model, and you end up with a wireframe that resembles the original object about as much as a mercator projection does a globe.

Throw in laser 3-d scanning, (of a bathing suited figure,) and it appears to me that machine fitting of garments becomes attainable.

Perhaps I'm too naive. I can see potential pitfalls with my idea. I just don't know if they're shallow or deep.

Kyle Corbitt Wrote:
-------------------------------------------------------
> reece.arnott Wrote:
> --------------------------------------------------
> If the extruded plastic for some reason doesn't
> work satisfactorily as etch-resist, it will be
> trivial to mount a pen with etch-resistant ink
> (there are a few leads online I'll be looking to
> follow up, such pens definitely do exist) on the
> RepRap print head.

Yep. I bought one of these pens at Radio Shack once, as part of a kit for making small circuit boards. Shouldn't be hard to pen-plot a circuit board, but it might be just as easy to print on transparencies, and use photo-sensitized boards; more or less how some mass-produced boards are made.