Many of you know that one of my primary interests in robotics is the development of a 3D printer for RoGR and DIY enthusiasts. 3D printing allows the creation of otherwise impossible-to-make objects. Here is an example from a great sculptor, Bathsheba Grossman:



The piece above was made by deposition printing, then 'sintered' in a blast furnace. Of course the model could be plastic, wax, starch etc, so long as it can be deposited one layer at a time. For sintered pieces, metal powder is included in the deposition media.

Needless to say, the machines that do this are very expensive! They are also quite small and are usually very proprietary, including the media. Plus, some of the methods seem downright stupid to me, well, I guess we'll see!

I had been contemplating on how to build a fabric printing machine that could print fabric by the bolt, with any design you create. I'm not yet ready to publish on that, but in researching I found the stepper-driven metering pump:

Click to view attachment

This pump is valveless, self priming, goes to 100psi and is adjustable from 0 to 100 micro-Liters (0.1 mL) per rev.. 1µl is the volume of liquid that a 1mm x 1mm x 1mm cube contains...perfect!

Normally the pump costs about $500, but I had the chance to acquire several dozen new pumps for substantially less due to a corporate bankruptcy (ahem, surplus).

Hence, the Lumenlab '3DP' is born! (Several of you lucky prisoner/pioneers will soon have the opportunity to try this at home!)

Now, what and how to deposit?

I've been considering many media...but we need to question what we are making. For the baseline machine, I decided on a durable plastic. Naturally the plastic must be a flowable liquid, but set very quickly on deposition. Hard waxes and thermoplastics were first considered, but I thought 'hey, if I gotta heat the stuff, why not use tin instead'? Alas, another time perhaps...

Epoxies, glues, resins; all have complications or undesirable properties...

Robin was musing on the use of UV lasers for lithography which jarred my addled memory, reminding me of the UV curing polyester resins used for emergency surfboard repairs (surf's up, emergency!)

Most of the resins have longer curing times of 15mins, too long. Research lead me to UV curing urea-formaldehyde resins which cure in 10s or less. The old brand name of this plastic was Urethane.

This resin has a viscosity of 1000cps (about like castor oil), and cures clear and resiliant in seconds when exposed to 10mW/cm of 400nm UV radiation. It is used to encase objects and trinkets in clear plastic, like key fobs, lapel pins etc...sweet, it just might work!

Next was to find the light....10mW/cm doesn't sound like much...

Well whaddaya know, a 12mW 400nm LED for $1.20; this is too easy!

So, the plan is to used the stepper pump to precisely meter the resin through a deposition tool.. I'll use a small hypodermic needle as the tool. Clearly the diameter of the needle is the max resolution. Trailing the tip is the UV source, however, if the lamp is down by the tip, it will interfere with the tip's resolution, so a fibre optic thread will carry the light.

That is all for now. I look forward to hearing your ideas!

-bc

Very cool idea, I'm curious to know how color would affect the drying time of the material. In other words would a tint mess with the uv spectrum?? can you hook up the stepper pumps to a LVLP head and lay down a "factory finish"? with a bot?

I'd imagine there are options for urethane colorants, or you could always "dunk" the part when finished. For robotic painting...sure, airbrushes come to mind. Precise pressure controls can be had with pneumatic transducers, allowing very fine art. For stuff like producing RoGR pieces an LVLP nozzle could be attached to the gantry and the pieces sprayed when the machining was done. There's nothing stopping anyone from using multiple tools at once on a gantry-bot, but most designs make it impractical, so you rarely see it. With RoGR, I want it to do as much as it can, so I've tried to leave the design very "open" and modular so it won't be hard to add things...the next robot I build is being made specifically to make the RoGR frames, and rest assured it will paint the pieces too!

OK, I know you can make stuff like that first picture using this process, but I have to ask, (and please don't hit me), why would you want to?

Why wouldn't you?

A few good reasons:

Click to view attachment
This photo shows a ceramic mold made by 3D Printing and an orthopedic knee casting poured into a similar mold. The knee casting has been polished on the back side as can be seen in the reflection in the mirror. The casting is made of a medical cobalt chrome alloy.
Photo Credit: MIT 3DP Lab

Click to view attachment
This photo shows a family of injection molding tools made by 3D Printing. The tool in the lower left corner is approximately 150 mm long. Many of these tools have conformal cooling channels printed within them. Such channels lead to far better temperature control during the injection molding resulting in faster production and high quality parts.
Photo Credit: MIT 3DP Lab

Click to view attachment
This object was printed directly from a mathematical model and would be extremely difficult to manufacture by conventional techniques. It is intended as an illustration of the geometric capability of 3D Printing. This model was made by Z Corp., maker of what are said to be the world's fastest 3D printers.
Photo Credit: Z Corporation

Click to view attachment
This photo shows a ceramic filter made by 3D Printing. The cut away shows the internal passages created as part of the printing process. 3D Printed filters are being applied to the clean-up of power plant emissions.
Photo Credit: MIT 3DP Lab

Click to view attachment
This 3DPTM model of the dome at MIT's Great Hall demonstrates the intricacy and detail possible with the process.
Photo Credit: MIT 3DP Lab

Click to view attachment
By using different printer jets to produce different parts, the 3DPTM technology is capable of creating color-coded objects, such as this model of a compressor.
Photo Credit: MIT 3DP Lab

Reproduced from: http://www.nsf.gov/od/lpa/news/02/3dprinttip_images.htm

I think you just hit me!!!

You do know I was kidding right?

In all honesty it's an awesome project and I wish you every success, and can't wait to see your results as always. The mathematical models aside, while cool, are not in my mind as cool as the more practical objects that can be made with this process. I think the ceramic filter and compressor model are pretty cool.

(OK, did you just edit and put in the first line, or did I miss it first time round?)

No worries, I'm not that "sensitive" . I did go back and edit though, only because I had to get the "copy and paste" operation over with first.

I use math models because they tend to show what can really be done with the tech. The practical uses of 3DP are momentous, but without the "cutaway" on the ceramic filter, it'd be hard to comprehend how cool it really is.

Ooops...

1l is the volume contained by a cube of 100mm x 100mm x 100mm.... Thus a cube of 1mm x 1mm x 1mm would be 1µl
1ml is the volume contained by a cube of 10mm x 10mm x 10mm

Just be careful with the amount of revolutions when depositing something with that pump ...

So when you are placing liquids with the technology does the point of the needle have to touch the product. like a sewing machine needle head releasing liquid so the volume of liquid held in the chamber under pressure would be the 10mm cube. Kind of like charging an airless paint sprayer with fluid.
the airless sprayer I have has to charge 25 feet of fluid.
sorry if these questions sound stupid just mumbling out loud.

Oops, I just boofed copying the specs...all fixed now. Should have read .01-.1ml per rev, so no worries. Thanks for catching that!

For my first attempt I'll have the needle simply deposit a small amount of urethane directly upon the surface of the model. The pump will control the rate of flow precisely. The needle will most likely be sheathed in a bundle of very fine fiber optic cable which will carry the UV light to the tip of the needle. It is important for resolution to keep the tip as small as possible.

just read a bunch of the mit labs papers they al talk about laying down dust then the liquid material or slurry

Yes, that's one of the 'stupid' ways I refered to...but until I succeed, I should refrain from judgement.

Had another funny idea; chocolate deposition printing!

yeah chocolate, using the MIT lab tetrahedron maping and different flavors of chocolate adding in flavors like ancho chile and caramel and creamy nugat center
you could lay down the flavors so as you nibbled through them it would enhance and heighten your awarness of the chocolatey goodness inside there by filling the god shapped hole into which all chocolate must enter.

tempering and controling the set temperature would be the variables. with colored chocolates you could make an edible visible woman fore instance. just sayin..

When I worked for a company that made guides for steel rolling mills, making "test" castings were very expensive as they needed to create tooling for the casting. We saw a demonstation of a deposition printer that worked in this way. The material was laid down starting from the bottom and working is way upward. The binder, or activator was applied via the nozzle and the final part was then "dug out" of a "box" of media. It was really very cool to see this done so quickly. These parts were then used to make a single mold for creating the test casting, but was destroyed during the process, so it was pretty much a 1 off... of course that was it's proposed use... we only ever got 1 or two test "prints" from the deposition printer, then it was abaondoned, due to it's cost...

I know that the local Sam's Club uses this technology (of sorts) to create an edible image on cakes. You give them a photo, they scan it in and print out the image within a thin layer of icing that is then applied to the cake... imagine one of the LL robots actually decorating the entire cake. Design cake, order online, LL robot prints (decorates) cake to your exact specifications. I love cake.... and beer...

Cake printers print on a thin sheet of sugar/gelatin with a paper backing that is peeled away and the sugary layer deposited on the top of the cake. It uses a bubble jet to lay down the food coloring onto the sugar-paper. This is technically 2D. With a RoGR and some hacking you should be able to print the actual cake.


That sounds very tempting...

I was thinking last night the dust bed probably supports the form so the goo doesn't ooze all over the place. Then I had a flash of meringue and cookie crumbs Nilla wafers, chocolate wafers, graham crackers crushed and the meringue deposited into a bed of cookie dust then sintered in the oven low temp like 150 /200. Remove the baking pan and shake out the crumb. You could probably make some very cool edible designs.

Yea, I've actually been thinking similar thoughts since the "food printing" issue arose.

Yay, got the micro-metering-stepper-pumps! How can life get any better??

Click to view attachment

Nice Brain!, i think 3d printing will be my primary reason for building a Rogr, Ive been looking into different 3d printers and came across the Reprap which seems pretty neat, but i think i want more precision. Maybe some sort of Laser Sintering?

I cant wait to start experimenting!

Hi Phife... A RoGR can make a big-ol' deposition print for sure! (Well I haven't tested my plan, but it's one of those "made in the shade" plans that always works...heheh [devil]).

For the actual sintering; on this I need more experience, but the "big" issue is the blast furnace to ~vape-off your model. Like most things to me; it all seems so simple....

Now to just build a time-machine that can amplify my daily hours to 3-4x what they are....

The "food printing" concept has been haunting me...I am somewhat obsessed by the notion that a "pizza" could be "printed", or even a "sandwich". Rather than "squirt-baking" some plasticky foodish stuff, I've wondered on the larger concept of printing "digital samples" of real foods. Most everything in the Universe can be reduced to a handful of elements (99% of the mass of the human body is made up of just six elements: oxygen, carbon, hydrogen, nitrogen, calcium, and phosphorus); but the fantastically complex sequencing of the elements is "near-random" as is currently computationally possible....or is it? Can a human be printed? There are roughly 100 elements that comprise everything you've ever seen, eaten, driven, birthed etc....

Agh, how??

Well, given that it either takes or releases energy in order to take those elements and turn them into the various compounds. (2 H2 + O2 -> 2 H2O +heat for example) This creating the water is likely to cause a spontaneous combustion of all the other hydrocarbons if you're going to do it at an elemental level.

From a molecular level is a BIT more feasable, but then the variations on those atomic compounds is several orders of magnitude more complex, and that's before we get anywhere near the RNA/DNA level.

I'm going to leave the whole spiritual/metaphysical thing out of it, but if you could create an absolute copy, to the atomic level of an existing human being, you'd also have to impart the correct amount of electrical and thermal energy at the same time, or else you'd be producing a corpse. This is perhaps less of a tragedy if you're producing a corpse of an animal which has been domesticated for food. (Would there be ethical issues around this? Given that the raw chemical materials for a steak vary only slightly than the ones needed for tofu and the input energy would be identical, gram for gram. The steak would even have more food energy in it, if it has very nicely marbled fat...)

Basically, at this point, you're no longer looking at any kind of printing process. For one, you have literally no time to do it. Leave a brain, for example, without a heart and lungs for the time it takes to "print" them, and it's going to be past it's "best before" minute. Similarly the heart won't so so well without the lungs and brain, and the lungs need the heart and brain.

No, even the most precise 3 dimentional printing (printing being any kind of sequential layout) isn't going to be up to the job.

Just FYI..

One pound of soybeans yields about 3-1/2 to 4 pounds of tofu, whereas 12-20 lbs of soybeans are required to produce just 1 lb of steak.

This is why steak is rarely eaten in impoverished cultures (sans USA ghettos)...

Otherwise the protein content of the two is nearly the same, and tofu isn't normally associated with bad health.

To be fair, since cows are ruminants and are meant to eat GRASS, it may be that the processes involved in getting them to digest what they were not evolved to digest may have something to do with the health issues related to eating their meat.

I'd certainly agree with you. When cows ate grass, people were more healthy, the meat was leaner and probably too "gamey" for domesticated "tastes". The cows' habit of eating insects played a large role in the "depth" of the nutrition profile; yet the econo-agri-complex focuses on "poundage"; ie quantity over quality: increase the watery chemical pain! (In the higher-end organic markets this is changing).

That said, and without intended provocation, if I can live the same quality-of-life without causing (visceral) harm to another being...why not? My biggest neglect as of late is to simply take care of myself....most animals get better treatment from me!

And for what it's worth...I'm not trying the "print a human thing" right now...

Wait, you're saying 1 lb. of soybeans becomes 3.5 - 4 pounds of tofu? You're neglecting a serious additional component here (or violating the Law of Conservation of Matter and Energy).

I believe that what Supra was trying to say is that you're "printing" the beefsteak. Thus your concern over cattle consumption to yield ratio is irrelevant. If you can simply re-arrange the COH molecules from something - like the very grass you'd normally feed to a cow, you've got a massive net win.

Additionally while they may be relatively equivalent in protein, the steak would have much more food energy as fat is almost twice as calorically dense as a carbohydrate, not to mention the additional mineral content.

Sweet! Violating laws late at night! ( PS: The rest is water )

Well, however your steak-printer works is up to you..

Well I'm thinking that either the missus would think that it's been a long 9 months, or will be asking that you get the thing working before the next Brainchildchild.

... it sure is hip to be square(d).

Heheh, counting the days. Though we've decided to stop at two kids, a man can still try, right?


DOWN WITH HUEY LEWIS!!!!

Famous last words... I said the same thing - and our third is due in three weeks.

Best to your family!

Oh man...

To yours as well, cheers.

Oh man, this sounds great.

Watch out for the type of fiber optics you use. Many of them will filter out UV naturally (many plastics, some glass) and others have doping agents to tune their optimum internal reflection frequency.

Indeed, thanks for the comment. A UV-400nm strand was specified though, as determined by the f/o glass by composition (silicates, boron etc)...though by no means is the final design of the DP determined!

Everyone on this forum is so ridiculously crazy.

BEST FORUM EVER.

Concur.

The dental industry uses a uv hardening process to create temporary caps/crowns. The composite used probably isn't adaptable but maybe the uv generator / delivery system is.

Well, the urethane I selected cures nearly instantly with a simple UV LED.

Alright! I finally have something to show here....hey, I take a day off to work! Kellie and I are getting close to understanding this 3DP deal, I can tell it won't be long before the last few pieces click into place. The tech is so young that there really isn't an easy path to 3D printing, but if you can hack a bit you can make it happen. Since I can see we are about ready I decided I'd better have some crude DEPCAP* lest I be the laggard, so I set to creating one from an empty "precision oiler" tube, some vinyl tubing, heat shrink, a stainless bike spoke, and some hi-temp hot glue but of course!

Click to view attachment

The spoke holds it all straight and supports the DEPCAP:

Click to view attachment

In case you don't remember, I have a high-precision metering pump that is step-driven:

Click to view attachment

The capillary output is about ~.5mm. The screw head to the left is ~8mm. This is probably a huge DEPCAP but I don't mind crudity for a proof of concept.

Click to view attachment

Ya'll may remember that one of my plans is to use UV curing polymer resin to compose models in 3D space. The DEPCAP deposits the resin in voxels or what ever, and a high-intensity UV LED comes behind and zaps the stuff into rigidity. The range of viscosity for the materials is 600cp to 20,000cp, which is within the capabilities of the pump as I tested several materials with similar ratings, like dish detergent and glycerine. Now I can remove one of micRo's Z bearing blocks and attach the DEPCAP; removing the block gives a little over 4" of Z for printing, which is actually considered a "large print".

*(my made-up word for deposition capillary)

How did the prototype 3d printer work? If you are looking to miniaturize your DEPCAP this is probably worth a look:
NanoFil Injection System
Who wouldn't want a 25-50 micron id needle and a pump with nanoliter resolution?

It is ready to run; just awaits the time/push to finish. Maybe I'll have some "free time" soon? It is only a dream right now.

Funny, that NanoFil was the exact type of system I was going to suggest for dispensing, accurate, small amounts of liquid. They have repeatable dispensing capability, high reliability, and wouldn't be that difficult to prototype yourself (minus the syringe itself). Basically they use a stepper / servo to run a linear actuator for the piston with a set of electovalves to control uptake and dispense functions.

Any progress on this experiment?

Pretty much ready to try, but a few hurdles remain; finishing the micRo run is top.

This is so cool, if personal business picks up again this will be a project for me to pursue.

Could these 3d printing processes potentially print with clay dust+something to hydrate it with? Or just very liquid-y clay, over a long period of time. Or perhaps hydro-cal plaster, since it can dry quite quickly.

It would be very interesting to use such a process as the "first draft" of a sculpture and to finish it off by hand. There's something about the natural feel of kiln-baked clay that I don't think can be replaced.

I've had many similar ideas, and there is no reason per se that they won't work.