DLP Projector Stereolithography 3D Printer
Articles Blog

DLP Projector Stereolithography 3D Printer

October 17, 2019

In this video I’ll highlight some of the
engineering design of this machine: the Ember Precision Desktop 3D Printer. It was manufactured and sold by Autodesk,
but now discontinued. All its details, though, are open sourced
and so it might well live on. There are many way to perform 3D printing,
but they all have in common this: they take a 3D model, cut it into many thin, two-dimensional
slices. And then . . . they print these slices one
at a time, one on top of the other to create a three-dimensional object. Three D printing promises a revolution in
manufacturing. One company that uses 3D printing to manufacture
turbine blades reports that it reduces pollution, take a quarter of the time to develop a new
blade — partly because of rapid prototyping with 3D printing — and that they can repair
the blades 60 percent faster. Futurists have imagined mobile production
plants rapidly deployed to disaster zones, that can print arm splints, tent stakes, and
even buildings! One company has pioneered 3D printing of concrete
buildings. While this promise is enormous, today I’m
not creating anything exotic, but focusing on printing this: Odile the Swan. It’s one of several items used to test and
benchmark 3D printers. On this printer the swan is created here,
on the lower side of what is called the “build head” or “build platform.” Odile will be printed upside down here — like
this. The printer will first build the base and
then work its way to the swan’s head. Let’s watch the printer in action. The print head descends into this amber tray
. . . which contains a liquid, called resin, which will solidify . . . there you can see
the print head settle into the resin … and let’s watch the action right here on the
print head . . . here’s a close up . . . the printer exposes the first layer — that’s
the flash of light — then the trays slides sideways. . . and then returns so another
layer can be added. . . You can’t see anything yet, but I’ll
speed everything up and then …. There . . . you see in about thirty-five to forty minutes
the base of the swan appearing — remember it is printing upside down… and by about
an hour the swan is done! Let’s examine the key steps used to print
this swan. Specifically, I’ll look at how that flash
of light creates a solid layer. Why the machine’s build head moves upward,
and then explain why the tray moves sideways. That will take us through the essentials of
the machine’s design. After that I’ll look at a few details of
the chemistry of the resin. Let’s start with that flash of light: the
flash creates a solid layer. Here’s how. The tray is filled with liquid resin. It solidifies when exposed to blue light,
which appears green when photographed through the orange tray. That orange tray shields the resin from room
light, which might solidify the resin inadvertently. To show you how this works . . . I’ve filled
this watch glass with a bit of the resin — I printed the swan with clear resin, but here
I’ve mixed in some black resin to show solidification better. Watch what happen when I shine this blue laser
on it. . . . almost instantly the resin solidifies
— its called curing. What I’m doing with this laser pointer is
pretty crude, so to aim light precisely this printer using a DLP Projector — DLP stands
for “Digital Light Processing.” The tray has a clear window, so light flashes
can cure a layer. This is the base of from a printer we took
apart so I can show you what’s inside. Sitting on the base is the amber tray that
contains the resin. Note that the silicone-coated window of the
tray is positioned over a window in the base . . . And inside . . . there’s a bundle
of electronics that contains a powerful LED — a light-emitting diode. It produces blue light of a narrow range of
wavelengths . . . then some optics spread that beam of light and shine it onto . . . a
device called a micromirror that creates the light pattern appropriate for a particular
layer, then a mirror reflects the layer pattern through the window in the tray and onto the
resin. Micromirrors were developed by Texas Instruments
for use in projectors for computer. This the optical train of the printer. You can see the micromirror if I take off
this lens and look down the barrel. It’s a small chip about three-quarters of
an inch by a quarter of an inch. It is comprised of a bit over a million tiny
mirrors — each about 8 microns by 8 microns square — recall that a human hair is some
eighty microns! — a micron is a thousandth of a millimeter. This drawing shows two of the mirrors. Each of the mirrors can be controlled separately. An electrostatic force generated by a small
voltage pivots a mirror plus or minus 12 degrees. This directs reflected light either onto or
away from the resin. Because these mirrors are so tiny light can
be directed to fifty micron sections of the layer, so the printer has a high resolution
in the xy plane. Underneath the mirrors is a slender post — it’s
a mere one micron or so — and below that an elaborate, yet tiny hinge. It would seem that such an assembly would
be fragile. You picture a million mirrors clattering back
and forth as this machine is moved. Yet, that’s not true: the mirrors are so
small that they don’t respond to the vibrations and shakes from normal handling. Nor, does their pivot wear out easily: tests
show that the mirrors could flop back and forth for eleven years of continuous operation
before the pivot fails. To see the layer images created by this micromirror,
I’ll put a yellow card right here where the layers are created. Remember that what we are looking at here
are the layers of the swan, which are printed one-by-one on top of each other. As we watch the layers, keep in mind the printed
swan: here the thin red line in the yellow block on the swan indicates the cross section
being printed. In the projected images the bright blue areas
are where light is reflected on the resin and cures it; the dark areas reflect no light
on the resin and so it stays liquid. Notice the round circles: these are the support
posts created when the swan prints — there’s more easily seen here when highlighted in
blue. These are later snapped off the swan. Let’s sped up the printing and watch the
swan being formed … all the tiny circles coming up are the support posts at the very
bottom … here the base prints … and then the body … you can see the wings begin to
take shape … these are the top of the wings … the round
blue dot on the right is the swan’s neck … and then the blue flashes disappear as
we reach the top of the swan. Now let’s turn to why the printer’s build
head move up rather than down. It’s called “bottom-up” printing. This dramatically reduces the amount of resin
needed to print the swan. When using light to cure resin there are two
main ways of doing this: the bottom-up — used here — and top-down printing. In top-down, a platform moves down into a
vat of resin while light is projected down from the top. The great advantage of this top down is that
it is easy to expose a layer to light and cure it, but it also has a very big drawback:
the vat must be as tall as the part to be printed. And this requires a large volume of resin
which can be very expensive particularly if the resin has a finite lifetime. In contrast, the bottom up method uses a shallow
tray that requires a much smaller amount of resin, even for tall parts. I find it stunning to watch a large object
being drawn out. Next, let’s look at why the resin trays
slides sideways after each layer. In a bottom-up print the layer is built on
a window. This machine would fail if the cured resin
stuck to the window. Most 3D printing resins have the property
that oxygen hinders the chemical reactions that cause them to solidify. To allow oxygen into the layer this window
is made from silicone — a material highly permeable to gas. Oxygen residing in the window diffuses into
a thin layer of resin just above the window. This layer is 5 to 50 microns thick. But the concentration of oxygen in this layer
is enough to prevent the resin curing directly on the window. Yet even with this silicone window, the newest
layer will still partly adhere to the window. If it does then no fresh, uncured resin will
be able to be added — and so no new layers. So, the newest layer must be separated from
the window. There are several ways to do this. Some printers just pull the layer up. Yet this “direct pull”, as it’s called,
can create problems. This layer is like a suction cup. Now, you can see that by putting a cup in
a puddle of liquid on a plate. As I lift the glass I can feel some resistance
— you can even hear when it separates. And so in a 3D printer to separate layer and
window requires a large motor to lift the printhead. This action might damage the layer. I find it easy though to slide the glass to,
raising it slightly it is follows the contour of the plate. In general, the force from pulling up scales
as the fourth power of the radius of this opening, while sliding scales as the square
of the radius. So, typically sliding requires one-hundredth
the force of pulling up. And for this reason this printer slides the
tray to separate layer and printhead. Other printers peel the layer off with a rocking
motion. The key takeaway here is that printing parts
with large cross sectional areas is very difficult. This lattice has a fine, detailed structure
and looks impressive, but because it’s mostly open space it’s actually one of the easiest
types of objects to 3D print. That’s why demos from 3D printers rarely
show a thick, brick-sized object like this. The forces are much greater, print speeds
are slower and it’s impossible to print for many stereolithography 3D printers. We’ve been looking at the mechanics of how
the printer works, but this printer relies on both mechanics and chemistry. The combination of precise motion, micromirror,
and fine-tuned chemistry creates the resolution of this printer. The resin contains three main ingredients. First, there are two types of molecules that
will together form the rigid network. These molecules come in two sizes: a monomer
and an oligomer — the latter is just three or four monomers bonded together. Second, a photoinitiator that, when struck
by blue or UV light, starts a chemical reaction that links together the monomers and oligomers. This solidifies the liquid resin. It is the balance of the monomer and oligomer
that yields a piece that is rigid and well-printed. For example, here’s a tiny boat — it’s
often called “Benchy” — that we printed using only momer, no oligomer. If I squeeze it you see that it’s no longer
rigid, and in fact, it’ll even breaks. And equally dramatically is leaving out the
the third chemical — a UV blocker. This chemical prevents the blue light from
penetrating too much past the layer being printed. Here the boat is printed with the UV blocker. You can see the resolved details in the boat. Keep your eye on the circular hole in the
middle. Here is the same boat printed without UV blocker
. . . compare the hole here and here …. Without the UV blocker the light cures regions of
the boat that were meant to stay liquid and lowers the resolution significantly. There’s of course much more to be said about
an engineered object like this, so I’ve linked to other videos that might interest
you. Including several that describe the micromirror
in detail. I’m Bill Hammack, the engineerguy.

Only registered users can comment.

  1. Hi Bill, I just wanted to say that you inspired me to study engineering. I love knowing how sertain things are build and that is mostly because of your video’s. Keep up the great work! By the way sorry for my english, but I’m Dutch ?

  2. You look like Andrew Carlssin, a real self proclaimed time traveler who was interrogated by the FBI in 2003 after he had an impossibly perfect run of good luck with stock exchange that helped him turn $800 into hundreds of millions of dollars in a large streak of highly suspicious investments. The story goes that, Andrew Clarlssin confessed to the FBI in a 4 hour long interrogation that he was from about 240 years in the future and something about wanting to start a company in 2007. In exchange for his release he offered to reveal a cure for AIDS and the location of Osama bin laden who, in 2003 was still in hiding. Andrew Carlssin was arrested but a mysterious stranger paid his $1 million bail and he hasn't really been heard from since except supposedly in a phone call made years later.

  3. I have intimate knowledge of the process, as I work with SLA/DLP printers, but I still took a great pleasure in watching this video.

  4. I wonder what it would be like if engineerguy and AvE were to collaborate
    Two very different worlds colliding that should be interesting

  5. Fantastic video Bill, I really saw the effort put into this one.

    I believe it may be your best interest to choose a more exciting thumbnail and adjust the title to infer you will be giving a break down.

    I almost skipped this one because I wasn't able to determine the content before clicking. It's a great video and deserves more attention.

    Thank you for another insightful learning experience and I wish you the best.

  6. 3d printing is fine for use in some ways but 3d printing metals can never beat forging and plastic is similar story

  7. Greetings Bill,

    I have a quick question, i know the reasoning behind the rotation of the platform, and i get why rocking and pulling are bad for the models being printed, but what if you tapped it. Or had a tiny hammer strike the center of the printing plate to remove the vacuum? Or had a small vibrating motor? Would this cause issues with resin not flowing into the cavity or would it just not be enough of a lateral force to break the bond. What if the surface of the printing plate was hydrophobic? Would it repel the resin? What if you did the printing in a vacuum, would that cause structural issues when you take it out or could you build in stress to make the structure stronger like a prince Rupert’s drop?

    Crossing my fingers to hear from you. Either way great job love the videos.

  8. Are you still alive after touching the STL Resin reservoir… I had a boss who wiped up that resin and now has chronic skin irritation and the guy in charge of cleaning the STL lab died from cancer shortly afterwards

  9. Superb. This guy needs a TV program on discovery or sci channel. What a wonderfully described process of applied science.

  10. The difference between an engineerguy video and a Vsauce video is, you won't have an existential crises after an engineerguy video.

  11. Great video. But I personally would have chosen benchy over the swan. Although I can't say that it'd be a particularly effective test print for an sla printer, since the whole printing upside down means that the arches wouldn't be a good test for drooping. That and drooping isn't really an issue with sla printers

  12. The amazing bit, is that if something is small enough, it is not affected by motion as supplied by shaking etc. Application to other 'tiny things', although other forms of energy will affect them.

  13. It's all about nano technology… Thanks Engineerguy…. been following you for years… since You were on tv….(where I found you).

  14. The silicone window has borosilicate glass on the bottom which would eliminate O2 diffusion through.

  15. If you're interested in these types of 3D printers, Applied Science just made a video about them as well. He seems to have gotten his hands on a rather large one. I was gonna mention his video on DLP, too but it seems it's already linked in the description.

  16. These videos need to be more widespread.

    I feel like things are explained so well, but it's obvious that it's still over a regular person's head. [Like…how micromirrors are made, how our understanding of physics and atoms plays a role, etc]

    If we demystified the processes for more people, even though some mechanics still look like magic, it would really go a long way. If anything, it would give a higher bar from which people then make theories as a whole, and lower the level of ignorance in the world.

  17. Hi, how about the software? I would guess the software that used by each manufacturer of the printers different. And software ease of use must be also an important part of the printer.

  18. The first episode that i watched was about aluminum cans and I don't even know why I clicked on the video in the first place but after watching it i fell in love. I have been going through and watching all of the engineer guy's videos because i absolutely love how everything is explained. I watch all sorts of videos on huge machines and how they work but I rarely retain any of the information but with these videos it always sticks. I don't know if that is because he explains it better or if its because he takes his time explaining it. I really like these videos and the community seems exceptionally nice as well which is really wierd for youtube. I subscribed and will always be looking out for new videos! Also i haven't read a good book in a while so does anyone have any suggestions?

  19. Can you , with this 3D printer , create an upside-down cup ? Can this printer create the surface without previous weight-taking layer?

  20. Excellent! Another simple to understand presentation on a complex technical product. One area of 3-D printed that perplexes me is body parts. I'm referring to aortas and heart valves produced from human cell material. The technique is similar but how does it layer build a part employing biological material?

  21. Why not use microLED screen on bottom? Also they can use micronano tech that plastic can reassemble easier and have memory.

  22. Thank you even though I had a decent crude idea of how they worked and some of the mechanics involved seeing it this way and how you broke it down really put it all into perspective. Much appreciated

  23. I find this channel absolutely fascinating. Thank you for the time you spend to accurately explain the workings of different objects that even us nitwits can understand.

  24. Hey Bill, nobody could have explained the DLP & SLA technologies as crisp and fast as you have. In just 12 minutes you have empowered me with the knowledge that many in the trade would not even have. A big thanks to you… Keep up the good work…

  25. Well at least now I know about the chemical compounds that make that resin, but I think that I would like to see some examples, maybe you could include a couple of references.

Leave a Reply

Your email address will not be published. Required fields are marked *