3D Printing the Eiffel Tower in 7 minutes

http://www.washingtonpost.com/news/speaking-of-science/wp/20 15/03/16/this-new-technology-blows-3d-printing-out-of-the-wa ter-literally/

... i like that!

<a href="http://3dprint.com/51566/carbon3d-clip-3d-printing/" target="_blank"></a>

Thanks for the post!

Although this looks new, I doubt they will be able to get a patent on it because, from what they're relieving, this is nothing new. We already use an oxygen permeable membrane to cause a liquid layer to form between the part and the oxygen permeable membrane. Also, we already control the amount of oxygen that gets to the resin by adjusting the thickness of the membrane when it is constructed. They may be doing some things that they are not revealing at this time and that may be of some novelty, but from what I can see from their current discloser there is nothing new here.

What's more is, they are not going to be able to move the part continuously as it is forms for every type of part geometry. They can for things like the Bucky ball and the Eiffel Tower, but not part geometry that has a large cross sections. The reason is, the liquid resin needs to flow into the exposure area, yet, due to hydraulic limitations cannot flow in at an accelerated rate that can keep up with a continuous Z movement. Part geometry that has a small cross section will work, but not large cross sections. This limitation can even be seen in their timelaps video of the Eiffel Tower, you can see the Z movement slow down for the areas that have a higher cross section.

One last thing I noticed is, they are not printing in a continuous analog process, they are still printing in layers and those layers can be seen in the electronmicrograph. There is no way that I can think of to do it completely analog using the current technology we have available.

One thing that they ARE doing is they are printing faster, but everyone with this type of printer, including myself, are going to make adjustments so that we can print faster too. I didn't know that could be done, and that is commendable, I highly appreciate their work here, but if they want to get an enforcible patent, they may need to get more creative in the way they are describing what they have discovered.

So you already own this technology?

No, it's public domain.

Really? All previous SLA that use a projection from below require a "peel" step, moving the part up and then down, or the tray side to side or rotating... this one seems to require no peeling (no contact with the bottom of the resin container), but oxigenation control instead as well as precise up motion to take into account the geometry at every given moment.

If it was public domain and in use, why other SLA techs are taking hours instead of minutes for similar models?

Yes, everything you are saying is true for most of these types of top down SLA machines, although some do not use and oxygen permeable membrane and use something like Teflon as the nonstick material. However, no one knew one could get away with moving in the Z axis based on the cross sectional area. Before, weather the cross sectional area was large or small we ( we being us humans ) used the same process and time for every layer. I guarantee you they cannot print a solid cube in 7 minutes with the same Z height as the Eiffel Tower print. The reason they cannot has to do with the hydraulic flow limitations of the resin.

Let me say it better. Let's say I get a conventional top down SLA machine that uses an oxygen permeable membrane to keep a liquid layer of resin between the membrane and the forming part. I print a solid cube that is 50 mm on a side and lets say that takes seven hours, and now, I print a cylinder that is 2 mm by 50 mm and it still takes 7 hours. However, on their machine, to print the same cylinder takes 4 minutes, lets say. Yet the 50 mm cube on their machine would take 7 hours. This is because their machine is speeding up the cycle time when the cross section of the part being formed is small, whereas everyone else has been spending the same time per layer regardless of the crossectional area of the part being formed. See?

Recall that a bottom up SLA machine has no sticking issues whatsoever, yet their control software could be used with one of these types of SLA machines and could get the same speed results.

They may be able to get a patent on this if someone hasn't mentioned trying that somewhere in public in the past, because processes can be patented and it does appear to be novel, but everything else they are claiming hardware wise is already known and in use and has been in use for a long time.

Oh, I just reread your statement and thought of something better to say. The step you are referring to as peeling, think of that as getting new liquid resin in there so it can be hardened by the light source. It is never "stuck" to the surface in a machine that uses an oxygen permeable layer. Also too, a bottom up machine has the same problem even though there is nothing to stick to. The problem is, moving new material into the area so that it can be hardened. If one had a way to magically instantly make that new unhardened resin material appear in the exposure area, THEN! Then you'd have an amazing invention because we could print at tremendous speed regardless of the cross sectional of the print being formed.

Getting new liquid in place, or trully separating the solidified resin from the "glass"? If the issues was just flow, Teflon would not be a big issue, nor would be pollution that requires the glass to be replaced periodically (or picking different print locations every run so the replacement time is bigger, because the pollution is spread all over it). That seems to be what happens with common avaliable models.

Their magic seems to be that they keep a "thick" buffer so resin is always flowing in, and the polymerization takes place away from the "glass". Immersive systems could use the system using oxygen as cure inhibitor... but they don't seem to, do they? Maybe because that is the new step, curing at depth with special control.

You should ask them to demo the speed of a full block, a worst and irreal case, but limits are always nice to see. If they still make it at a fraction of the other printers, they truly got something going.

I do see what you're saying about a "thick" layer. If the layer was thick enough then that would render my view on it erroneous. A thick enough layer would allow more room for the resin to flow in and therefore it would be a great improvement.

I'll ask them if they'd like to print a solid block for us.

http://www.ted.com/talks/joe_desimone_what_if_3d_printing_wa s_25x_faster has a comparison of "classic from below" SLA (with peeling, or delaminating) vs their new "just a bit beyond the window" SLA. The trick must be in making sure the curing is always in contact with the previous solid parts, but not in contact with the window (and the use of special Teflon is proably a safety measure, otherwise they could use anything that lets O2 pass).

Lets' see if they have time to answer, I'll bet they're pretty busy right now though.

Here's someone doing something similar with a bottom up SLA printer.

I saw the TED Talk you posted long after you posted it. Very interesting! After watching I went searching for the first patent covering the oxygen permeable membrane, but got to doing something else and never finished looking. I'm sure there was patent covering this though. If I get in the mood again I'll spend more time looking for the patent. I would love to do my own experiments in this area, but my machine is dismantled right now as I'm doing some upgrades to it.

Oh and, no they didn't write me back.

Edit: Had the wrong URL.