Anyone who has done any amount of 3D printing with SLA printers is probably well aware of the peeling step with each layer. This involves the newly printed layer being pulled away from the FEP film that is attached to the bottom of the resin vat. Due to the forces involved, the retraction speed of the build plate on the Z-axis has to be carefully tuned to not have something terrible happen, like the object being pulled off the build plate. Ultimately this is what limits SLA print speed, yet [Jan Mrázek] postulates that replacing the FEP with an oxygen-rich layer can help here.
The principle is relatively simple: the presence of oxygen inhibits the curing of resin, which is why for fast curing of resin parts you want to do so in a low oxygen environment, such as when submerged in water. Commercial printers by Carbon use a patented method called “continuous liquid interface production” (CLIP), with resin printers by other companies using a variety of other (also patented) methods that reduce or remove the need for peeling. Theoretically by using an oxygen-permeating layer instead of the FEP film, even a consumer grade SLA printer can skip the peeling step.
The initial attempt by [Jan] to create an oxygen-permeating silicone film to replace the FEP film worked great for about 10 layers, until it seems the oxygen available to the resin ran out and the peeling force became too much. Next attempts involved trying to create an oxygen replenishment mechanism, but unfortunately without much success so far.
Regardless of these setbacks, it’s an interesting research direction that could make cheap consumer-level SLA printers that much more efficient.
(Thumbnail image: the silicone sheet prior to attaching. Heading image: the silicone sheet attached to a resin vat. Both images by [Jan Mrázek])
Carbon doesn’t use an lcd, they use a projector. I don’t think this would be possible with an lcd since the lcd screen needs to be right up against the membrane to reduce light bleed from light that isn’t parallel. If the light from the legs was perfectly parallel then maybe it would be possible.
Source: I’m an engineer at a company that makes lcd printers.
I suggest reading the original blog post where everything is explained. TLDR: The whole idea is to make oxygen inhibition via system that is 200µm thick so it can be used with LCD. Using it with LCD was the goal (so you can use it on cheap printers), not replicating CLIP at its finest.
dumb question – is an LCD printer actually cheaper compared to DLP, given the horrible attenuation of a LCD?
They might not be. It depends – if you want 22µm voxels and print area of 220×120mm, then I LCD is much, much cheaper. Even when you consume tenths of them.
LCD is, however, also interesting from another point – unlike DLP it doesn’t require optics and thus, it doesn’t suffer from image distortion. You get the amazing precision of the n-million USD machine that produced the LCD nearly for free and without any work on your side.
What would happen if you dissolve oxygen in the resin and/or put the vat in an oxygen enriched environment? Maybe ramping up the pressure of a build chamber by attaching it to an oxygen tube.
Needs an ignition source. Good start though.
Then none of the resin would cure properly. You only want a thin layer at the bottom of the vat to be exposed to oxygen so the layer immediately above it cures, leaving a film of uncured resin as a kind of release agent under it
Oxygen permittivity of silicone is not as high as the permittivity of the amorphous fluorocarbon used in CLIP. You can make silicones work if you enrich the environment beneath the membrane…but that can be dangerous to do ‘ar home’. The atmospheric quenching with CLIP, using the amorphous fluorocarbon film, is only about 2 microns. DeSimone had to use enriched oxygen environments to get thicker dead layers and show resin flow under the part. In normal atmosphere, CLIP doesn’t have very strong chemical adhesion to the vat bottom, but still has slow retraction speeds based upon resin replenishment under the part (viscosity limitations as thick fluid runs through a capillary gap). Also, many resins are not CLIP compatible (i.e., they don’t care about oxygen and cure even when its present).
Likely, you had a temporary effect of the silicone de-wetting the resin/part. Eventually this wears out a few layers in when you’ve sufficiently roughened the film surface with resin defects. Looking at Nexa, you can see they basically used a silicone film as a ‘sponge’ filled with silicone oil. So it’s always ‘leaking’ lubricant. Eventually runs out, but not for a long time. Much longer than 10 layers. 😉
I believe I was able to achieve pretty long-lasting effect (I can’t prove it is infinitely lasting), but the problem was elsewhere. The inhibition worked, just it wasn’t practical and didn’t fulfill the goal of lowering peel forces. All details are in the blog post.
From memory the Right Answer(TM) to this problem is oxygen-permeable teflon, but while it’s theoretically available as a product it’s never been practically purchasable.
It also feels like there’s some optics needed to get an image plane off the base of the vat.
The material is just ridiculously expensive (even for Carbon) because of the manufacturing process. You can cast your own sheets or but then….at hundreds of USD per gram. So not really good for hacking.
You are also correct about light – if you have a ‘thick deadlayer’ that’s like 200um above the vat bottom you need to focus the light there. LCD stops being a great option in this scenario because you optically fuzz a bit. That’s why DLP is used. For Nexa silicone process, LCD is fine because your focal plane is very close to resin…
Not PTFE in particular, but fluoropolymers more broadly grouped as “highly oxygen-permeable ionomers” seem to be the keyword when looking for such materials in the field of fuel cell research (e.g. 10.1149/MA2020-02342219mtgabs). Does that sound right?
“LCD Resin Printer stays for what ?
==(Thumbnail image: the silicone sheet prior to attaching. Heading image: the silicone sheet attached to a resin vat. Both images by [Jan Mrázek])
single image only
I literally just ordered a few sheets of nFEP and some PTFE lubricant to combat this problem. I have had a real problem with failed prints as of late. I have also heard that RainX will do the trick to prevent sticking.
Are you sure that you have problem with sticking? Sticking creates models that are torn other than by the layer direction. If you have you model completely on FEP, the cause is usually delamination (layers not sticking together) caused by too thick layer (the printer has hard time squeezing the resin into appropriately thick layer). For more detail, look up blog post with name “The Winter is Coming – B̶r̶a̶c̶e̶ ̶Y̶o̶u̶r̶s̶e̶l̶f̶ –Prepare Your Resin Printer And Prevent Print Failures” (HaD doesn’t allow links in comments)
When I first got my resin printer all prints failed till I used the RainX trick and so far I have far fewer failures (most failures now look like they are down to my fault for not properly supporting the part or possibly the room/resin being too cold).
What about a polymer like Nafion or HOPI that’s being used in fuel cells?
One of the ways I have improved the peeling issue on my system was following a suggestion of using PTFE lubricant on the resin tank. Has worked wonders.
Great hack but not a new idea. One of the first home resin printers, the B9 creator, solves this with an oxygen absorbing silicone which is solid sideways out of the resin into free air to breathe for a second before sliding back under the build plate for the next layer exposure. I have one and it works very well.
(No criticism of the work here, just sharing another person with a similar solution that made it through into a product and a moderately successful cottage industry splitting printers to for the jewellery industry)
I’ve often wondered whether it would be possible to turn the while thing on its side and have a circularly rotating silicone plate on a horizontal axis so half immersed for exposure and half above the resin level replenishing it’s oxygen.
I’ve seen a much simpler method with a circular FEP sheet being rotated here awhile ago
https://hackaday.com/2021/09/16/continuous-resin-printer-shows-the-speed/
https://www.reddit.com/r/resinprinting/comments/pcbvbv/diy_continuous_resin_printer/
Was thinking of modifying a standard resin printer to have a roll of FEP film go across the screen instead.