Brumbaer / 3D Printing / FEP Vat
 

June 2019.

What's this ?

I do not use my 3D printers continuously, but project wise. So there are long stretches when I do not use them at all followed by shorter stretches when I use them 24/7.

When I started my latest project, I had only one vat being freshly coated and so I had to recoat some, to be prepared for the inevitable swap. I will recoat at once all that are not freshly coated or not in use and so I had 4 vats to recoat.

Recoating is simple, but still I try to postpone it as much as possible. It's more the thought about what you have to do than doing it. So whenever a recoat is due I think about a method to avoid it.

The printing process on the B9 as for all DLP based printers is simple and well understood.

Theory

The model is cut into horizontal layers. They are printed one on top of the other until the model is complete.

There is a vat with a transparent floor, which is filled with uv-sensible resin. Just above the floor hovers the build table. An image of the current layer is projected - in our case using light with a high portion of uv-light  - through the floor of the vat. The light will cure the resin where it is hit. The depth of the cured resin depends on the amount of light, which is determined by time and intensity. If all is well the depth will be deep enough to cure the resin up to - for the first layer the build plate - for successive layers the layer printed just before.

After a layer is printed the build table is raised and and the image of the next layer projected and cured. 

That's the principle, the process is in fact a bit more elaborate.

When the resin is cured it will cling to whatever it touches. On the first layer that would be the build plate and of cause the transparent floor. That bond will be so strong that the build table may not be able to rise or it may stick to the floor and come off the build table.

So a coat of PDMS is put on top of the floor and the resin is now trapped between PDMS and build table. After curing the resin will stick to the PDMS and the build table. But resin bonds much less to PDMS than to the build table, so the chances are good that when raising the build table the cured resin sticks to the table and is released from the PDMS. PDMS has a weaker holding force in horizontal than vertical direction. To make use of this and improve the chances of resin sticking to the build table and being released from the PDMS, the PDMS - in fact the vat - will be moved horizontally until the cured resin is off the PDMS. To make this possible, the vat is much wider than the build table and the PDMS covers only the half of the floor that will be under the build table while projecting an image. After projecting the image the vat is moved far enough to move the cured resin over the floor where there is no PDMS, dangling in thin air. Now the build table is raised higher than needed for the next layer. The vat is moved back into its original position, the build table is lowered to the next layer height and the next layer is cured. And so on.

The vat is mounted on a metal plate called the shutter, because it has a hole under the part of the vat where the image shall be projected. This will shut out stray light from the projector while the vat is moving and prevents resin to be cured while the vat moves.

Problem

When the resin is released from the PDMS it will rip minuscule pieces from the PDMS. This will show as fogging - the uneven surface of the damaged PDMS will break the light and the transparency is diminished. This will happen where cured resin is ripped off. So if you print a model multiple times at the same spot, the fogging will be noticable faster than when you print it at different spots - spreading the damage.

Idea

The idea is to replace the PDMS with something else that is not prone to fogging.

FEP (Fluorethylen-Propylen) is a plastic which is available as transparent foil and has similar non stick properties as PDMS. It has a much harder surface and does not fog. A year ago I tried replacing the PDMS with FEP foil and it really didn't work for me. 

But time passed and nowadays FEP foil replaced PDMS as standard method on all current dlp printers. If everybody else gets it to work, why shouldn't I ? So I decided to try again and design a completely new vat instead of converting an existing one.

The force needed to rip of the resin from FEP horizontally or vertically is identical, so the horizontal movement of the vat is not necessary any more. That means that the vat can be smaller than the original. Still the printer software expects the shutter to work, so the new vat will be mounted on the side rails and not on the shutter. The shutter will be able to move freely, but the vat will not move with it.

The new vat is still a tub with a transparent bottom. The bottom will be covered by FEP foil. The resin will be cured between foil and build table and after exposure the build table will rise and rise and rise ....

The FEP has some elasticity and will be drawn upwards where the cured resin clings to it. At some point the fep will peel off from the resin starting at the edges. That's the idea.

The FEP foil will be mounted in a metal frame, which will be fastened in the tub. That's it. Quite simple.

Go about it

The foil has to be tightend as much as possible, so that it will move the least possible way, when resin clinging to it is pulled away. 

The mounting must be in a way, so that the vat is water tight.

And to make the printing process faster by skipping the horizontal movement step, the printer software has to be adapted - it will work with the original software, it will just not be as fast.

With each iteration of the vat there were changes and now the vat consists of a baseplate which has a raised frame to mark the window and stretch the FEP.

The tub is a separate part which will be mounted on the baseplate and when doing so the foil will be tensioned by stretching it over the frame. 

After some tests I filled out the frame with a sheet of acrylic glass. When left open it was too difficult to calibrate the build table.

The tub is basically a 3D printed watertight tube, with stainless steel frames at the top and bottom which allow the screws to be fastened strongly enough to withstand the tension of the FEP foil.

The FEP foil is readily available as a replacement parts for different printers, it only has to be cut to the correct size. I use a foil with a thickness of about 120µm. It is mounted between the bottom steel frame and the tube. The FEP working as a sealant.

Experience shows that there are differences in the quality of the FEP. You might have to try different types. At the moment I use one from Wisamic sized 140x200mm. Thickness specified as being between 0,15 and 0,2 mm.  

This assembly is mounted on the base plate. And the completed assembly is mounted in the printer on the side rails. It may be the best to look at the images :)

I made more than a hundred prints without replacing the foil.

Print times are a bit lower than with the original vat. They can be drastically lower with the right kind of resin, model and tension of the foil, but I settled for something more reliable.

Hardware

The materials used are stainless steel, PLA for the 3D Printer, acrylic glass for the window and screws and nuts type A4.

Tub - sub assembly

The most obvious  part is the tub. It's walls are 3D printed from PLA.

The Dimensions are just big enough to hold the build table. You could make the tub wider, but deeper would be difficult because of available space. The height is 18mm. The height is derived from the thickness of metal parts, nuts and available screw lengths.

The tub will be fastened directly on the FEP. To make this connection water tight, you will need some pressure. I didn't expect the PLA to stand up to it. For that reason the top of the tub and the FEP will be supported by two stainless steel frames. They will distribute the pressure and the tub serves mainly as a water tight spacer.

 

The frames are made from 2mm stainless steel. I had the lasercut by an online service.

There is a sandwich made from frame, tub, FEP and frame. The whole thing is held together with 16 M3-30Start screws. Again made from stainless steel (A4). Why 16 screws you may ask ? Because it works. The screws will stand out by about 6mm - not counting the nut.

When the parts are assembled I set this aside and fill it with water to see whether it's really water tight. If you haven't any confidence in the whole thing, let it rest for a night and check in the morning. 

Base plate - sub assembly

The base plate is made from 2mm stainless steel. On the front there are mounting holes, but on the back are slits. This allows to insert or remove the vat without removing the screws at the back. You just loosen them and slide the vat in or out.

 

The wider of the two frames will be glued on the base lining the opening. The slimmer one will be glued o top of the wider one. Both are 5 mm high. They will build a pedestal that will hold the window and will press against the FEP and tauten it. The small frame is missing in the picture above. I should have had a spare one, but ... .

The acrylic glass window is inserted without glueing it. When you will replace the FEP be prepared to change the window as well. Being acrylic glas it will suffer from the resin, ipa, heat and whatever. The window is 5mm thick. The holes shall let air between FEP and window and shall lessen the FEP being stuck against the window. I haven't tried it without them, so they might be useless. I think about having a real glass window without holes made. But up to now, I'm still content with the acrylic version and it is no hassle for me to laser cut a new window whenever I have to replace the FEP.

Putting it all together

You will realise that the window is not in the middle of the tub. This will cause more strain at one side. That's how it is, you will have to work with it. Once you stuck the screws through the holes in the base, you will fasten the nuts. It is a good idea to put on all nuts just as far as you can turn them without applying any force. Some screws will have more of their length stick out than others.

Now fasten each screw by 2 or 3 turns starting with the "least showing" middle one. Than the one to the left, than the one to the right. Follow the pattern till all had their turns. Than fasten the screws following the same pattern.

Now it's time for an other water tightness test.

The pudding

 

 That's the vat installed. You can see spacers under the base. Whether those are needed or not depends on the length of the screws used.

The Resin is Spot-A with some dye.

You will need some large overlift about 3mm should see you ok.

You can set the shutter movement as high as you want, as the vat doesn't move anymore.

Don't fill the vat to the top otherwise the build table drive some resin over the edge.

You might want to stir the resin after a period of rest.

The FEP is quite forgiving. But it doesn't like pointy things. Like remains of the supports from failed prints driven into it. The good news so, is that it is damaged - as in not even anymore - but I had never one leak.

There is a FEP variant or alternative from Epax, which after first tests seems to be absolutely fantastic. With Spot-A, a "fresh" lamp and Epax I get up to 40mm/h at 30µm. That's a far cry from the usual print rates, which were less than 20 mm/h.

So where to go from here ?

Don't know, there are some minor annoyances so.

The most prominent is that the vat is so low. This means that you will have to refill more frequently than you're used to. The little height makes it easy to get resin spilled when the build table lowers. But I'm loathe to make it higher, because right now the inbus key to fasten the build table screw rests nicely on the vats top edge, which eases the build table process.

There is no nozzle or such to make emptying the vat more controlled.

I only have tested z heights of 25 and 30µm, but can't see why it should be different for thicker layers.