A proposal for
Cocoon Supports
for 3D FDM printing.
The slicers as of early 2023 are remarkable applications that can transform a 3d solid model to G-code with minimal input.
The goal of the slicer is to allow any user to create parts without having to develop complex procedures and tricks to produce the part without failure and at minimal cost.
I have had great luck with the latest printer designs and the open-source slicers that are available.
The problem:
The one style of part that “Standard” and “Tree” supports do not seem to work well on are
TALL and THIN parts with a small base.
The logic of the slicer does not “see” any significant overhangs that need support, and when support is manually set up, the resultant converging towers do not really solve the problem.
After I had a couple of TALL and THIN parts fall over during printing, I realized that printing a large brim around the base was not enough to support the part because of the leverage exerted as the part gets taller.
My research:
I evaluated the possible causes of my part falling over. My part was not jiggling about during printing, it was possible that just having a small amount of hot filament drool being dragged across the top of a part while the platen moves in the XYZ directions can topple the printed part, but the particular part that I was trying to print had a unique issue of having the center of gravity creep far away from the base as the height increases. It would take many points of support to follow the center of gravity as it travels around.
I had used a technique recently to support a straight shaft that I thought could be modified to be scalable to work with any tall part, including these.
I came up with a solution that allows me to print this type of part with minimal sacrificial support material, no flushing material wasted because of filament swaps, the process if foolproof enough that the parts can be printed with lights out ( I only have to go to the machine to remove my finished part ), it just works. It is also scalable (to be automated) for use in a slicer.
A comparable solution?:
I searched online for a program to automatically create the type of support I was using, but I could find no mention of this technique. I realized that I may have stumbled upon a NOVEL METHOD of SUPPORT that could be used in the cases where “Standard” or “Tree” supports would not work.
Since I created it, I think that I should get the privilege of naming it:
I am calling it “Cocoon Support”.
The basic concept is to create a minimal shell completely around the part with an air gap on all surfaces. Since the support never touches the final part, there are no tags or support marks anywhere, the part comes out with a pristine surface. The Cocoon also controls the temperature of the part, so the upper layers look the same as the lower layers and the part does not lean or collapse when printing a thin section underneath a heavier structure.
The shell is created with a single outside wall, a single bottom surface, no top surface, and a maximum infill of 1% using a rectilinear pattern.
I have found that using a Boolean gap of 0.4mm in all directions will always work, but in some cases I have gone down to 0.2mm gap to support very fine features.
Using 0.4mm allows easier removal, in almost all cases the part literally falls out of the support when you remove it from the plate. If your part has overhangs and your printer is tuned up properly, you could use 0.3mm gap or less. If your printer or filament is sloppy at all, then stick with 0.4mm.
Here are some examples:
I usually make mechanical parts, but they aren’t very dramatic as test pieces.
So I decided to make a Quadruple Portal Toy by Ada Cohen on printables.com .
This is what the finished model is supposed to look like.
It has four screws of various helixes with complex profiles.
Here are the four screws shown without the housings or nuts.
THIS is the screw that I chose for this example.
It is relatively heavy, has a small base with an undercut for a guide pin, and is a helix that goes out and around and ends up almost directly above the base.
This part can be printed without using supports if the printer is slowed down.
It can also be printed using paint-on tree supports.
So this part is printable, but it could be difficult.
Here you can see that this part is a true helix.
The 2nd picture shows the base with it’s undercut keyway.
I use a conventional support for this undercut so that it is shaped correctly.
[Cue the suspenseful music]
This is how my FIRST swing at this part turned out (without any supports).
The spaghetti recognition feature stopped the printer…
[ Pause for dramatic effect...]
[ Cue the dramatic "everything is going to be OK" theme ]
Here is my next attempt using Cocoon support.
The 245mm part height just barely fit in my printer.
This is with the print plate removed and placed on top of the printer.
The finished part sticks out above the cocoon by 15mm.
Plate and part tilted towards the camera.
You can see the finished part at the 9 o’clock position spiraling clockwise down.
The inside and outside walls of the cocoon have a thickness of 1 layer.
A single layer on the bottom,
0 layers on the top.
The fill in this photo is 2% with a rectilinear pattern.
The fill percentage can be reduced further and will still function fine. I have found that I now use 1% for everything.
It is possible that there may be an occasional error when a slicer is calculating a percentage less than 1% due to rounding errors, but I hope I am wrong about that.
When I tried to remove the print, the part slid out of the support.
The finished part glides easily through the mating part.
No snags or rough spots.
This view shows the bottom of the base with a pocket that was supported by a standard support.
There is a single layer of support interface material between the surfaces.
This is the base with the support material popped out.
This is a video showing the procedure to remove the support material from the finished model.
Arguments for and against Cocoon Supports
Disadvantages:
It has limited applications. So far I have only been using it for tall and thin parts that are difficult to make with traditional or tree supports. This is not for cosmetic parts with complicated undercuts.
It uses more filament than using no support (of course). With a tall-thin part, the Cocoon itself can actually use more filament than the finished part.
It can be slower. You are printing more, BUT: In many cases the way to deal with tall parts is to slow the printing speed down. With a Cocoon, you can return to full speed. ( I still uncheck the box for "reduce retraction on infeed" because I don't want to push my luck, but you can experiment.)
More waste. It bothers me to throw away what looks like a perfect print of the Cocoon.
BUT: It bothers me MORE to throw away a failed print because it needed support.
If the part has complex surfaces the slicing calculations can take a while. Mating complex surfaces actually have to be calculated twice because the surfaces are duplicated in the cocoon.
There is no programming to create Cocoon Supports automatically... YET.
Advantages:
It is useful for tall thin parts that need to be geometrically correct without blemishes.
50% of the time a cocoon support will use less material than other types of support ( you can guess about the other 50%).
A cocoon support will not leave snags, pits or other remnants of other types of support because the cocoon does not touch the finished part.
It is an easy concept to understand, so it will be easy to modify to the user's needs.
Although it has limited uses (so far), it seems to work exceptionally well for the subset of parts that are tall and thin.
It just WORKS! That is a good enough reason, being able to simplify the process of making parts.
If you would like to print some samples to see how they work,
download these files from printables.com
What this page is about.
I am presenting this proposal in the hopes that a few people will try it out and evaluate whether this is worth writing a routine for embedding into a slicer.
It is far simpler to program than tree supports (but not as cool to look at).
There should be options for the size of the center hole, side zippers, and clearances needed for undercuts or steps.
I have used this procedure for a number of parts and have found it to be extremely useful, but I am very comfortable using any solid modeling program so they are easy for me to create. Many people using 3D printers do not have experience with 3D modelers so I think that this type of support should be created within the slicers.
Do you agree?
Please email me at cocoonsupports@warrenw.com
OR add the link to this site to whatever forum you are on and send me an email with a link to the forum. Thank you. (ww)
2023-05-22, WarrenW