More clever supports idea


#21

This Vat has been used for only half a liter of print and I am already having print failures due to clouding.
To me, it is evident that the vertical supports concentrate the wear in “failure triggering spots”.
See how they diffract the sunlight ? Well they do exactly the same with the laser’s light :

I’m not sure if the new Vat will also be affected by this, but if that’s the case, smarter inclined supports would make it’s life span not just 20, but maybe 30x or 40x better than baseline !


#22

It’s a question of how much time FL wants to invest in coming up with a more efficient structural support design algorithm.
Locating the object from the base and then designing a vertical support to that object with diagonal braces across the verticals is complex but a lot easier than a spaceframe.
So what is that cost vs less wear on the tank surface.


#23

I think that the best way of solving this is to use a live physics engine like Kangaroo for Grasshopper.

There’s no use in re-inventing the wheel : the tools are out there already.


#24

Go on - show us


#25

Hi Bill,

I’m investigating these plug-ins :
http://intralattice.com/case_studies/

Cheers,


#26

I haven’t had any issues with the supports buckling. I think the supports do more the during the peeling process where the direction of force is going the other way rather than the compression of them where you would be right as they would buckle.
So far what FL has seems to work fine here and cleanup isn’t too bad.


#27

The main issue is the vertical struts which focus the wear on spots (see pictures above)


#28

First test : VIDEO


#29

I can see how that is a problem.
So far I really haven’t had that problem with supports marking my tank but I use OpenFL and clear 02 resin. Also my prints are relatively short for the most part but occasionally have tall prints.
I like what you did grasshopper, I have to figure how to do that in FormZ when I get some time.


#30

Isn’t this going to be too flexible for the forces it will see during the peeling process? There are several compression and tension forces and nothing is stiffer than a straight pillar when you pull it.


#31

Hi Fantasy 2,

This is just me fooling around with a new tool right now to show the process of :
-Lattice generation
-Form-finding
-Meshing the wireframe struts into a watertight mesh
It needs much more thought and experimenting, in particular some FEM analysis to check for the best material usage/stiffness ratio with various lattice designs and form-finding options.

Peeling forces are mostly shear forces since the vat slides purely horizontaly, contrary to the F1.
Therefor, a support structure with high bending inertia is much more efficient than slender vertical supports with erratic bracings.


#32

The peel force is a combination of vertical and lateral. There is actually a pretty good vertical pull before the slide (likely resin dependent).


#33

Hi Fred,

I did not observe any vertical displacement of the build platform before the vat slides, therefor I didn’ think there was any vertical pull involved, but maybe it is too small to be noticed ?

If there is a vertical displacement, does it occur during or before the slide of the vat ?


#34

It might just be me but I think there is a bigger lift during the base build on the plate then a slight lift for the rest of the print. I’ts not much but it’s there. Lift then slide. Might be lift then (slide and lift) it’s hard to tell.


#35

Yep, it lifts first(probably to create a bit of tension), then slides. This applies to at least the first 5mm of base layers.


#36

On the F1+ there is not lateral movement (old school).
I was wondering if your getting burn marks from the supports then maybe the layers are slightly over exposed.
I don’t remember seeing that with any of my tanks but I also use Clear 02 which I think FL has really optimized the exposure for.

Does OpenFL work on the F2 yet? If so they try changing the exposure there, slightly less should really help.


#37

Just for the fun’z


#38

I love me some grasshopper, but you could just use a free pre-process program like autodesk’s meshmixer to do the same thing with more control. I use mesh mixer for FDM printing. Sure it may not be a pretty as a minimal surface grasshopper structure like that, but if your an average Joe you don’t have to spend $1,000 for rhino, learn rhino/grasshopper just to do some pre-processing.


#39

Hi GiantSquid,

There’s something fishy when you can read “Autodesk” and “free” in the same sentence.
Take Fusion 360 : it was free just to attract the little fishies, then it was “free under certain conditions”, and then if was free only for students.
All the makers and small companies who had invested time learning it were suddenly forced into a monthly subscription.
Rhino is 1000$, sure, but that’s a one time purchase.

My attempt at form-finding is a bit primitive though. One would need to consider :
-Normals to the printed object’s surface
-Collisions with the object or other objects in the batch
-Shear forces during peeling process
-Etc…

…or maybe there’s no need to go so far, but I’m sure of one thing : the present Formlabs support structures are not structurally optimised, which means we pay $$$ for un-needed resin in the supports, or that with the same amount of resin, supports could be made much stiffer.


#40

Seeing those pictures of the vats held up toward the sky, I couldn’t help thinking: new side project for used resin vats, print star constellation maps!