can someone with a stereolithography printer help me test this structure?

[helge]

This "foil" structure is the central piece of one of my latest hobby projects but I have no way to manufacture that structure. I've considered making square pillars which could be produced with a diy wafer saw for which at least the spindle motor exists... see where this is going, recursive problems
The aspect ratio is key here, if the structure turns out to be too high res (0.2mm pillars, each 1.0mm high, 0.07mm gaps in between them) the model can be tweaked.

I've attached the initial test .stl which covers an area of 20x20 mm²

OpenSCAD code:

Code: [Select]

module hexagonal_prism(r,h){
polyhedron(
       points=[
        [r,0,0],
        [ 0.5*r, 0.5*sqrt(3)*r,0],
        [-0.5*r, 0.5*sqrt(3)*r,0],
        [-r,0,0],
        [-0.5*r,-0.5*sqrt(3)*r,0],
        [ 0.5*r,-0.5*sqrt(3)*r,0],       
        [r,0,h],
        [ 0.5*r, 0.5*sqrt(3)*r,h],
        [-0.5*r, 0.5*sqrt(3)*r,h],
        [-r,0,h],
        [-0.5*r,-0.5*sqrt(3)*r,h],
        [ 0.5*r,-0.5*sqrt(3)*r,h]],
       faces=[[5,4,3,2,1,0],[6,7,8,9,10,11],
        [0,1,7,6],[1,2,8,7],[2,3,9,8],[3,4,10,9],
        [4,5,11,10],[5,0,6,11]]
);
}


pillar_radius=0.1;
gap = 0.07;
pitch = 3*pillar_radius+2*gap/cos(PI/3);
pillar_height = 1;
base_height = 0.5;
union(){
    for(y=[-30:1:30]){
    for(x=[-18:1:18]){
        translate([x*pitch,(2*y  )*pitch/3*sqrt(3)/2,0]){
            hexagonal_prism(pillar_radius,pillar_height);
            //hexagonal_prism(pitch*0.25*sqrt(2),-base_height);
        };
        translate([(x+0.5)*pitch,(2*y+1)*pitch/3*sqrt(3)/2,0]){
            hexagonal_prism(pillar_radius,pillar_height);
            //hexagonal_prism(pitch*0.25*sqrt(2),-base_height);
        };
    }
    }
    translate([-10,-10,-base_height]) cube([20,20,base_height]);
}



[Artlav]

Hm, it looks too fine for mine, but i might try it when i get home.
What is the purpose of this structure?
Should it flex? Photopolymers aren't very flexible.

[helge]

No major mechanical forces - shouldn't have called it a foil then  it's a surface tile that gets glued onto a flat ceramic plate. Actually it would be best to just start out printing it directly onto that surface but the substrate is a different matter altogether.

[tggzzz]

Upload your .stl file to http://dirtypcbs.com/store/print3d and see how much it will cost. I suspect £10 delivered.

[helge]

@tggzzz nice hint, I'll give them a try too. The next step may require a custom substrate but that'll be something for 2017 ;-)

[helge]

Is SLA still that rare?

[helge]

Alright, well here's what it's all about:
https://hackaday.io/project/18987-transient-thermochromic-contrast-camera

[CatalinaWOW]

If you are going to stereolith this, why make the pillars constant thickness?  You typically want to minimize conduction in a bolometer camera, which this is, with a liquid crystal readout mechanism.  So make the pillars as thin as the process and mechanical requirements allow up to the liquid crystal supporting platform.

One of the benefits of stereolith is that it enables shapes that are difficult or impossible with other fabrication techniques.

[helge]

I guess what I'm asking (150-250µm pillars and the 70µm in particular) is just on the edge of being feasible with lower end (that is to say within the reach of hobbyists) SLA printers. The material itself does not impose such limitations but some printing services will not manufacture such feature sizes and spect ratios of 5-10.
And then there's the trade-off between settling time, SNR and dynamic range. I'm shooting for 0.5-1K resolution, a whole cycle should not exceed 10 seconds. The time constant itself in air is notably larger than that so some heat capacity and finite thermal resistance is actually a good thing.
I should do a simulation with FEMM before a larger prototype is built.

[Artlav]

Sorry for not getting back to you.
Mine can only do about 0.2mm, and it's a top-down, so there will be issues with liquid settling - there would be a semi-spherical bulge over the whole area.
I might try to move the projector closer and refocus it to get higher resolution, but still best i can do is to make rectangular pillars, not hex ones.
Another problem is, i don't have any thermochromic materials around, so even if i print it, there won't be much i can do with it.

[helge]

I'll see if I can do some simulations over the weekend to figure out where this is going with respect to aspect ratio and pillar size. Square pillars should work too.

 
If the distortion doesn't get out of control I can imagine that a contact printing tool with 3/4 of the pillars removed might work so one can transfer the paint in four passes.
Depending on the solvents used photoresist might work.
If all else fails you'll see me paint the little mesas one by one under the microscope ;-)

[Artlav]

Speaking of etching, can the pillars be conductive?
I.e. made out of copper?

I've recently seen a PCB with 0.7mm (20oz) of copper on it:
https://www.reddit.com/r/electronics/comments/5i9wxd/i_just_received_my_20_ounce_pcb_soldering_is/

Apparently these are made by plasma etching process, which can produce high aspect ratio patterns.

That is, you should try to look at capabilities of PCB places and machine shops, perhaps, rather than stopping at STL.

[helge]

The things they offer these days.. but that 20oz PCB got me thinking... one might as well have a copper mold produced that way and cast some silicone rubber. Ben Krasnow over at applied science did something similar on a much smaller scale (Gecko tape, see ). I guess these PCBs are hellishly expensive.

As for material selection: ideally you'd have the equivalent of an RC T filter with voltage (heat) sources on both sides and while copper has a respectable volumetric heat capacity
( 3.45 J/(cm³K) https://en.wikipedia.org/wiki/Heat_capacity#Table_of_specific_heat_capacities ) its superb thermal conductivity would suck the heat right out of the pixel. Ideally you'd like to have a copper cube polished on five sides floating in vacuum and a way to retract it onto a reservoir surface for resetting while the thermalization would be governed by radiative heat transfer only.
So the trade-off would be to have pillars that have both usable heat capacity and conductivity. The conductive heat transfer resistance of the bridge to the reservoir needs to be on the order of the radiative heat transfer resistance ( https://neutrium.net/heat_transfer/radiative-heat-transfer-coefficient ).
An interesting implication is that the pillar mantles are also radiatively coupled. So maybe there really needs to be a low emissivity coating on them (thin Al coating will work great, epsilon ~ 0.04, Chromium coating 0.06). Physical vapor deposition of Al will also inevitably lead to an accumulation of Al on the pillar tips which works in our favor.

Definitely need to do these simulations next.

[helge]

New structure suggested. Please head over to hackaday.io :-)

https://hackaday.io/project/18987-transient-thermochromic-contrast-camera