I’m exploring taking a greyscale photo and converting each pixel to a variable reflectance shape on a surface. I would like to have each pixel fit in a 1mm x 1mm square of the surface. This means I need a shape in that area which reflects a variable amount of light. So I’m using a solid trapezoid with the surface portion of variable width, matching the grey scale value. I would like to get these tiny trapezoids to print at least at 0.1mm detail, and preferably down to 50 micron details.
Here is a sample such surface stl file: https://drive.google.com/open?id=1BAcQ-u3aH2kz2bs5a8HxApgKfQ8kb_gL
I’ve tried passing this file to 2 printing services so far, I think both resin prints from Form 2 printers. But the details are quite washed out. I feel I’m getting perhaps 0.5mm details, even with a requested 50 micron resolution.
What advice can I use, to actually get 50 microns for this file?
well… that’s smaller than the laser spot the Form 2 uses to cure resin… so I don’t know how you can get that resolution.
the what it sounds like you are doing to me is printing a solid shape… but with each pixel represented by a flat plane, angled at different incidences to the incoming light so that each will have a different angle of reflection… right?
But for some reason your model has deep grooves in between rows of pixels. those are seriously complicating the print as they each will tend to hold resin in the groove thru capillary action and diffract the laser spot.
So First of all, I would eliminate all those grooves by having the lower edge of each inclined plane reflecting surface at the same height so that the block below that lower edge is a common solid.
Then- the worst thing you can do is orient this part flat to the build platform like this.
Many of you planes are angles so close to parallel to this surface that you are only going to get 2 or 3 layers to indicate the entire angle of the surface. that is simply not enough samples.
TO maximize the detail the form 2 can achieve in subtle angular changes, you need to orient the print nearly vertical.
think of the pixel as a vertical cross section in terms of how wide it is, versus how much its height changes across that width to form your angle.
It might be Fifty 25 micron layers wide… and fifty 25 micron layers tall at its tallest point- sloping to Fortyseven 25 micron layers at the low side of the trapezoid.
That means you are trying to capture that angle with only 3 layers… very poor resolution.
But, if you turn the cross section 90 degrees to print vertically- you are making Fifty slices thru that angled face. much better resolution of the angle.
Your primary limitation is the size of the laser spot- at 126 microns. The laser in drawing a profile is NOT rastered like the vertical layers are- it can step over a fraction of its spot width to capture details smaller than 126… however, the spot size does determine the sharpest corner radius it can draw.
So- what you want to do is make sure each Pixel is as large as you can get away with. the larger it is the more accurately the Form2 will be able to capture the edges of the angled faces. And orient the part so that it is printed just enough off of vertical that you don’t get supports on the detailed side.
The back you can sand and polish flat.
Thank you so much for the detailed advice!! Lots to try!
You would need to print at an angle, but most likely some of those gaps are going to fill in due to indirect exposure to the laser.
On my parts i have printed rows of small pillars with a base of 0.5mm and a tip of 0.1 mm and a height of 0.5 mm, this is printable ( i used the flexible resin). The biggest problem i had is the space between the pillars, i needed a minimum of 0.2 mm/0.3mm. Otherwise the laser would cure the resin in between as well.
I also noticed that i had to hand clean the part, since the space between the pillars will be filled with uncured resin, soaking the part in IPA will not be enough to remove this. I used a brush with short hairs, and gently brushed the part clean with some IPA before post curing.
Also i print in the same orientation you did, but i made a custom support on the part. The form 2 compresses the first layers (first 3 mm or so) firmly onto the bed to ensure it is connected to the bed. This results in a loss of detail since al the layers are compressed together.