3d printing lense shims

D

Daimon

Established
Local time
11:28 PM
Joined
Jun 9, 2023
Messages
74
Location
Poland
View My Gallery
Some idea I already tried myself - still an experiment but haven't seen similar topic here so sharing my thoughts. I'm getting lots of very useful information on this board, thought I can try to share some ideas as well.

2 lenses required much different shim compared to the one already installed and I don't have other spares but I have a 3d printer so though I'd give it a try.

Model
Model is trivial - shim is just a cylinder with outer diameter and height and hole inside which is inside diameter.

model.png

3d printing precision
theoretical: depends on concrete 3d printer. I have very standard extruding printer - 4 stepping motors (x, y, z + material extrusion). Theoretical precision depends on stepping motor step size and gearing between motor and actual movement of nozzle. In my case minimal "Z step" is 0.04mm - i.e. shim thickness can be theoretically set to any value which is multiplication of 0.04mm. So if I want a shim which is exactly 1.02mm I'll get 1.00mm or 1.04mm

practical: in practice many other factors have to be taken into account. 3d printing is essentially precision pouring of molten plastic (240 Celsius in my case) which shrinks while it cools down. A lot depends also on material quality and printer calibration. It may happen that 1.02mm thickness model ends up being 0.992mm or any other arbitrary deviated value. But that's not a problem, can be taken into account during shimming process

3d printing time
once I got everything setup it took about 4 minutes to print ~1mm shim - much faster than "manual" cutting and can be used to prepare "any" shim

material
I chose PETG because from "cheap" materials it's the most temperature stable one, is elastic (does not crack) and pretty strong (similar to one from which plastic drink bottles are made - but if you increase thickness it becomes much more stiff)

process
PETG is more elastic than brass or aluminum - may be squashed a bit when lense is assembled. Prints also won't be super precise, expected shim dimensions measurements (50mm Jupiter 3 f/1.5 Information — Jason Howe - thx Brian 🙂 ) may be off as well. Instead of trying to print super precise "additional" shim we just replace original shim with a new one. Here's the process

1. Get current model shim thickness (first model is same as original lense shim)
2. Measure how much lense is off with current shim and calculate expected shim delta (amount of increase/decrease)
3. If delta is less than 3d printer precision (0.04mm for me) - use sand paper (it takes 4 minutes to print a new one so if anything is wrong start over) - if lense is fine, process can be finished here
4. If delta is bigger, apply delta to model (that's important, we don't try to print exact dimensions each time, just apply delta), print a new shim and go to step one

It took me 2-3 prints to get lenses exactly right once I figured out all the details. Here some printed shims from Jupiter 12 and Jupiter 8 fine tuning.

IMG_20230827_183450_707.jpg

shim temperature stability
all materials shrink/expand when temperature changes. It's not only important during printing - we also want to ensure shim maintains its dimensions when air temperature changes 🙂 I used PETG which has one of best stability in "cheap" 3d printing world but it's still 3-4x worse compared to aluminium and brass

Theoretical values (linear thermal expansion coefficients):
PETG: 51-68 µm/m-K
Aluminium: 21-24 µm/m-K
Brass: 18-19 µm/m-K

So for 1mm thickness of PETG shim and a difference of 30 Celsius/Kelvin degrees in temperature we get:
0.001m * 60 µm/m-K * 30K = 1.8µm (0.0018mm) linear expansion

For practice check I used actual printed shim, marked measurement point (to make sure I'm measuring same area) and used micrometer:
"+27 Celsius" - average 1.178mm (from 5 measurements, 1.177mm, 1.179mm, 1.179mm, 1.177mm, 1.179mm)
Around "-15 Celsius" (my freezer) - average 1.176mm (from 5 measurements, 1.176mm, 1.178mm, 1.175mm, 1.175mm, 1.177mm)

0.002mm between +27 and -15 Celsius is negligible - actually I doubt my micrometer is precise enough to measure that value - it can be just micrometer error which was introduced by turning it off and on. Anyway that value is in the ballpark of calculated theoreitical value so it looks fine for both theory verification and actual usage.

IMG_20230827_183332_714.jpg
 

Attachments

  • IMG_20230827_183422_748.jpg
    IMG_20230827_183422_748.jpg
    173.2 KB · Views: 3
Nice, I have considered this but haven't tried it yet. I'm guessing based on the width of the ring (not height of the shim) the slicer is treating it all as external wall to spiralize the printing?
 
I just used exact inner/outer diameters as original shim and printed it as solid - layer after layer (but used 0.04mm layer height - that's minimal for my printer). This results in a slight "seam" build up at a single place but that can be easily scraped off.

TBH haven't checked how precise inside/outside diameter turned out - after printing - it fit lense properly so haven't put too much attention to it.

Spiral mode sounds interesting, haven't tried it. The most standard solution worked good enough
 
I need to get a solid printer.

How strong is the material at 0.1mm, 0.2mm, 0.5mm, and 1mm?
These are the most common thicknesses that I've used.
 
PETG is similar to PET Bottle (hence the name) elastic when thin but gets really sturdy as a solid shape.

Here are some characteristics with actual numbers - Ultimate 3D Printing Material Properties Table

There definitely are better materials (I just used what I had) and better printers (I have simple one from few years ago) but it was already calibrated for my material so turned out to be good enough.

Don't expect machining precision - 1.00mm model may turn out to be 0.90mm if printer is not calibrated right. Even with calibration there are tolerances which are "wider" than what is required for exact shimming. But with a bit of hacking (not expecting exact values but working on deltas) for me it turned out to be good enough.
 
I wasn't thinking about inside/outside diameter precision, but spiral increasing height precision due to less chance of layer interactions cause surface defects. Looking at your rings it looks like it already printed that way, probably as it is all 'walls'. The difference is the print head moving in a circle vs back and forth like it typically does for the bottom/top of a larger item.

The other thing that might be interesting (though may need a larger nozzle) is to print in vase mode as it would put it all down in one continuous extrusion. That will only be one wall thick though which is why a larger nozzle might help.
 
I use Polyamide 12 Nylon for all my prints, not familar with PETG. There are also versions with glass bead fills for parts that need higher strength.
 
Did you check if the thickness was uniform? When I tried making shims they had some taper due to slight misalignment between bed and rails.
 
splitimageview said:
yes, can the desired precision be reached?
Click to expand...

I think it will depend upon the shim height itself, and also on the focal length you are shimming. The wider you go the larger a small difference makes.

If you need something like 0.05mm you aren't likely going to be able to do that with typical FDM just due to first layer height. If you need a 1mm shim that is much more likely though it might take some iteration to get actual 1mm. Things like Z offset will change the designed print height from the actual print height. When I'm building Kraken's I print single layer shims on occasion to get the last little bit dialed in.

I've never printed with resin but it would allow considerably more granularity as well.
 
I think you look at the problem on how to get most precise 3d print from model - that's not the problem to solve here.

Goal is moving focus plane where it should be. You don't solve it by calibrating ideal prints. You solve it by adjusting model values to get you where you want to be - see description from first post.

With method I described in first post was able to improve 50mm lense at 1m like this:

DSC03457.jpg

DSC03460.jpg

Good enough for me. Alternative was leaving it as in first image 🙂
 
We are saying the same thing. What I am saying is if you need a 1mm shim to fix your focus you may or may not be printing a '1mm' shim. You might have to scale it to 1.1 or .9 or whatever to get a shim that is actually 1mm. As long as the the shim is consistent all the way around that doesn't really matter as it is iterative. If the shim you need would be 0.05mm you very likely aren't going to get that with a typical FDM printer just due to first layer height.

You can pre-determine what height shim you need by measuring the lens length with a caliper at its indicated focus position, then refocus to optical focus and remeasure the length of the lens. The difference is the difference you need in your shim. If the lens got longer that is how much to add, if the lens got shorter it is how much to subtract from an existing shim. Doing it this way the more accurate your printer is (defined by what you feed it is what you get) the quicker/easier this becomes.
 
Such measuring is not solving the whole problem here. It may be fine for aluminum or brass shims which don't compress - or at least not much.

3d printed shim will compress "some" amount. How much - I'd say impossible to calculate with the precision needed here and each material type (or even batch) may behave a bit differently.

That's what I'm mentioning that it's not about getting perfectly sized print.

It's about getting whole assembly right and by measuring final outcome, not print
 
You must log in or register to reply here.
Back
Top Bottom