What's new
What's new

Macro Photo of 3D printed surface finish

Pete F

Titanium
Joined
Jul 30, 2008
Location
Sydney, Australia
Recently somebody asked in another thread about the surface finish of 3D printed parts. Unfortunately that thread became polluted by people not involved with 3D printing, hence I walked away from it. Nevertheless I'm always keen to try to help those with a more open mind who are keen to see if 3D printing may be able to assist with their needs.

I've attached an extreme close-up of a 10 mm test cube, shot with an extension tube to try and get in as close as I can for those interested. It was just a quick shot and even though I stopped down, those who know anything about photography will appreciate the depth of field is difficult to attain with such macros, and I don't have time to focal stack the images. However hopefully it will convey enough information to be able to judge whether the finish is good enough for their purposes. The focus is about a third of the way back (ie 3-4 mm) from the nearest corner to the lens.

The part is a test cube that I use when experimenting with different plastics and when tuning the printer. It was just sitting next to the printer is not selected or reworked in any way, indeed it wasn't even dusted off! The profile shows the bottom and two random sides. The part was modelled 10 x 10 x 10 mm and measures 10.14 x 10.22 x 10.35. However I have never properly calibrated that printer for precise size as it's not important to me. I expect I would be able to achieve +/- 0.1 mm reliably however based on past experience. I really should calibrate it and see, however it takes time I don't have.

This was printed quite some time ago so I can't recall any details, however the material looks like PLA and the layers probably 0.2 mm, somebody could perhaps count them, but I normally don't bother with 0.1 for most applications. The cube sits on a 6"/150 mm rule for scale and a blue shop towel for texture. Unfortunately I don't have a surface roughness meter to advise an objective roughness of finish. I'd suggest if you need one, FFF is not for you anyway.

Hopefully that helps some people.

Test Cube Macro.jpg
 
Sorry I should have included a photo from a more normal perspective, as macro images can just look awful and bizarre given how much they're magnified. Below is the cube taken deliberately holding it in my hand for scale, with the same 6"/150 mm rule and shop towel for scale and texture. I just bullseyed the cube in frame and tried to catch the reflection of the light to convey a better feeling for the surface texture. It's clearly still zoomed in a lot, and on my monitor (and likely most) the cube would be several times life size.

Without an objective measurement available, I'd describe the surface texture as possibly similar to an LP vinyl record perhaps? I tried not to run my fingernail across too many of my records back in the day, but that's the best I can come up with.

Test Cube 2.jpg
 
Sorry I should have included a photo from a more normal perspective, as macro images can just look awful and bizarre given how much they're magnified. Below is the cube taken deliberately holding it in my hand for scale, with the same 6"/150 mm rule and shop towel for scale and texture. I just bullseyed the cube in frame and tried to catch the reflection of the light to convey a better feeling for the surface texture. It's clearly still zoomed in a lot, and on my monitor (and likely most) the cube would be several times life size.

Without an objective measurement available, I'd describe the surface texture as possibly similar to an LP vinyl record perhaps? I tried not to run my fingernail across too many of my records back in the day, but that's the best I can come up with.

View attachment 156330

That is probably a perfect explanation. We just printed some parts on our Fortus 400mc and that is a very good description of the surface texture.
 
Thanks for the datapoint. I was just reading up on the latest in FFF machines, as my young son has become enamored with them after seeing a demo of a cheap XYZ printer. We brainstormed different things we could make if we had one, and were both pretty excited. Both functional stuff and toys, and imagining what the future might bring.

As I was looking up specs, I've noticed some pretty nice improvements in what the 'average' machine is capable of, in layer thickness and resolution. Big improvements since I last tried to bone-up on FFF machines. Also, when did we switch from calling it FDM to FFF? I didn't see FFF much and now it seems to be the industry accepted term outside of higher commercial and industrial contexts.

The "budget" printers are getting pretty respectable. I'd really like to see what kind of soft jaws I could make with them, and how well they do in locating 'funky' parts in a Kurt vise, or comparable. That's one of the biggest shop solutions I would like to see. It's not terribly expensive to mill aluminum soft jaws with funky profiles, but if you have to do any 3d profiling it gets time intensive. Plus, time spent milling tooling is not money being made. Better to keep the spindles cutting deliverables...

For the applications I anticipate, the roughness you photographed would be absolutely acceptable, I would think. 0.2mm/200micron isn't even the 'average' layer thickness anymore, with some going down to 15micron layer thickness, even among the more affordable FFF machines. That would surely pull the 'vinyl record' texture a little tighter.
 
Just to quickly answer a couple of your questions. FDM is a trademark name owned by Stratasys. FFF is a term that isn't associated with and company and is instead a generic term for the process.

In terms of the machine capabilities I haven't seen any huge advances to be honest. Every time you decrease the layer height by half, you roughly double the print time. 0.2 layer height tends to be the sweet spot in compromise between the two. What some people may not realise is that prints are typically measured in hours, often in days, and I've heard of week long builds for larger models on bigger machines.

Beware of the hype associated with any particular machine. At the end of the day, a 3D printer like this is mechanically VERY simple. It uses stepper motors to drive a hot end across a platform, and another stepper to extrude the plastic. That's it! I haven't measured what one step on my machine actually equates to in distance, but as you can imagine it's not a lot, especially on the Z axis. There would theoretically be no reason that I can see not to single step each layer, but you are already well within the randomness of the plastic extrusion by this point, and the build would take forever. There isn't the huge difference in texture on a vertical surface with a low layer height you may hope for, instead you use this if you have surfaces close, but not quite horizontal to avoid steps in the finish.

I haven't ever printed soft-jaws, but I will keep it in mind if I do and try to remember to post here. The thing to remember with 3D parts are they are not (normally) solid, instead they are a shell filled with a honeycomb or various other patterned infill. Therefore you would need to consider crush strength on a part like that. There would be nothing stopping you from inserting internal bracing and all manner of support within the model that is completely invisible from the outside to aid strength. I have however made collets and tapers of various types and they worked fine, despite the apparent need for extreme accuracy. The slight pliability of the plastics and the randomness of the errors can work to provide an accuracy of the overall part in service that well exceeds the accuracy of any particular area of it. No it's NOT precision ground, but can sometimes assist in work holding in some situations where uber-accuracy isn't required and there's no practical alternative. If you need extreme accuracy/strength, in any particular area, this is probably not the solution you're looking for. However in my experience many of the things we use in our workshop don't need to be made to high precision, just because we can. A T-slot cleaner that comes with the label "milled from solid titanium billet" won't clean your t-slots any better than one 3D printed in plastic! It's that type of hack work I think these types of printers excel at, and there seems to be a lot of hack work going on around workshops every day!

Finally, there are many sides to successful 3D printing. Just because you own a DM DMV5000 5 axis machining centre, doesn't mean you plug it in to the nearest power outlet on "poof" out pops the parts. Don't let the low price tags of some of these 3D printers fool you into thinking that they are a domestic appliance like a toaster. There is still a lot to learn and know about operating these printers, and experience in this area is typically what separates the good from not very good results off them.
 
Great points @Pete F. There is plenty of application for 3D printers and price does not always mean quality. To the original post. Here are some macro images of structural prototype parts printed on a sub 1K printer in Carbon Fiber PLA (the CF is short chopped fiber).

unnamed (1).jpg
unnamed (2).jpg
unnamed (3).jpg
unnamed.jpg
 
face.jpg

I like the idea of this thread Pete. I also had a fellow contact me more so on the detail possible with my printer. Either way, this also shows the layering and could be helpful for some people :) This was printed from the back of the head to the tip of the nose at a .1mm layer height. Hopefully it can be of assistance to people :)

Thanks Pete

Chris,
 

Attachments

  • 1 size.jpg
    1 size.jpg
    82 KB · Views: 374
Does anyone know what the price point is for a machine which can produce consumer quality finished parts in durable plastic without re-work and which technology that would be?
Thanks,
Nick
 
Does anyone know what the price point is for a machine which can produce consumer quality finished parts in durable plastic without re-work and which technology that would be?
Thanks,
Nick

Probably the best answer in terms of suitable technology for producing a decent surface finish and decent speed is employing a DLP mirror array to modulate a UV light source, rather than scanning a point around:


The advantages of using a thermoset rather than a thermoplastic are considerable, elastomers are also possible using a UV process.
 
the most precise plastic parts I've seen printed (in terms of finish and tolerance) are done with 3D Systems HD machines... they make really nice stuff! Not mechanically functional but great for prototype.

20140320_194857.jpg


20140319_184851.jpg


20140321_120845.jpg


20140321_120824.jpg
 








 
Back
Top