a 3d scanner for 3d printing?

Any topics that don't fall into above categories. Please note that your thread may be moved to other discussion boards.
luksj
Posts: 1
Joined: Mon Mar 22, 2021 3:16 pm

a 3d scanner for 3d printing?

Postby luksj » Tue Mar 23, 2021 11:42 am

Hi!

I'm new to 3d printing and not sure if I need a 3d scanner for my prints. Do you use a 3d scanner?

Last month I tried to make my own mask inspired by this article. But my mask didn't turn out to be exactly my size and then I found another article about making a mask using 3d scanner and now thinking about buying a cheap scanner or rent it. Have you succeded in making high-quality 3d printed products with the perfect size without using a 3d scanner? Is it too complicated to use a 3d scanner with a 3d printer or my concerns are groundless?

Jetguy
Posts: 3165
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: a 3d scanner for 3d printing?

Postby Jetguy » Wed Mar 24, 2021 1:06 pm

No, you do not need one and no, they do not generally work like the magic you think they are.
Background, if you don't know who I am, I started in the early days of 3D printing circa 2009. I've seen the evolution of scanners, spent thousands of dollars and watched this now for all this time.

First is understanding how most scanners work in the price range you can afford.
They use a line based laser and expect that a straight line on the object deforms from straight and draws a silhouette outline of the shape of the object in front of the camera. So here is the fundamental problem- the laser is pretty bright and the object has a a non uniform reflection of light so the laser beam that should be a nice perfect crisp thin focused line of reference balloons and blooms in the camera view from high reflective spots, and can be hidden or masked by certain object shapes. The key point here is that the capture is flawed because what you really need to do is coat the object with a dull non reflective and uniform coating and that mostly defeats the users idea of not having to prep the object, possibly damage or affect it by having to paint it or cover it in a powder. Again, the biggest problem is the line laser used can be too bright, or on some dull object too dim compared to ambient lighting, but then the flip side too is the object paint or color or just surface in general is not a uniform reflection or tends to bloom and spread the laser line making the silhouette outline not crisp.

So great, now you have this burry capture and that results in a lot of extra non true noise 3D data points in the capture or massive cutoff areas where the scan just couldn't get data off the deformed silhouette line structure. Now you have to clean that mess up and make a unform smooth closed object that can even meet the rules for 3D printing. It takes hours and hours of cleanup.

Yes, there are better non-DIY tech solutions like real 3D scanners that use holographic produced laser lines in the infrared spectrum and because the laser is less bright, the laser color and thus reflection is different, they tend to work better in some cases and worse in others. I'm talking about using an X-box Kinect scanner and Skanect from Occipital. The problem I've seen with this solution is still the same, it works in some cases- lager objects like full body size, and still requires a very methodical cleanup to get a result.
https://all3dp.com/2/kinect-3d-scanner- ... -tutorial/
And the Great Fredini taught me a lot about his setup https://thegreatfredini.com/scan-a-rama/
https://www.youtube.com/watch?v=80M45YxtLR0

I also have the David scanner setup (one of the earlier ones) and probably spent well over $1500 in upgrades over time and it's still to date never given me the scans it promises https://en.wikipedia.org/wiki/David_Laserscanner.

I also built very early on some open source scanners and had bad results.

Last, the worst thing I ever bought ever in 3D printing, the https://www.xyzprinting.com/en-US/produ ... ci-1-0-aio
It was a pretty sloppy printer with lots of problems in the hotend and feeder, not to mention the slop and problems of the mechanics and bed, and then the scanner was just a complete and utter abortion.

So, have I ever 3D scanned- yes and the one project that worked required a specific computer and graphics card build to get the best performance and the system works best with people, not objects you typically want to scan and print. It in the end is a several thousand dollars setup. The licenses, the software, the hardware, it all adds up. And then, even after all that, the results take hours of time to cleanup.

No, generally this movie like dream of you scan an object as is and then a screen pops up, you load that into the slicer and print is still a dead folks. It's not generally reality.

Jetguy
Posts: 3165
Joined: Tue Mar 22, 2016 1:40 am
Location: In a van, down by the river

Re: a 3d scanner for 3d printing?

Postby Jetguy » Wed Mar 24, 2021 1:27 pm

Again, what I'm trying to share is how difficult the challenge is of 3D scanning for 3D printing.
You said you were new to 3D printing. Given that it's taken me all this time to master and know what I know about 3D printing, the thousands of dollars I've spent in the industry, the countless nights and hours, and here you are starting out.
I'm not saying these to put you down in any way, I'm saying that this skill- 3D printing is a massive thing to master.

I'm asking about just understanding the machine, upgrading the machine, understanding volume calibration, how the extruder works, what various materials need.

All this to say you assume you are starting with a given printable STL file, and it's now up to you and your skills to orient the object in the slicer, get the right settings to slice it, getting the machine prepared before the print, calibrated, upgraded, and so forth, and then starting the print, monitoring it the entire time, and ending up with a result. Can you read and understand gcode? Can you write your own custom gcode? Do you really understand how the extruder works, the details of a good hotend design VS a bad hotend design? Do you know the firmware and electronics? Do you really know how to master and tune he mechanics of the 3D printer? Can you design and build a better 3D printer?

Then designing for 3D printing requires understanding 3D printing and limitations and all that in getting CAD to export an object and then import that object and have a slicer slice it the way you know it needs to be printed and then setting up the printer and printing it.

But the next hurdle down the road- scanning an object, getting a good 3D point cloud with minimal noise, then still cleaning that up, correcting the mesh, filling the holes, making and basically morphing into a printable 3D object- well you have to have mastered the above previous skills. https://3dtotal.com/tutorials/t/cleanin ... -meshmixer

Again, a huge piece of this is all the software required, and sure you can play around on trial versions, but at some point it cost $$
Are you then prepared to take that leap?


Return to “Others”

Who is online

Users browsing this forum: No registered users and 1 guest