At our cancer center, we are thinking about getting a 3d printer. Very exciting! It is my job to investigate 3d printers. I tried printing a few things on a MakerBot Replicator 2X but ran into some artifacts. I happen to have free access to one I can use to try things out, but we are actually looking at buying the Lulzbot Taz 5 once we get our bearings straight.
I am very new to 3D printing so I could use some help in understanding and dealing with the artifacts. I am asking on these forums because (1) I hear these forums are helpful and (2) I want to ask if anyone here has had similar issues and if anyone knows what to do about them since we are looking to use the Taz 5.
I made an album to show the artifacts I ran into. Pictured are (1) a Meissner tetrahedron, (2) a jig to help us with some calibration measurements, and (3) an organ for patient visualization.
Basically there is a vertical spine/ridge that runs from top to bottom and some ridges just randomly appear.
Also, the left side of a flat bottom object seems to get lifted up. I saw another user with a similar problem. I understand it is a cooling issue, but I am too new to the field and could use some more explanation on strategies to resolve the issue and their implementation.
Finally, I tried printing the organ as a solid object, but about 10% into the job, the MakerBot just went bonkers, making a glob of filament. Any ideas on how to adjust the settings?
I took a quick peek at the album… A lot if the issues may be from the slicing settings. The “ghosting” may be the machine, reducing the print speed may help. Maybe a little more insight into the slicing software and parameters would help.
As for the lifting of the print, that’s a common issue and struggle which is unique to the machine, filament and print environment. The TAZ5 PEI bed should help with adhesion of PLA and ABS. Other print strategies such as brim / skirt and proper nozzle height would help mitigate the warping / lifting. Ultimately an enclosure for the printer will greatly reduce a lot of print issues simply by creating a consistent and predictable print environment.
That Vertical spine issue may be due to the starting point of your extruder. Many printers out of the box tell the extruder to start each layer at approximately the same spot on the outer perimeter. At the start of each layer there is typically a very small amout of excess plastic. The easiest way to get rid of that is to tell the extruder to start each layer on an inner perimeter at a random location. then the excess just gets eaten by the printer.
Lifting, exactly what kcchen_00 said. Depending on what you are doing with these, PLA might be a better plastic for this if you don’t need a lot of structural strength.
For the blobby part, that could be a couple of different things. If the organ model had a void in it, such as a Heart chamber, you might be running into the “trying to print a piece of something before the attachment point exists” issue. For example, if you were printing a model of a person doing jumping jacks. you could print their outstretched arms in the up position because the shoulders would print first, then the arms then the hands, etc. You could not print the same model with arms in the downward position without using support materials, because the printer would try to print the fingers then the arms then the shoulders before the shoulders actually existed. Since there wasn’t an attachment point, it ends up just dripping plastic all over the model.
You also could be dealing with a model error. Depending on how you ended up with that STL file, it might not be “manifold” (a solid model with an intact surface mesh with no gaps in the mesh) . You can try repairing the model with the free version of netfabb, or the new online Microsoft version of netfabb that actually works better (https://modelrepair.azurewebsites.net/) but be aware that actually alters the model, and could end up inadvertently closing very small holes that you were trying to model to show a puncture, etc. I’d have to take a look at the model to be sure, though I realize there would be patient confidentiality issues that could make that infeasible, etc.
Depending on what you are planning on doing with the models, you might also want to look at one of the resin or laser sintering printers, such as the Form1 or the Autodesk Ember. The items they produce aren’t nearly as structurally strong as what you can make with a TAZ or as big, but you can get very high resolution with them. Don’t get me wrong, a TAZ is a great printer and I wouldn’t hesitate to recommend them, but for high precision medical models that don’t need to be large or have all that much in the way of strength, there might be other things to consider.
Other print quality things to be aware of:
Make sure your filament diameter Is set correctly. If you have 1.75mm filament set, but the filament actually measures 1.82mm diameter, you may be inadvertently overextruding 3% more plastic than you actually were intending, or the other direction if it was smaller than actually set. It’s not unusual for filament to vary somewhat between roll lots and manufacturers.
Moisture can affect filament. store it in a sealed dry place, preferably with some dessicant. If you see lots of tiny pin head sized blisters on the surface of your print, that can be moisture vaporizing and blowing outwards.
Calibration calibration calibration. You cannot possibly spend too much time or effort dialing in your printer. Start from a solid base, follow the various extruder calibration guides out there, and print all the torture test prints you can find and you’ll be amazed at the difference it can make.
I am not sure what you are referring to as “ghosting”. By slicing settings do you mean what the up-down resolution, % infill, speed, etc.? In order to specify these type of settings, I have to use some software from MakerBot. I can try to use a slower speed and see how that goes.
What is the PEI bed and how does it work? Is it different from the MakerBot’s print bed? I am unsure what you mean by brim/skirt. The nozzle height settings would be available in MakerBot’s software or the equivalent from Lulzbot? The MakerBot I tested on was enclosed.
Interesting. What should I look for in the settings to adjust this?
PLA is affected by this warping to a lesser degree?
The model looks like it is really just the outer shell/surface and empty inside. Should I prefer making it a solid model before printing it? All of the models printed with this honeycomb structure on the inside.
The Meissner Tetrahedron, the acorn looking thing, I just obtained from thingiverse and printed. The jig with the holes in it, we made from TinkerCAD. Blobby organ thing, I made by (1) using 3DSlicer to convert the contoured structure from the CT scan into an STL file then (2) used Meshlab to smooth out the layers and finally (3) I removed some messed up points from the smoothing and had Meshlab fill them in. All of these looked hole-less before I printed them (except for the jig which has intended holes).
I will check those out. Some of the structures to be printed will be big, ie around 30cm x 30cm. None will have to support weight on top of them though.
Interesting. So I should take some calipers to measure the diameter and enter that into the software’s settings? Also, sounds like this would be the type of issue that, more-or-less, uniformly affects the print. Good to keep in mind.
I did not notice anything like that, but the filament is stored on the back of this printer. I will keep an eye out for this.
The MakerBot is not mine, but rather it belongs to the college where I work. I assume the company calibrated it when they installed it this spring. I guess it needs regular recalibration? Quarterly? Semianually? Anually? I am not sure what their situation is, but the print jobs do not seem to be failing. If I understand correctly, we would be calibrating the Lulzbot ourselves once we install it? PTV_Contour_ClosedSurfaceModel_Fixed.stl (6.11 MB) copy_of_chamber_hdr__inter-comparison_jig.stl (235 KB)
In Slic3r it’s "Start peremeters at Concave points and Non overhang points is are both checked, with optionally “randomize starting points also checked”
PLA does not contract as it cools, which is what causes the warping. PLA has its own issues, it can actually expand slightly and lock itself in the extruder bore if it overheats in the wrong place so you need a dedicated extruder fan, and you also need a fan on the extrusion as it comes out otherwise it tends to make a lumpy soupy print. It’s also more brittle and not as strong. Many people prefer PLA though and its commonly used as a “beginner” plastic over ABS. I personally prefer ABS for most projects, but I have a really well dialed in printer with a PEI bed.
Most slicer programs will interperet a solid surface STL as an actual solid model. Hexagon infil (or honeycomb) looks cool, but it is also one of the more difficult infil types to print because there is a fairly high amount of retraction that occurs on the vector changes. You might try changing to “rectalinear” infil and see if that works better for a while
The hole Jig part is not manifold. it has a slight error down the bore of the small hole and in the face of one of the two larger holes. Easily fixed and probably not large enough to really effect printing of it.
The blobby surface thing, that is fully manifold, but its also going to be really hard to print without support. What I would do if I were you, would be add a round 3-4mm thick cylendrical base about 70mm in diameter, extrude a 25mm stalk up from it, then park the organ model on top embedded in the stalk and export that. That way you have a good bed contact surface that should let you print the whole thing.
Yes, exactly. A good set of digital calipers should be your next purchase immidiatly after buying a 3d printer.
There are various types of calibration that need to be looked at. Extruder calibration will generally not change over time once you have it dialed in correctly, though fillament width and extruder latch spring tension can affect that somewhat if they aren’t checked. Bed level is important to check fairly often, but you will know when it’s out of true, mainly because nothing sticks to the bed correctly. Some printers have automatic bed leveling, the makerbot may be one of those, I don’t know for sure either way. A makerbot also may have other quirks that I don’t really know enough about to give you good advice on what to look for either way. What I would reccomend doing is print the calibration objects and then verigy with your calipers. If they are off significantly, or the prints look odd in quality, post a picture and ask for advice on how to get better quality.
Ghosting is the problem you point out in your third photo, where a ‘ghost’ image appears in the surface around features like around holes. Kind of like echo images, or ripples in a pond.
PEI is a type of fiberglass composite sheet or plate. The TAZ 5 has a .01" sheet of PEI that is adhered to the surface of the glass build plate. It performs better than blue tape, glue sticks, ABS juice, or about any other common printing surface to hold parts while printing, but (usually) releases the parts easily once cooled.
If you are buying a new Taz 5 it will probably come with the fan already.
Several different ways to check for non manifold parts. I find the easiest way for me is to download the free version of netfabb from http://www.netfabb.com/ and open the STL file with that. If it isn’t manifold, it will prompt you to repair the file. Depending on what you use for model editing there might be a plugin that does something similar.
A raft will do essentially the same thing so if you already know how to use them I wouldn’t bother with the manual version. I was just wondering if the print failed because it rocked itself off the bed and maybe a bigger manually created one might help.
I run a lot of PLA on my older Taz with the PET bed surface and on my Mini with the PEI surface. PEI is better for large bed contact / stiff parts. It releases much more easily than the PET. For smaller or flexible parts, it doesn’t really matter.