6. 3D Scanning and printing

This week we were tasked with testing the design rules of our 3D printers.

Research

There are many settings that can be tweeked to improve or fix issues in a 3D prints results. This can be overwhelming to a beginner whilst being a near obsession with seasoned makers, modding their 3D printer to get the best results.

Even once you have great settings for your specific printer, when you change the material it can have errors once again. Many slicers have some default profiles which preadjust some of the main settings such as temperature and speed.

In order to test the results of your printer there have been many models created for this purpose.

The most famous one is the 3D Benchy. This model has many features built into it to display potential issues. These can be read about in detail here.

I do personally have a attachment to the calicat test model.

Neil also provided us with many individual test STLs which will focus on specific topics and considereations when it comes to 3D modelling.

The 3D Printers

In our lab we have two Ultimaker 2+ machines. They are getting a bit old but still work very effectivly. They have a built in exclosure which provides some safety from entraption and help maintain the heat. They have a boden tube extruder system which can be a issue for flexible materials and a glass bed which can make removing prints difficult. It takes 2.85mm filament and has a build volume of 223 x 220 x 205 mm.

I personally have a Prusa i3 MK3S+ that I got secondhand. This is a direct drive extruder which allows for a better chance of printing TPU and a removable flexible plate which aids in removing parts. It however does not have a enclosure. I takes 1.75mm filament and a build volume of 250 × 210 × 210 mm. It also have the nice feature of automatic bed levelling.

Set up

For the Ultimakers we use dimafix 3D printing spray glue to make sure the prints stay stuck.

For the Prusa I use isopropil alcohol to make sure the bed is super clean before each print. It is also useful for seperating TPU prints.

The Tests

As materials can affect the quality of a print we will be sticking to PLA for the most part, the most commonly used filament we use. Unfortunatly, we are unable to use the same roll due to them taking different diameter filament.

The ultimaker is using dark blue and the prusa is using turquise filament. I used the default profiles provided by the slicer other than adjusting the layer height.

Benchy

Overall the results are or quite high quality and comparable between the two machines.

Ultimaker - Pros - Nice surface texture - Port holes are circular - Base text clear - Cons - Some issues on overhangs on the roof, drooping - Gap on the top layer of the back square - One off center layer line, very minor

Prusa - Pros - Nice surface texture - Chimney streight - Coped well with the overhangs - Cons - Brim difficult to remove - Bottom text is messy - Some stringing (Probably filament is moist as that was a common issue)

As a bonus I printed a 50% benchy out of PETG on the Prusa. It had quite a lot of issues with stringing and the chimney didn’t keep its shape, however it coped quite well considering its size.

Overhang

This test compares what degree of overhang is possible without the part deforming and drooping too much. The standard number I have is to not do steaper than 45 degrees, however these results surprised me. Both printers would cope down to about 30 degrees without deformaties. The Prusa could do 20 degrees satisfactory.

Wall thickness

I only had opertunity to test this on the ultimaker. The uprights were all printed except for the 0.1mm. However due to using a 0.4 nozzel on the printer, walls 0.7-0.2mm were all one line thick and did not accuratly match the design model. They were all roughly 0.5mm, slightly expanded from the extruded 0.4mm.

The walls above this 3mm-0.8mm had more walls and slightly closer to their intended size.

In conclusion, if you are printing particularly thin walls it would be best to swap to a fine nozzel such as 0.1mm, expecially if you want them to be accurate to the designed thickness.

Bridging

Bridging is similar to a overhang but it connect on the over edge allowing for longer distances. The speed and flow settings of the print affects their quality. The Prusa preformed very well on this test. The test piece went from 2-20mm gap. Although very was some drooping occuring on the final 20mm gap, it correted itself quite quickly. Add another 5mm or so and I thick that would be reaching the limit of acceptable.

Dimentions

If the printer is not calibrated fully there is quite the high chance the print warps or develops a elephant foot which can change a parts dimentions.

This test has a simple 20mm outside and 10mm inside hole.

From experience I always add some clearance because the parts tend to be larger than the hole you want to put them in. However this cube printed basically spot on this time.

Clearance

I printed the clearnance model on my Prusa, however I had to create specific suports in the slicer. I knew from experience that the supports are difficult to remove so I didnt accept the auto generated versions as it entered the gaps between the bar and the squares. I used the ‘support enforcer’ feature to ensure that the supports were only generated in the middle of the bars between the squares.

The reults were somewhat dissapointing. The minimum clearance was 0.4mm. Any lower than this would not release from the bar.

I used this value when creating a hinge for this weeks individual 3d modelling project. The gap between the bar and the hole in the hinge section was set to 0.4mm.