A quick update about the 3D printed wing have been experimenting with. Turns out there are some interesting differences when sending the model to different printers.
You may have read my previous post on my printing a wing section with a flap. This week I tried printing the same model on an Ultimaker 3. As I wanted to try out using dissolvable support material. The previous printed wing section used the same material for both the wing and the support.
The Unlitmaker 3 can use two different materials. This allows you to print a model in two colours or two materials. So I went with PLA (same material I made the wing from before) and PVA. PVA (polyvinyl alcohol) is a water soluble filament which makes it ideal to use as a tempory support in the printing process.
The only differences between this print and previous one are the printer, the software they use and that the support material is different. But there are some interesting differences between the two.
On another note, you can now download the SLT file for the wing from Thingiverse.
The print was pretty uneventful. No problems. I just used the ‘normal’ settings when printing and selecting the support material to be the PVA. The print tool a little over 24 hours.
As you can see, the support has a few defects. This is okay because it is only the support and will be removed.
Cleaning the Printed Wing
How do you remove a water soluble support? Quite simply, you put it in a bowl of hot water (but not boiling). And that’s what I did. Although I might need a bigger bowl in the future as the one I found in the office was a little small.
I helped the support off the wing with a screwdriver. At this point, the water had also cooled down. There was still some support material between the wing and the flap as well as around the axels. At this point, the flap was not movable. I decided to take the printed wing out, dry it off and dispose of the water.
I then took the wing home and submerged it in a larger bowl of water. Every now and again, I wiggled the flap to allow water to seep into various places. I also topped up the bowl with some more hot water every now and again.
When I could, I pulled the slither of support from the between the flap and wing with some pliers. This freed the flap. I then submerged it back in the water and went back to periodically moving the flap.
I didn’t manage to remove all the support. There is still some around the axels. At this point, the printed wing had been in the water for hours and there was less and less support coming off. However, it does not impede the movement of the flap too much. But support material removal should be considered in future designs.
The Interesting Thing – The Differences
The first thing I did, after the print was finished, was to look at the newly printed wing to see if it look okay – and it did. Then I compared the airfoils of my two wings and found that…
They are different! At this point, I had no idea if the first one is short, or the second one is long. This may not be a problem if a product designer is using a 3D printed models to test design or show to clients. But aerospace is a precise industry and the differences here are clearly visible.
These are just some test pieces so there is not much of a problem. But it could cause issues if on a real aircraft. For example, you have designed an build a complete 3D drone or model aircraft airframe. You are then flying it far from your workshop and a wing becomes damaged. Your friend has a 3D printer and prints off a wing for you. Then you find out it just doesn’t fit. Or worse, you don’t notice it is loose fitting and the new wing comes off during a flight. You could destroy your whole aircraft.
So I decided to measure the difference between the two printed wing sections.
The first printed wing is 1.5 mm short. With the chord length meant to be 150 mm that’s 1%. Not an insignificant amount. I didn’t expect them to be exactly the same, but this a large difference.
Also, removing the support material I foud another difference.
You can see the infill on the top solid part. Another thing that was not expected. The outside of both the flap and main wing section are both meant to be smooth flat surface. Well, as smooth and flat as 3D printing gets. However, when checking the slicing in Cura it appears in the preview.
So what’s happening?
I don’t know the exact reason why these differences happened. But I think there are two factors at play:
- The model file.
- Different printer software and settings.
The model file
The STL used for printing the wing is one made from an assembly saved as one file. The assembly is made of two parts. If I had saved the parts as separate STL files and imported them both into the Cura slicing software it may have sliced them differently, with flat surfaces. Or it could be selecting a different material for the supports.
I think this is what caused the infill to come over the top and bottom surfaces of the flap and cut out section the wing. I will have a play with different models and materials in the software and see what happens.
Different printer software and settings
On the day I set the printer running a came across a paper from NYU’s Tandon School of Engineering.
The relevant part of abstract reads:
With the introduction of these features, only a unique combination of processing and printing parameters will provide a high-quality component and any other conditions will result in a defective or inferior quality component. The high quality part will print only under a unique set of STL file resolution, slicing conditions, part orientation on the print bed and the printer operating parameters. It is recognized that these features are not off-the-shelf technologies but represent design methodologies that need to be developed for the given component design and the desired [adative manufacture] technique.
It seems this is what I have done, if unintentional, and is now something to take into consideration when printing future wing sections. The full paper can be read here (if you have access to the journal).
Both printers have their own software to slice the SLT file into a format they use. The Wanhao software creates a .i file. Whereas the Ultimaker’s Cura software creates a G-code file. G-code is an industry standard way of controlling manufacturing machines and tools. G-code has been around for a long time. Originally being used in ‘Numerically Controlled’ machines in the 1950s.
So the differences in the machines & software, settings applied, will affect how any future wing are printed and be factored into the design. This what caused the differences in length between the two printed wings.