I recently got hold of a Snapmaker 2A350 3D printer, as I wanted to be able to print larger items than is possible with my Cocoon 3D printer from Aldi.
Although I had designed a number of 3D parts for the X-Class Loco model using Lightwave 3D, I haven’t designed any aeroplanes in 3D, so headed over to Thingiverse and downloaded 3D files for a Boeing 787 Dreamliner, created by a member going by the name of Lukys. He also has a video of his model on YouTube. Unfortunately for me as an English-only speaker, the video and most of the comments are in Russian (?), but still worth a look to see his attention to detail. His finished 1/72 scale 3D print looks fantastic!
STL and Gcode Files
Once 3D models are created in a CAD/modellingprogram, the standard way of saving and distributing the file is in .stl format (Stereo LIthography).
As 3D printers print by building up layers of plastic filament (usually ABS or PLA), stl files have to be ‘sliced’ into layers in a program such as Ultimaker Cura. The plastic filament comes on a reel, usually specified by filament type, diameter and weight.
All the details of the 3D printer to be used are loaded into Cura, so the virtual layer slices correspond to the actual printed layers on the 3D printing machine. The result is saved as a Gcode file, which contains all the 3D geometry information for the model, specific to a particular printer.
Also included in a Gcode file is printer-specific information such as the diameter and flow rate of the extrusion nozzle, the dimensions of the print bed (H, W and D), the desired temperature of the extruded plastic and temperature of the print bed, print speed and so on.
Fuselage section 1 stl file loaded into Cura for slicing, to suit printing on the Snapmaker. Cura reports that at the current scale of 1/100, this piece will take 3 hours 17 minutes to print and use 28g of plastic filament. Scaling up to 1/72 would take 8 hours 13 minutes and use 80g of filament!
Click image for larger view
Scale and Size
I decided to print the model at the provided scale of 1/100. As this is a 787-8, the model is almost 570 mm in length. I started by scaling up to 1/72, but realised that the model would be 787 mm long, use a significant amount more filament, and take ages longer to print. After printing out a 1/72 scale flight deck section (about 7 hours), I decided to stick with 1/100 (about 3 hours for the same section). This totaled 580g of PLA for the complete 1/100 model; probably a full kilo reel for 1/72.
The Snapmaker2 A350 3D printer has a build volume of 320W x 330D x 350H, much larger than the Aldi Cocoon’s volume 200W x 200D x 180H. In the photo below, the Snapmaker (foreground) and the Cocoon are printing parts for the engines. The Snapmaker print was superior, so I decided not to use the Cocoon for the rest of the model.
Below: Printing part of the nose landing gear. As a small part, the print only took a few minutes. Some of the larger fuselage parts such as the section where the wings join took over 8 hours!
I am using PLA (PolyLactic Acid) plastic filament, as it’s relatively cheap and is easy to work with. The parts are reasonably smooth, but print layer lines are visible, so finishing will involve cycles of sanding and applying spray putty and primer before the final paint coats. Some people also sparingly apply acetone on a rag to smooth PLA surfaces… must experiment with that!
Below: Printed PLA parts, awaiting joining with Superglue.
Click images for larger view
Most of the PLA parts were glued together using SuperGlue, but the wheels, wing flap actuator covers and antennas were attached with a hot glue gun. The wings have not been glued and remain as a jam-fit so they can be removed for transporting the model if necessary. Tip for using SuperGlue: wear nitrile gloves so you don’t stick your fingers together – and the glue won’t stick the gloves together either!
Click images for larger view
My engine fans were slightly too large to fit into the nacelles, but the problem was easily solved by sanding the blade tips when spinning the fans in a drill chuck. If I do another print, I’ll reduce the fan diameter slightly in Lightwave.
The stl files came with a basic stand, but I modified it in Lightwave 3D to include the Boeing logo and 787 as extruded parts. Click images for larger view
Below: Stand base in Lightwave 3D. Boeing logo and 787 added to the original stl file and base hollowed out a bit to save print time and plastic filament. Modified base exported as another .stl file for slicing into Gcode with Cura. Click image for larger view.
This stand design turned out to be unsatisfactory due to the effects of gravity over time. A new stand with a redesigned arm was produced in July 2021 (see below).
Painting & Finishing
Once the main parts were glued together, the painting process began. This involved many cycles of of automotive grey spray putty, followed by wet sanding between coats, then a couple of matt white primer coats followed by a couple of gloss white top coats. This removed the join lines in the fuselage and smoothed out the print lines, leaving a nice glossy surface. The silver along the leading edges of the wings, tail and tailplanes was aluminium spray paint, masked with Frogtape yellow (made specifically for delicate surfaces).
The main masking for the red tail paint was also Frogtape yellow. To make the masks for the tail Kangaroos, I created the shape in Adobe Illustrator as an outline, mirrored it for the other side of the tail, laserprinted the outline shapes onto copy paper and cut them out with a craft knife. The two Kangaroo paper stencils were then applied to the tail with a glue stick. When the red paint was dry, it was easy to remove the stencils by wetting them with water and rubbing them off, showing the white paint below the red layers as Kangaroo shapes.
The window ‘glass’ was formed by smearing ‘Micro Kristal Klear’ across the window holes with a toothpick. Similar to PVA glue, but a bit thicker, this product dries clear. The small side windows were easy to cover, but the larger flight deck windows took a bit of practise to form a film that remained intact until sufficiently dry, rather than popping. The solution was to hold the model by hand in order to keep relevant window flat and level for a few minutes; alternating having the window up towards the ceiling, then down towards the floor. This technique evened out the drooping of the film (due to gravity), producing a reasonably flat finish rather than bowed in or out.
I created the livery decals in Adobe Illustrator and laserprinted them on to some sheets of water transfer decal paper. The Kangaroo decals comprised 8 point red outlined shapes with grey shading. These decals finished off the stencilled Kangaroos on the tail with nice smooth lines. Qantas aircraft show the last 2 letters of the registration as small white letters on the top rear of the tail. Since it’s not possible to laserprint in white, I’ll have to figure out another technique to add them – perhaps some white Letraset will do.
I decided the stand looked better painted in satin black, with the lettering and Boeing logo detailed with a gold marker pen. As the stand was quite thin for the plane’s weight (about 640g) and was bending too far, I made a curved part to strengthen the vertical arm of the stand.
The 3D model with a recreation of a 1970 Qantas poster. When I was a boy, I had an original copy of the poster on my bedroom wall… wish I still had it, they’re worth a small fortune! I recreated it in Adobe Illustrator and printed it at A3 size.
Stand strengthening and replacement, July 2021
Since the model was finished in April 2021, it has stood on its stand without falling off. However, the effects of gravity were having their toll on the thin-section arm of the stand leading to significant warping. I designed and printed a piece that was glued to the rear-facing curve of the arm to strengthen it. However, this was not successful, as after a short time, the warping continued and the almost-600 gram model was in danger of falling over.
Although the stand base was strong enough, a redesign of the arm was needed. The flat cross section approach was ditched and I tried to design a conical shaped support which could be retrofitted to the existing base. It turned out much easier to modify the original design, retaining the base design but replacing the old arm with the new, and re-printing the whole thing.
Lightwave screen grab showing the old arm (black wireframe) and the new arm and old base in white. The old base design was retained, but modified slightly to accommodate the new arm.
Perspective view of the new arm and old base
Printing the new base on the Snapmaker 2
The finished arm is hollow; this photo shows the support material which was removed after printing was completed.
The old and new stands. Despite the addition of the strengthening piece along the rear curve of the old arm, the warp continued to get worse.