One forum entry pointed out that the advertised volume is 210x210x205mm but CURA only allows objects printed with a size limit of 178,2mm x 178,2mm x 200mm.
The reason for this is that Cura accounts for the skirt/brim or raft.
The settings can be disabled to print larger sizes, but I wanted to stick to Cura limits to show to my students how large an object can safely be printed.
In Autodesk Meschmixer I quickly wacked up and object that represents 178,2mm width, 178,2mm depth and 200mm height.
Fig.1 - Object designed in MeshMixer for size testFig.2 - Object in CURA, too large
I played with the size options in Cura to see how it displays the limitations of size.
Fig.3 - Here is a message telling me that the current size cannot be printedt
After taking it back to 178mmx178mmx200mm it was happy and I could proceed with the print.
Fig.4 - Printable size: 178,2mm x 178,2mm x 200mmFig.5 - Raft laid down by the printerFig.6 - Finished Object
After the week 5 lecture I saw students using and interesting STL test file with overhangs and brisdges, printed without support.
After a quyick google I found the ftest file here: https://www.3dnatives.com/en/thingiverse-platform-190220204/
This is the outcome.
Fig.7 - Test print in the makingFig.8 - Test print in the makinFig.9 - Finished Test Print
Summary
This was an excellent exercise in getting to know the slicer software Cura and the Ultimaker 3.
I was amazed that the Ultimaker 3 could print bridges and overhangs without any support.
Design and 3D print an object
This week's task was to design and 3D print an object (small, few cm3, limited by printer time)that could not be made subtractively.
While I was thinking about an object that would be difficult to create otherwise and that an additive process would be ideal I came across an elephant sculpture from India. The sculpture has been carved out of soapstone by hand. It is a subtractive process and takes a long time.
Fig.7 - Indian soapstone carving
The only 3D program that I really know well is Autodesk Maya, I would love to learn Blender. In order to save time I decided to stick with what I know best and started modeling the elephant sculpture using Maya. Maya is a polygon mesh and Non-uniform rational basis spline (NURBS) modeling software, ideal for animation.
Nevertheless object files can be exported in object (obj) files that then can be converted into a Standard Tessellation Language (STL). A STL files can be used in a Slicer software to prepare and optimise an object for 3D printing.
I took an reference photo and imported it into the 3d modeling software and created a polygon cube that I stretched the length of the elephant's body (figure 8).
Fig.8 - Everything starts with a rectangular shaped object
A popular way of modeling is mesh face extrusion, where faces get selected and extruded. It is best done in orthographic view windows to assure accuracy (figure 9,10,11,12 & 13).
Fig.9 - Face extrusionFig.10 - Face extrusionFig.11 - Face extrusion
Edges and Vertices make up the polygon mesh of an object, all of those elements can be moved and copied, inserted and combined. I applied all of those techniques in order to model the elephant (figure 6).
Fig.12 - Vertices manipulation
In figure 13 the elephant is clearly taking shape by moving vertices to match the reference image placed in the background.
Fig.13 - Edges and vertices moved to match the reference image
I split the elephant model in half in order to insert holes and a container that can hold the baby elephant (figure 13). The baby elephant needs to stand on something otherwise it cannot be 3D printed.
To add holes to the geometry it is best to apply a boolean/difference operation using other objects that work as templates (figure 14)
Fig.14 - Elephant geometry copied in halfs to allow for boolean operationFig.15 - Boolean operation using hexagonic cylinders
At this stage something went wrong, I didn't know what it was till I ran the model through the splicer software before printing. When putting together all the geometry to be one I should have merged the vertices. Later I would have found out that the mesh had gaps and therefore wouldn't print as planned.
Fig.16 - Boolean parts , inner cage and elephant halfs
I used a complete model of the elephant from previous versions of my built ( incremental safe) an shrunk it to fit into the bigger elephant.
Fig.17 - Complete model with baby elephant inside
Here is the final model. I smoothed the mesh around the legs, head and trunk to make it more realistic looking.
Fig.18 - X-Ray view and smoothed mesh
3D-Printing of the elephant
I booked the FlashForge Guider II S Series printer for Monday at the visualisation studio at the University of the Sunshine Coast (USC).
The heated printing table is covered with green masking tape (figure 15), the tape makes the plastic stick better. I am using pink Polylactic Acid (PLA)plastic . PLA is a polyester produced by fermentation under controlled conditions of a carbohydrate source like corn starch or sugarcane (source: all3d.com).
Next to the 3D-printers is a PC that runs FlashPrint slicer software that came with the printer. It is not as sophisticated as Cura or other slicer tools.
When opening my STL file it turned out to be impossible to print. Figure 16 shows how the print would have looked if gone ahead with that mesh. I asked the techy about a possible reason. He said:"Mate, your mesh has holes that's why".
It would have taken me a long time to fix the mesh and print the model. I decided to print a elephant mesh created before applying the boolean operation, because that's when things went pear shaped.
I optimised the print in FlashPrint, that means I applied the print resolution (figure 14), scaffolding, raft size and inner support shape (hexagonal).
Once complete a generated G-code was sent to the printer.
Fig.19 - FlashForge Guider II S Series Single Extruder and Creator Pro Dual Extruder 3D printersFig.20 - Print specs for the elephant printFig.21 - the heated printer bed treated with green masking tapeFig.22 - Bad geometry, would not workFig.23 - Slicer fix but still would not workFig.24 - Elephant without holes and baby but working
Well, this is the elephant model without holes, inner cage and baby in slicer.
Fig.25 - Slicer image: elephant with supports
The above figure 25 shows the support that was applied to assure that the liquid print material would not sink to the printer bed.
Fig.26 - Print time preview
Halfway through the print I noticed that the trunk came off (figure 27), I paused the print job and tried to put it back in place but that didn't work. I continued with the print because other students were desperate to use the 3D-printer.
Fig.27 - Trunk coming off
Below is the finished print, with a render spatula it was easy to scrape off the printer bed.
Fig.28 - Print complete within printer
I left the support on, it looks kinda groovy.
Fig.29 - Completed elephant print
I plan to fix the geometry and try another print. Lessons learned are that it always takes longer than anticipated. Good planning and knowledge is key and of course practice, practice, practice.
In 2023, my initial idea was to create a shape that is informed by mathematical formulas and I came across Andrew Marsh's website that let's users create interesting shapes. It turned out that the site has no export function, so, I had to find something else.
http://andrewmarsh.com/apps/releases/supershapes.html
Fig.30 - A shape based on mathematical formula settings
I thought that I could continue with the Fibonacci informed shapes such as the Nautilus and started tracing a shape in Illustrator.
The plan was to import the illustrator trace into Fusion 360 to continue makeing a 3D model of it. I visited the following tutorial to learn mor: Tutorial SVG illustrator to Fusion 360
Fig.31 - Nautilus profile
After the tutorial it became clear that the assignment is to design and 3D print an object that would be very difficult to make otherwise except through a six axis CNC router.
Last year I designed an elephant with an elephant inside and decided to continue with that idea.
Last year's printing didn't go very well, I wasn't able to print the holes because the mesh was messy and had manifolds. I only ended up with the outer shell and a broken elephant trunk.
Fig.32 - Elephant model in OBJ format from Free3d.com
I imported the mesh in Autodesk Maya and modified it so that it could hold four elephants inside.
Maya is not suitable for making models for 3D printing because it uses only shell modelling modes and not solid.
Again, being time smart and therefore I decided to stick with the software I know best. Hopefully, one day I can learn other software tools.
Fig.33 - Elephant with cavityFig.34 - Big elephant with four smaller elephants inside
After that I made a flat base that would make it easier to 3D print the object.
Fig.35 - Finished modell with flat base
Once the model was complete I exported it as an OBJ file.
Next I opened Autodesk Meshmixer to resize the elephant to 12cm length and exported it to an STL file.
The STL can be easily opened and read by CURA, the slicing tool I was using for this project.
Fig.35 - Elephant model in Meshmixer for resizing and STL export
I saved a STL copy that could be imported into CURA.
Cura is an open source slicing application for 3D printers. It was created by David Braam who was later employed by Ultimaker, a 3D printer manufacturing company, to maintain the software.
Cura 4.3 can be downloaded from here: https://ultimaker.com/learn/ultimaker-cura-4-3-available-now
Fig.35 - Elephant model in CURA slicing software
Cura offers a plethora of options to fine tune print outcomes by setting resolution, material, infill, support (scaffolding) patterns, rafts and more.
Fig.36 - Printer settings
I started out with a very fine resolution 0.15 mm but it created a very long print duration of over 7 hours.
By changing the resolution, I was able to reduce the duration to just about 5 hours.
Once I was happy with the set-up I hit the slicing button and received the following preview, which indicates aspects of the print such as shell, infill, raft and support as shown below and in the colour scheme figure.
Fig.41 - Elephant model display after slicing using Cura colour codeFig.42 - Cura Colour Code
Last step was to save the file in Ultimaker's G-Code. G-code consists of G- and M-commands that have an assigned movement or action.
The G-Code is used by the printer to actually print the object.
Fig.43 - Elephant ready to be exported in G code
Below is an image of the Ultimaker 3 that I used for this project print.
Fig.44 - The Ultimaker 3 printerFig.45 - Using a USB drive to load the Elephant G-codeFig.46 - Preparing to print and heating up the print plateFig.47 - Starting print: 5hours 2minutesFig.48 - Laying down the raft
The printer laying down the raft, which is a horizontal mesh of filament deposited directly onto the build platform.
Fig.49 - The elephant print seen with raft and support.Fig.50 - Print completedFig.51 - Elephant print out of printer with raft and supportFig.52 - Raft and support removed with Stanley knife and pair of pliers.Fig.53 - The print looks very rough around the hole, one tusk and the elephant's tail did not print properly.
Summary:
The quality of the print was clearly diminished by clipping off the support and damaging the plastic surface of the elephant. Some detail such as the tusks and tail of the elephant didn't come out as planned. 3D printing with the equipment on hand should be only used for prototyping. It is not suitable for mass production due to imperfections and time it takes to print objects. The elephant is just 12 centimetres in length and it took 5 hours to print plus one hour cleaning up support manually.
I think I should try printing the elephant model without support, just to see if it works and this would create a much more detailed print especially for the three small elephants inside the large one.
If used for prototyping and proof of concept or for visualisation purposes, it is truly an amazing technology. The revolution predicted over 20 years ago is still to come.
My lesson learned from this assignment is that I need a lot more practice with different modelling software and printing objects with different requirements and different printers to be able to help and advise in 3D design and printing. It always takes longer than anticipated. Good planning and knowledge is key and of course practice, practice, practice.
3D scan an object
At USC is a HP PC with a Sprout 3D scanner that unfortunatly produces very bad results. I therefore tried photogrammetry via the Trnio app.
I had put my model on a timber stump outside, the sky was overcast, ideal conditions for photogrammetry.
According to Wikipedia: "Photogrammetry is the science and technology of obtaining reliable information about physical objects and the environment through the process of recording, measuring and interpreting photographic images and patterns of electromagnetic radiant imagery and other phenomena".
I took 9 photos making sure that I keep the same distance and angle to the object.
Fig.24 - Telemetric scan of the soapstone sculpture using the Trnio app
Once all images had been taken they were sent via the app to Trnio cloud render service.
When the render was completed I was sent a message and could download an OBJ or STL file.
Fig.25 - Completed image files uploaded to the Trnio cloud server ready for rendering
The mesh resolution is huge and very messy and not so detailed as expected. In the app preview it looks stunning because of the image map overlay. The mesh model was so bad that I decided not to print.
Fig.26 - Completed mesh, unusable for 3D-printing, lack of detail
Lesson learned: It would be nice to have had access to some or even one high performance 3D scanner to see the difference and learn how to operate the scanner. I might try to get hold of one from a different source.