Week 8: 3D Printing and Scanning

3D Printing 

I bought for my birthday a budget 3d printer: the anycubic Mega. It cost £160 from eBay and has a build volume of 210mm x 210mm x 205mm. Below is the anycubic mega on their website.  I was  interested  to test the capabilities of this printer and also hopeful that I could use the printer to help with lots of home projects. One project i was especially interested in was using the printer to make molds to cast carbon parts into (see casting and molding page)anycubic mega 3d printer

Setup of the printer was very easy: it came in two parts: the gantry and the base. These were connected with 8 small machine screws. I also had to attach 3 cables up but the whole process was incredibly easy and intuitive. The printer came with an SD card that had a pair of owls as a test print. It also had a manual that suggested using Cura to export files to it. The printer can either be driven by a laptop plugged in or the SD card. The first print of the owls went well for the first 60 %. This was where I ran into my first issue.

owl failed print

I figured out that this was because the base of the owls was quite small in comparison to their height, when the printer got to a certain height, there was not enough area of the print to keep the model stuck down when the printer head brushed against the top of the model. There was too much force at the top so it unstuck the model and gummed up the top of the model. I fixed this issue by re-leveling the print bed. I had initially done this in a bit of a rush because I was excited to start printing, however, it meant the print head was too close to the model on some parts of the bed. I found a very useful tutorial on youtube and followed the steps:

I found that my bed was too high on the front left corner and too low on the back two corners, once I had dialed in this properly, I found I was getting much better more consistent results. I also had to change filament from the blue PLA that was supplied with the printer to a gold PLA that I bought from ebay. The change of filament also had an effect on the quality of the prints because even just from a visual inspection it was better quality: it was much less brittle and seemed a little smoother than the included stuff. I also went into cura and tried messing with some of the settings there. One thing that made a huge difference was the build plate adhesion setting.owl with raft

This is the raft setting: cura automatically adds a 2mm thick raft underneath the part. This increases the area for the model to adhere to the build plate and reduces the chances of the print failing. The printer needs to print support material so that it can make overhangs. I found most of this support material came away easily after the print was finished but there are some scratches on the feet where you can see I struggled to get the support structure out from between the claws. This is something I’ll need to consider when I’m designing parts in the future. Next I decided to try and make a set of fin blanks for my windsurf board. fin diagram

The picture shows the bottom of my windsurf board. It has 5 sockets at the back of the board that you can put fins in. The board can be used with 3 different possible fin set ups. This means that not all of the sockets always have fins in: e.g. If i was using the board in a three fin (thruster) configuration, two of the finboxes would be empty. This is a problem because when water is passing over the back of the board, these voids will increase the amount of turbulence in the water and increase drag. This decreases the acceleration, top speed and maneuverability of the board. I am going to make a set of 5 inserts that can go in the finboxes to plug up these gaps and reduce the drag.  fin blank design

I designed the finblanks in fusion: they are a pretty simple shape that I got by measuring the length of the finboxes on the board as well as the shape of the top of the fins. The design has symmetrical recesses on both sides just like the fins do to give the screws that hold them in place something to grip on. When I finished the design I exported it as an .stl file and then opened this in Cura. Cura is a software that takes various 3d file formats and then allows you to decide settings about speed, layer width, scale etc. and then exports a g.code file that your 3d printer can use. finblank image

In cura, I turned off the build-plate adhesion because this model is very low and flat and there is no risk of it tipping over. I set the material to Generic PLA in the settings area. It is possible to have a preset for a specific brands filament but my SUNLU filament was not on the list of presets and the generic settings have worked fine for me so far. I also am able to set the density of the print which obviously effects the time it will take and the weight of the part. I put this at 40% because the part is not structurally important and I wanted to save myself some material and time. I also set the layer height to 0.2 mm. I found that by changing the layer thickness you can greatly reduce or increase the run time depending on how much precision you need. Because there are no complex, small features I set this to 0.2mm but I could have definitely changed this to o.3mm to save some time.bottom of the board

3D Scanning

I tried to do some scanning using my phone (samsung s8), there are several apps on the app store that use a process call photogrametry. This is where the device takes a series of photos of an object  from different angles all around it. The app then uses an algorithm to stitch these photos into a 3d model. I tried to scan a few things including a portable speaker and a water bottle, however, I didn’t have any success actually exporting this or doing anything interesting with the scan. However, my Mum has one of the recent iPhone models which is equipped with a face ID sensor that uses a different scanning technology. I found this tutorial for the app scandy pro that I found really helpful to get better scans.

I used the app scandy pro which is able to use the face id sensor in the front of the iPhone X to project a laser mesh onto whatever is in front of it. This means the scan you get from the app is a point cloud which can be exported as a .obj or .ply file to a software like mesh lad to clean it up. I also watched a good tutorial on clean up of a point cloud mesh (.ply file) into a .stl which you can then export and 3d print. However, my home computer didn’t have the processing power to do this kind of clean up and just crashed whenever I tried to open the program. Therefore I wasn’t able to do this section of the process but I do feel comfortable with the basics of it.

Additional 3D printing

As part of my ongoing hydrofoil project I decided to try and remake the front wing using the CAD from week 2 to make a 3D printed mold. Because my printer has a bed size of 210 x 210 x 205, I had to split the cad model up into 8 pieces to be able to print it.wing pieces

I played around with the settings in cura to get the mold to print in a reasonable amount of time. All in all, I had to print it in 4 installments each taking an average of around 27 hours. I had to be especially careful to level the bed before each print because the sections were very long and thin and so it was crucial that the first layer stuck to the bed really well.nose pieces

wing centres

I printed the parts with 60% infill because they were large areas and I wanted to save some time and material. I chose the tri-hexagonal infill pattern  because I thought it would give better strength to weight performance.

cross section

Once I had printed all the pieces, I assembled them on a wooden board and used screws to hold wooden blocks applying pressure to the pieces. I also had to use some fiberglass and epoxy resin to keep all the mold pieces of the the top together and in the right shape.front molds

Once the mold was ready, I lined it with clingfilm so the epoxy didnt stick to it and then layed up carbon fibre and fiberglass to make the wing. I have found it works best to use a layer of fibreglass on the outside, Then twom or three of carbon, and then fill the core with fiberglass (and a 3d printed core). This seems to make the final product easier to sand (because the other layer is glass not carbon) and also strong and durable.front wingtrimmed and sanded wing

I used lots of weight to keep the two halves of the mold pressed together and this was acheived by doing all the wet lay up and then piling loads of bags of compost and bricks on to of the mold. Having lots of force on the two molds will ensure a tight  bond  and  this  will  help  prevent  delamination  between  layers  during use.mold with weightnew wing

Link to the CAD file: https://a360.co/2yreILW

Link to G.code for 3d printing: https://drive.google.com/drive/folders/1_3XBFZjkt4yyaBJYcbXfNj0cl_3gAhhj?usp=sharing

Print Friendly, PDF & Email