Goal
The goal of this assignment is learning how to use with the vinyl and laser cutters and also to design and make a press-fit construction kit.
Last week MIT lecture review
For the rest of the semester the first hour and a half will be dedicated to review last week assignment. I noticed that Neil insisted on a couple of things. He asked people using commercial packages software if they had tried the free open source software alternatives. The reason is that paid software has some limitations. The price is one of them. Another one is the huge community of developers that keeps the software evolving fast, much faster than paid software. What happens is that this free open source software used to be very limited in the past, limited to computer geeks I mean. But now, I don't think MS Office can beat LibreOffice features anymore. Their only weapon now is keeping changing formats so their docs can only be opened with their software. But the users are slowly saying that they had enough of this.
After the break Neil talked about cutting machines, specially the vinyl cutter and the laser cutter. I specially remember two things from the lecture. One: Laser actually cuts because it burns, so keep watching it at all the time and if necessary use the emergency stop button. I agree with him because we had two fires this week using the laser cutter.
He also talked about something I had never heard: Flexures. It looks like you can make dynamic structures by just cutting things in a specific way. I have to investigate this further because sounds quite interesting.
Important notice for Apple users: Do not mess with CUPS
While trying to send an .epi file to the Epilog Laser Cutter via Ethernet from the Fab Modules we noticed that we were blocking the Epilog ability to send and receive files until a hard reset was made. We tried setting the printer in different ways: from Mac OS System Preferences, from CUPS, all of them without success. Then I noticed that my version of CUPS was 1.6svn and I thought that maybe the problem could be caused because I had an outdated version of CUPS. Never think such a thing. What I did then is going to the CUPS website and downloaded and installed version 1.6.1, I didn't bother in reading that this was not my OS. Result: I open Chrome, it crashes. I open Mathematica, it crashes. I open any app, it crashes. Hummm, they all say the same: libcups.2.dylib is wrong architecture. Now I cannot even make a Time Machine backup, the fan of the computer is blowing at 6000 rpm and there is no way to stop it, there is a process named ReportCrash which I cannot kill because it reappears and consumes 125% of the CPU. As I write this lines I am performing a full reinstall of the Mac OS X. Later I also read that 1.6svn was newer that 1.6.1. Someone inside my head says: Did you go to school Francisco? Isn't it the very first thing they teach you, to read?
Press-fit construction kit
Thinking what to do
The hardest part of this week assignment is choosing what to do. After the lecture, in the local session, Tomas told us two kinds of press-fit we could make. It could be either a 3D model that we slice in some way to press-fit. Or it could be some sort of Lego pieces repeated in order to form modular structures. Structures and modularity are two things I like so much so I will go this way. The truth is that I would like to make a desk lamp as well. I think that this cardboard we have in the lab is a very good material to make lamps. Cardboard it's lightweight and quite rigid. But I will do it another day when I have more experience. This week I'm going to experiment with modular press-fit structures. That is actually one of the reasons why I came to Fab Academy. In a future I would like to find something that could be usable to make DIY Open Source Prefab Houses.
Castellated modular structure
I started with a design from Tiffany Tseng, a student from MIT's MAS.863/4.140 How To Make (almost) Anything in fall 2011. She made a nice design with a connector and a curved beam. I downloaded it and made some modifications. First of all I scaled the items so that the slot width measured 6.5 mm. Sara Ezz (one of my mates at the lab) did a test and found that 6.4 mm was optimal for our 7 mm cardboard. I set it to 6.5 mm just because I like round numbers. Then I added some straight pieces to the set. The lengths of the straight pieces is not arbitrary. I made them so that they reach different nodes of the 3D grid. Finally I made the set castellated -what else- by adding holes to the beams and connectors. It looks much nicer to me now. I used Rhino to make this set, but I would like to repeat it with kokopelli or OpenSCAD.
Making speed/power tests
At the beginning I was going to use the Epilog 36EXT. But when I came today I saw that Jose Pablo had an accident and a fire burned the lens of the machine. Then today (Monday, 24 hours before committing deadline) some people were doing some tests with the Spirit GE (100W). We tried different combinations of speed and power but we could not properly cut the 7 mm cardboard. Anastasia told us that this machine was not able to cut the cardboard. She was right. These are the trials we made:
- Power 85, Speed 40: Engraving.
- Power 85, speed 32: You might name it half cutting or deep engraving.
- Power 90, speed 32: This is the closest setting to cutting, bad finished though, with some parts that need to be manually finished to cut and quite burned borders.
- Power 90, speed 30, cutting a circle: fire.
- Power 90, speed 30, cutting a square: a lot of fire.
Then Luciano told us that we should use the Multicamm 2000 laser cutter. It is the big 400W laser machine that can cut you into pieces in no time. The settings I used to cut the cardboard were power 150W, speed 30. It cuts cardboard like a hot knife cuts butter. It doesn't even leaves black marks on the edges.
Removing double lines and nesting
One important thing to remember when making a design that is going to be cut by a laser is removing double lines. If you put two lines, one on top of the other, the laser will pass two times. We started a micro-fire inside the laser cutter because of this last day. Cardboard is very flammable. If you force the laser to pass over the same place again and again it will eventually start to burn. Also when cardboard starts burning you cannot stop the fire by blowing. If you blow then the fire will become bigger because of the voids inside the corrugated structure. Having some water near you is never a bad idea.
Nesting is placing all the pieces in the design so that you use the less possible material. In the past, I have used some programs from Vectric that have auto nesting. But I haven't found any open source nesting software. So I am nesting manually. If your pieces are regular it is easy to nest. Just follow a square array for rectangular pieces and a hexagonal array for triangles and circles. For more complex shapes I think the best is to print some of them in a piece of paper, cut them out and try manually to place the items so that they are nested. As you can see nesting is good, because very little material is unused. But what was not so good is arranging the items like I did. Continue reading to find out why.
Cutting strategy
I have to admit that I followed a bad strategy when cutting. I arranged my items so that there were all the connectors in one place, all the curved beams in another place and the straight beams at the end. After being one hour cutting someone came into the room and said: Stop the machine now because we are closing in 5 minutes. At that time I had a lot of connectors, a lot of curved beams and very few straight beams. This shouldn't happen. Lucky me, there was someone else who did not finish and I could eventually cut almost all the straight beams. If you are going to make a modular press-fit kit, your cutting strategy should also be modular. It est, arrange the items and follow a cutting order that makes you cut an entire set in batches. And then repeat the batches as many times you need. This way if you run out of time or you need to stop the machine, you won't waste the rest of the material and also you will have whole sets of items to play with.
The result
The best feature of modular structures is that you don't even have to know what you are going to build with it. You just start assembling pieces and then you can make a lot of shapes and elements. Below there three of the infinite possibilities of the construction kit.
Also when you are tired of assembling structures (if you are like me you could be doing this for hours) you can use your press-fit pieces as chips in a poker or casino game.
Improving the press-fit kit
If someone wants to continue this work, here you are a list of things that could guide you in improving this kit:
- Modify the ends of the beams so that they are slightly chamfered. I found that sometimes it's a little bit difficult to insert the pieces.
- Create connectors for 45º and 60º and their associated beams. Remember that you always should end in another node of the 3D grid or sub-grid.
- Create A 3D connector.
- Modify the connector so that it can be used as a base for a beam as well.
- Engrave the different parts so that you can identify them better.
Vinyl cutting
Design
For the design I went to The Noun project (a collaborative icon project) and I downloaded a couple of vector images. Since it is Carnival this week in my town, the first icon is a mustache designed by Ade Harnusa Azril and the second one are 3D glasses designed by Fabio Grande.
Setting the Roland GX-24 under CUPS
You need to setup a raw printer named vinyl in order for
the Fab Modules to send the .camm file to the Roland Cutter. Go to
your browser and type http://localhost:631
which is the address of the CUPS administration panel. You will be
presented with a screen with some tabs. Connect the Roland GX-24
with the USB cable to your computer and go to Administration tab and
then click Add printer. Then setup a printer with the
following settings:
- Name (this is important!): vinyl
- Make: Raw
- Model: Raw Queue (en)
Result
As you can see in the pictures below, something went wrong. Two things at least. First thing: It was very difficult to remove the background because the blade did not cut all the way through the vinyl. I can fix it next time by adjusting the blade length and force and making some test before cutting the piece. It doesn't worry me because it was caused because of lack of time, which I can control.
The second thing is worse. If you look at the fab modules picture below you will see the png at the left and the path at the center. Notice that the path lacks some lines (marked with red arrows). At the beginning I thought that it was a rendering problem, but when I saw the pattern cut like the path I knew something was not converting well. It might me be resolution of the png (is was a 400px width image) or maybe the image needs a white border in order to be recognized. I need to do more tests. I had to manually cut these voids with a sharp blade. But the results are quite poor, as you can see.
Fablab Ink
I also want to laser etch my laptop so that I can remember the Fab Academy experience. I don't know if I will put the digital tattoo in the case bottom or in the back lid. Please find below the idea I have, this is not a final design, laser etching last forever and I have to carefully plan it, but should resemble something like this:
You can download the fab lab EPS logo file from brandsoftheworld.com.
Tomas told us that it is safe, that the laser won't go through the
aluminum, it is not that I do not trust him, but given the lucky I
am being so far, I want to make some tests before I cut my computer
in pieces like a Jedy. I have a piece of aluminum from an old Apple
battery, so I will make a 3x3 array of squares and text with a
variety of speed and power. Something like this, but with the
appropriate values.
If you see no week 04 assignment for several days it means that I failed miserably.
What I learned
Having a good cutting strategy is very important. The basics: Engrave before cutting, cut the inside holes before the cutouts and finally arrange your production in batches if possible. Another thing that I will have in mind for the rest of the semester is spiral development, if you have time, cut a basic set, do some tests and refine your design before going to production. This method might be apparently slow, but in the end it will be faster and produce much better results.
Download files
You can download all the files related to this week here.