Week Three: Computer-Controlled Machining

Goal

The third week goal was to design, make and document a press-fit construction kit using the laser cutter and cardboard.

Week Three Objectives

• Think of an idea for the construction kit
• Choose a software package that uses parametric design
• Create design in the software
• Laser cut the kit
• If necessary repeat steps

What I did?

The idea for the construction kit was to design a opensource Hot Wheels track that can be put together by kids. I wanted to have have straight-a-ways, bends and rails for on each side of the track, similar to a train track kit for kids, that I see in Barnes and Noble. For the design and the software, I wanted to challenge myself, I could use Solid Works but I know SW and it would not be that difficult. Since, I liked kokopelli in the previous week I chose kokopelli, although I know it would be more work. Plus, our instructor said we use a lot of Python in the upcoming  so I wanted to par\y it forward and learn more on Python.

Kokopelli

The first question I had was about the way the coordinate system is set up in kokopelli. What is the spatial framework for kokopelli? Is the working space in quad one or three. Some programs like Inkscape and Khan Academy work in quadrant four, and at first I did see the coordinate system switch. Therefore, I wrote and ran the small program below to find out. The red rectangle is long across the x axis starting from 0 to 6 units. The blue rectangle is long across the y axis starting at 0 and ending at 6. Each rectangle is 1 unit in the orthogonal direction.

 

kokopelli code

--------------------

from koko.lib.shapes import *

r1 = rectangle(0,6,0,1)

r1.color = 'red'

r2 = rectangle(0,1,0,6)

r2.color = 'blue'

cad.shapes = r1,r2;

--------------------

kokopelli image

--------------------

The positive coordinate system is in quadrant 1. Once I did this, I found out that you can turn on the axes with a click of a button under view -> Show axes.

Questions:

 

What units does kokopelli work in?

Kokopelli works in inches, I asked Francisco and verified this when I used the laser cutter.

How do I convert inches to mm?

I could write a function, however there is a command in kokopelli which is below.

However, when I changed the units to mm, the screen did not refresh using the new units.

Why does this happen?

 

Next thought

It is interesting to me when I started modeling a press fit gauge I found what i think is a bug in kokpelli. The height of the slot wa 3 mm. It does make a difference how you input the value in into kokpelli. If the height is (3.) you get a different outcome if the height is (3).

Pressfit Gauge:

A neat littke trick I learned is kokopelli can exepct a empty variable. "variablename = None"

 

The move command is based off the local coordinate system and the rotate command seems to be based on the global system. Further investigation required.

I came up with a workflow strategy working in kokopelli which is outlined below:

    first create item on 0,0

    rotate or reflect if necessary

    then add or subtract while move command

    loop and add move to make multiple parts.

This workflow is very convenient. It allows the program to reuse the item over and over again.

Program as of 2/14/14

Don’t know why the loop is ignoring the first item in the loop? the first number in the parentheses is bypassed by the program

Click HERE to download the program. This program identified a bug for the right_triangle command, which had no bounds for X and Y axis, therefore could not be exported. Matt Keeter was made aware of the problem and was quick to respond. The bug was ficed. Below is what the program output, it should be fine to print.

http://kokompe.cba.mit.edu/downloads.html

It works. COOL!

I started research flexure design for the side or the turn and found a lot of academic papers on flexure design, which is not what I was looking for. I found out the name is typically called a living hinge. I did not find much in the slot separation, so I took an educated guess. the below illustration is

 

Below is a kokopelli view of the final layout of the track.

The overall experience with kokopelli was a learning curve for me. I got what I wanted out of it and now can use it a lot more effieciently. I modified the program with a fair amount of easy when I needed to when I had to modify the tabs and the slot. Furthermore, I was able to work through the printing problem I had with little trouble. The program seems powerful, however, there are some things which I miss from SW, one being dimensioning. I cannot tell how big the part is quickly. This can be frustrating at times.



The above image shows the laser cutter in action cutting out the test pieces for the track. The circular tabs and circular slot fit together nicely, a tolerance of ten thousands of inch was parametrically built in, however the small tab and the slot did not fit. I needed to go into the code and change the fit. When I got to the code I noticed the tolerance variable was tied to both tabs. Therefore, if I modified the number I would modify both tabs, which is clearly not the intent. So, I changed to code to take to this into account. The tolerance I tried next was twenty thousands for the small tabs. After this, I cut out another test and the press fit worked, see below image.
   
The final press fit construction kit is shown below. I tried to make a living hinge for borders, which did not turn out the way I liked. I was hoping for more bend-ability for the hinge. I believe I made the grooves to far apart. however, I also would like to try the hinge on wood to see how to behavior changes.
Fab Modules
   
Inkscape?

Vinyl Cutter

For the vinyl cutter I chose to do three different stickers, one fairly large, one with thin features, and one that was multicolor. This was I could get a feel for many different types of stickers. The experience was eye-opening to say the least. Since, I never used the machine before, I am lucky to have my Lab Assistant Chris help me set the machine and walk me through some of the basics. For example, when using the machine how to set the origin and how to load the paper. Working through the Fab modules were fairly straight forward, since I worked on them for the laser cutter they were somewhat familiar to me.

A couple of things to note, the default force was 45 grams, however, the Lab Assistant said they have been cutting using 80 grams using the final and it has been working great. (So I changed the default setting) Another default setting was the speed was set to 5 (cm/2), which I believe the units are a misprint and should be 2 cm/s (minor, but noteworthy.) I modified the 5 to 20, due to the Lab Assistant's advice.  Below are pictures of the process and my experiences with the vinyl cutter an  the sticking process.

Black and White FabLab Sticker

Once cut out I two pairs of tweezers which is a must to remove unwanted material. There was no issue or challenge at this stage.
The transfer paper is very interesting, it has more stickiness from itself to the top of the sticker than the bottom of the sticker to the wax paper. This allows the transferability of the sticker off the wax paper and onto the surface you would like to stick it onto. I never knew this.
As stated above really no supprises here, just a simple transfer, however getting the logo around the apple was a small task, which was made easy with the transparency of the transfer paper.

YETI Robotics Sticker

Right from the get go, I had problems with the thin pieces coming off the was paper and sticking to the unwanted parts of the sticker. This is where I needed both tweezers, one has a dull point to it, which allowed me to apply pressure to the thin pieces I wanted to keep. The other tweezers are sharp, which were used to cut the thin pieces away from the unwanted material. When doing this, I felt like a surgeon, although life was hanging in the balance. As one can see from the above picture one of the teeth on the yeti peeled of and was difficult to stick back to the wax paper, as much as I tried. So, the strategy was to try and fix it when applying the transfer paper.
As one can see from the above picture the strategy work better than I anticipated. Problem averted. Both tweezers were very helpful. Again no problem with the transfer.

The little piece of chest was a problem, the piece wanted to keep coming off the wax paper, in fact it did when cutting the extra material off. So I pulled the piece away from the unwanted material and placed it back on the wax paper, then it transferred nicely. Again the tweezers were the key to success here. While transfering the sticker to the place you want the sticker to be I took a small robotic wheel and went over each of the thin pieces, in this case was practically everything.
Above shows a picture of the process transferring the sticker to the final surface. One thing to note is the angle of attack when pulling off the transfer paper needs to be large and as parallel to the surface. This seems to help peel off the tranfer paper while leaving the sticker. Furthermore, it is better to go from thick to thin when peeling away the transfer paper.

Final YETI turn out better than I thought, I recovered from the mistake when I transferred the sticker onto the transfer paper I flattened out the tooth beforehand.

Resources:

Found this, which you can learn a lot from this example. I was looking for for loops.

1. https://github.com/mkeeter/kokopelli/blob/master/examples/gik.ko

Found this for hinges using the

http://www.mit.edu/~calisch/linkage/3/3.html

Inspiration for the track.

http://fab.cba.mit.edu/classes/863.13/people/robert.hart/Week2/RobertHartWeek2.html#

Living hinge

https://www.snijlab.nl/nl