The base structure was inspired by and built using Jennifer Jacob’s Codeable Objects library in Processing. The structure is approximately 20 inches in height, and about 12 inches at its maximum diameter. This went beyond the parameters which the codeable objects library allows you to use, but with some tweaks I was able to achieve the my desired size. Here is an image of the resulting 3D design as well as the 2D parts that were laser cut:

The base structure took quite a bit of trial and error to laser cut. At first, the design was too large to fit in the laser cutter. Also, the notches on the top and bottom pieces were too deep but after some adjustments and experimentation, I was able to laser cut the pieces correctly.

The 36 white LED lights were attached and wired to the lamp through the use of solder and copper tape lining the ribs of the base structure .  The proximity sensor was also wired using copper tape as well as some insulated wire. Both the sensor and LEDs were wired to the LilyPad Arduino board using copper tape and solder.

Initial test of the lights after the Arduino was attached:

Attaching the laser cut flower shapes:

The finished product!

Link to Presentation: https://www.dropbox.com/sh/034kcsgpomeje1d/yZrIpkNK13/FinalPresentation.pdf

Link to the Arduino code: <tbd>

I also have purchased some metallic silver cardstock to use for the flower structures.

Here are samples of what I tested out using the Arduino, conductive thread and a test flower:

Instead of this I have opted to create a spherical base structure now with the help of Jennifer’s codeable objects program and use copper tape to attach the LEDs. I  will be attaching a flower structure similar to this but will not be using a hanging lightbulb:

This is the base structure I created using the Codeable Objects:

I have started to test out laser cutting the parts for the spherical structure. I wanted to start small before I laser cut the real pieces. I also worked with Jennifer to find out what the maximum size I could achieve with the codeable objects parts would be and she supplied me with some updated code that allows for larger components.

Here is an image of what I tested out using the laser cutter:

I also learned how to use the blade saw in the lab so that I can work with a larger piece of wood.

My plan for then next few weeks:

• Cut the base structure using the laser cutter
• Laser cut the flowers
• Attach all the flowers and LEDs to the base.
• Program the LEDs using Arduino

Due to the way the code was written for this program, it ended up that I was not able to generate this file as a pdf to preserve the vector graphics.  After spending a lot of time with the “record” functions in Processing to try to do this, it was determined that it was probably the renderer that I was using that was preventing it from being recorded as a pdf.  I also tried to do a live trace of the image in Adobe Illustrator but had no luck in preserving all the lines.  In any case,  I was told eventually that this was a fairly complex design and it was probably not going to work out well on the embroidery machine afterall so I moved on to a new design.

I decided to simplify my design and do something completely different. This time I opted to created a voirnoi diagram out of a grid pattern and spirals.  Here is what I created by modifying some of Jennifer Jacob’s voronoi code:

I uploaded this to Corel Draw and while there I decided to fill in a portion of the pattern with the intention of using a fill on the embroidery machine. I also cut out a part of the image to save some time on the machine.  I then exported it as an .ai file, then imported into the Drawings 4 software.  I decided to use a triple running stitch for all the lines and used green thread. I also set it so that the fills were different for several portions of the pattern. I attached a piece of linen fabric and interfacing to the embroidery hoop, attached it to the machine and started the embroidery process.  For some reason though (I think it was due to a glitch in the software where “none” was not deselected on the stitch menu)  my fills did not end up getting filled in.  Here is my finished product for this design:

I also experimented with a different design. This time it was a pentigree L System design as seen below:

I followed the same process as mentioned above to attach fabric to the embroidery machine and start it off. This time I only did a single running stitch so the lines came out much thinner and I did no fills.

Here is my finished product.

While I thought these designs turned out nice, I was still not satisfied and wanted to create one additional design that I had thought about.  I have been fascinated with the fact that voronoi diagrams are seen in nature such as on a giraffe’s spots or a butterfly’s wings so I wanted to create an outline of a giraffe and fill it in with a generated voronoi digram.  So that is what I did.  :)

Here is the voronoi pattern that I created which has a good mix of small and large shapes. This was a a grid pattern layered with spirals using the voronoi code.

I then found an image of a giraffe online and deleted everything except for the outline of it using Adobe Illustrator. I then layered the voronoi pattern I generated on top of it  and cleaned up the edges.  I then filled in some of the spots with the intention that they would be filled in with a satin stich once I got to the embroidery machine.  Below is the resulting image.

Once in the Drawings 4 software, I set all my fills to be a satin stich with different patterns so that I can see a variety of stitches on the resulting embroidered image. I left everything else on the default which apparently came out to be a satin stitch for most of the edges.  I used all brown thread for this. One thing that I discovered as my pattern was getting embroidered was that I had uploaded my image with a couple different colors in the image so the software interpreted this as separate parts of the pattern, and therefore somewhat inefficiently embroidered from one edge of the giraffe to the next leaving me with a lot of extra threads I had to cut off in the end.  Overall though I was happy with my pattern. The work in progress and final result is seen below.

Link to all Processing Code.

I can imagine having multiple of these peaks and valleys in a larger piece would make for a very interesting texture.

Here is my completed piece for part 1:

For part 2, I decided to knit an image pattern using the DesignAKnit software.  I used a pattern that was already in the DesignAKnit library of patterns and contained images of flowers that looked like this:

Here is what the entire pattern looks like:

I chose this pattern because I knew that I would be using the fairisle setting on the knitting machine which knits only on one side and carries the yarn over on the other side, so I wanted something that would gradually increase and decrease the switching of colors in the pattern to decrease my chances of dropping a stitch in the process.

I loaded up this pattern on the software and chose to have the swatch 30 needles wide and 100 in length.  After casting on the first color of yarn (white), I added the crimson colored yarn in front of it on the carriage.  I started up the DesignAKnit program and starting knitting, making sure I was on the “F” setting on the machine and in the “triangle” position for the levers on the side.  I watched as the software kept track of which colors it was using for each row and how it switched off. Amazing!  After I knit the first few rows, I added weights to the ends of the knitting and then made sure to move the weights up throughout the process to prevent the machine from dropping a stitch.  There were a few times that some stitches were dropped, so I tried my best to fix them before moving on to the next row but inevitably the finished piece still contained a few dropped stitches that I did not see until the end.  I could have possibly prevented this if I watched each row more carefully and adjusted the needles or perhaps adjusting the tension on the machine may have helped so that the stitches were not as tight, which makes it hard for the machine to catch on to them each time.  In any case, I ended the pattern by knitting a wide stripe of the crimson colored yarn and a thin stripe of the white  yarn which made my piece look like a pattern for a sock. I made the mistake of keeping the machine on the F setting during this part, so the opposite color yarn was carried over in the back. I fixed this by cutting the extra yarn at the end and tying a few knots.

Appropriately, my swatch happens to be the Harvard University school colors.

Here is the finished product:

This pattern is interesting because the gradual shift of colors in each row makes it very suitable for the knitting machine since it is not jumping around and switching colors too frequently.  The continuity in the colors makes it so that the pattern looks very sleek and organized.  I played with switching colors manually at the end of my pattern and noted that it took slightly more effort to get this done right than having to deal with one color when knitting without the software. Looking back though, I think a nice effect would have been achieved if I flipped the colors of my yarn halfway through the pattern which the software was controlling so that I could see the opposite effect on the pattern or maybe could have used completely new colors.

For my final project I would like to create a Tord Bootje inspired chandelier.  His beautiful work with lights have really sparked my interest and have inspired me to take on this project idea.  This will be suitable to a textile based project as I plan to computationally design and laser cut a pattern that will cover the main body of the chandelier. I am considering using either card stock paper, linen fabric or similar material for the pattern and I then plan to line the chandelier with LED lights that will be programmed with the LilyPad Arduino.   Here are a couple of examples of Boontje’s work that have inspired me. I’d like to incorporate color in my design as well. This can either be in the form of different colored lights or different colors of material.

Here is a brief timeline for my planned work:

April 11th – 18th: Research materials, computational patterns, and plan out construction of chandelier

April 18th – May 2st: Generate computational pattern using processing, test on the laser cutter, and find out how much material I will need.

May 2nd –  16th: Start and finish construction of chandelier, begin inserting lights.

May 16th – 22nd:   Program the lights using the Arduino and document process.

________________________________________________________

[OLD]

For my final presentation I want to combine concepts I have learned from various projects, specifically the computational design assignment, the color changing textile assignment and using the LilyPad Aruduino to program the behavior of lights.

For my project I would like to program a clock that you can hang on a wall that has different behavior based on the time of day. For the base of the clock I would like to create a lace pattern using Processing and then either etch or laser cut this on wood.  I will then attach conductive materials to this so that I can put LED lights on it to represent each minute (or hour) on the clock or paint this on using thermochromic ink. The lights will be attached to the LilyPad Arduino so that I can control when the lights turn on.  The clock will display the correct time through the lights but will also change color depending on if it is daytime or night time.  (e.g. yellow for daytime, blue for nighttime).

Link to Presentation: FinalPresentationProposal_GracieElqura

I then realized that for my belt, I would have to scale down the design so that it would fit the dimensions of the belt. I measured the belt and modified the pattern accordingly by adjusting the size in the code.

This is what the design looked like:

I also decided I wanted to create more than one design so I can laser cut both on the belt to see what looks better. So this is a spiral pattern overlaid on top of a grid pattern:

Once at the laser cutter, I realized I would have to work in pieces because the belt was too long to fit in the laser cutter.  Before even starting on the belt, I tested my design on a piece of cardstock to make sure I had the dimensions correctly and that the design turned out how I wanted.  I also tested a portion of the design on a piece of scrap leather so that I could determine the correct speed and power settings for the laser cutter. I started the laser cutter with very a very high speed and low power and noticed that it only etched the leather. I then opted to decrease the speed significantly because I knew the leather would need some time to burn away. I adjusted the settings to 10 for speed and 100 for power and ultimately settled on 20 speed and 90 power after testing out a few pieces with the leather.

Test pieces:

Finally I placed the belt in the laser cutter, made sure it was aligned, and started it up. The spiral pattern was going to take up about half the belt. Once this was done, I moved the belt down so that I could continue with the second design and laser cut the grid pattern that is shown above, on the second half of the belt.

Here is a comparison of the design on the cardstock and the result on the belt:

The laser cutter burned through the leather in most places, but there were a few places it did not. This was probably because the belt may have moved a bit while in the laser cutter because it was hard to fit in there. Also in certain areas it was not aligned perfectly so it tried to cut through the part of the belt where there were seams which probably made it harder to cut.

Here is my finished product!

As a final project, I laser cut the Voronoi diagram I designed initially for this project in cardstock:

Here is a link to my Processing code.

I chose to use Twizzlers, Nerds, and Dots (Gum drops) as the candies i was going to cast and wanted to make some fashionable bracelets/bangles out of them. To get the right shape I first laser cut a mold out of acrylic that was about 1/2 inch thick.  To prevent any leakages in the mold, I bonded another piece of acrylic behind it so that the mold had a back to it.

I then poured silicone into the molds so that they were filled about half way, placed the candies in, and then poured more silicone on top to make sure they were fully coated. The Twizzler bracelet was a bit more difficult because of its texture so I had to make sure that I was covering all spaces in the mold.  I had some extra silicone so I cast a Swedish Fish candy and more nerds in a small rounded dish to make a pendant that could be worn on a necklace.

I let the silicone cure for a few hours, then pulled the bracelets and pendant out of the molds. I trimmed off any excess silicone using a sharp knife.  Unfortunately I damaged the end of one bracelet while pulling it out of the mold but overall they turned out looking great.

Here are some pictures of my process and the finished product.

I started out prototyping on one piece by attaching the nitinol by sodering it to conductive fabric. I then tested this by using alligator clips and attaching it to my battery to make sure that I can achieve the movement that I was looking for in the square of felt. Based on the properties of the 0.006mm nitinol, since it needs ~410mA to achieve 1 second of contraction and I was using a 3.7V battery, I calculated, using Ohm’s law, that I needed approximately 9Ω of resistance in my connection from the nitinol to the power supply.  The resistance of just the nitinol and conductive fabric was only ~5Ω so I added four additional 1Ω resistors to the connection.  I tested this, and I was not achieving as much movement as I would have liked so I removed one resistor and it seemed to move better. I recalculated the total resistance at this point and it was actually closer to 9Ω while before it was closer to 9.8Ω.  I repeated the same on the other square but used all four resistors because my measurements using my multimeter indicated I needed slightly more resistance. Perhaps my nitinol was a tiny bit longer.

I connected both squares to the mosfets and then connected them to separate Arduino pins.

Finally I programmed it so that the pink and purple “kites” alternate in their movements every few seconds and speed up in their alternations. They then progress to slow down again.

Here is my work in progress:

I added some accents (bows) on the kites to make them look more visually appealing and I also cut off some felt from the sides of the pink kite to see the effect that reducing the weight had on movement (moves up slightly higher).

Here is the finished product and accompanying video:

Calculations:

Required resistance: 3.7V/0.41Amps = 9.02Ω

I experimented a lot with several different projects before settling on this one and learned quite a bit during the process. One thing I learned was the need for a good plan for the circuitry so that it would be organized and not messy when making the paths to the LilyPad.

Here is my plan for the circuitry for my puppet (while laid flat):

After I created the bend sensor, I made sure to measure its resting resistance using a multimeter to make sure that the resistance would decrease when I bent the sensor.  At first, my measurements did not return a very high resting resistance so I added an additional piece of velostat so that the sensor would be less sensitive and return a higher range of resistances when bent.

As I sewed the circuits on the felt, I made sure to measure for continuity as I went along to make sure everything was connected before I got too far. I also wrote a few test programs in Arduino to test out the connections to LED lights.

I sewed all the circuits in place and this is what it looked like before I assembled the puppet.

I then assembled the puppet and added the hair with the stroke sensor.  Since this sensor would be digital I had to make sure that it was assigned digital pins on the LilyPad. I also had to create a “ground pin” in my program since I could not reach the Negative pin without crossing paths with another circuit.

I finally then programmed the LED lights in Arduino so that when the mouth is wide open, the LEDs light up in a “chasing” pattern. When the mouth is half way open, the lights slow down, and when it is closed, the lights turn off.

When you stroke the hair on the puppet, the LED on its head lights up when the conductive threads touch each other.

Here is my finished product and accompanying video:

Measurements: (Note: my calculated measurements did not seem to line up exactly but this could be because the resistance of the resting position of the bend sensor was taken when it was outside of the puppet, but voltage was measured while inside of the puppet. The curvature of the mouth most likely made it so that the resting resistance was slightly less when inside the puppet, thus throwing off my predictions).

Measured:

Calculated:

Measured:

The thermochromic ink requires at least 100mA of current to heat up to change the color of the ink, so given that I was using a 3.7 volt battery, using Ohm’s law I calculated that the resistance must be at most 37Ω. Therefore I tried to find a way to make the resistance in the conductive thread as low as possible. I discovered by using a multimeter that the thread is less resistant if you double the thread over. So I decided to sew using 4 layers of thread (doubled over twice), just to be safe.  I did a couple of tests with a few pieces of thread using a multimeter just to be sure that it did not have a high resistance.

I began by sewing my pattern for the tree/circuit starting from the trunk and up through the branches. I at first wanted to sew the branches across (breadth first vs. depth first) instead of up and down to give it the effect of growing branches but then I opted not to because I was worried that I would not be able to avoid crossing paths with the thread which would cause the current to take an alternate path.  While sewing I made sure to measure the resistance using the multimeter as I went along to make sure it was not going beyond 37Ω.

After I completed sewing, I measured the trace from the start of the circuit to the start of the switch and got 10.6Ω. I then measured the resistance of the trace from the switch back to the battery and got 1.8 Omega for a grand total of 12.4Ω.  I measured the voltage of my battery and got 3.77V.

Using Ohm’s law, I calculated that the current should be ~304mA but when I attached the battery and measured the current using the multimeter I got 119mA.  I made sure to measure in a place where the multimeter becomes part of the entire circuit.  In any case, this would be enough to change the color of my design.

I then went ahead and painted my tree design using the thermochromic ink, mixing the blue with yellow to form green, and mixing blue with red to make a brownish color.

I measured the current again after the paint dried and it remained ~119mA

Here is my finished design and accompanying video of my tree changing color!

Summary of measurements:

• Voltage of battery: 3.77V
• Resistance: 10.6Ω + 1.8Ω = 12.4Ω
• Calculated current (Using Ohm’s law): 304mA
• Measured current (using multimeter): 119mA