Sonntag, 27. Februar 2011

More Experiments with Nitinol Paper

We are accustomed to objects behaving the same way not matter what our mood or how we treat an object. A book behaves the same to someone who loves it and treats it as gentle as possible as it would to another person who hates the book, almost tearing out the pages while reading it. What if that where to be changed?

Take a look at what we are trying to do with paper:

We want to create paper, which moves depending on inputs it receives. It may shy away or close up altogether if you reach for it to fast. It may open and revile its message if left alone, or if soothed. The content of the paper itself will be projected on the inside of it. The projection will be of very graphic nature, playing with the phenomenon that one often is drawn to gruesome content and cannot look away. Here, one may be forced to look away because the paper my physically prevent you from seeing its content.

Animating the Paper

We animate the paper using Nitinol Memory wire. Memory wire has two phases, the Martensite phase at room temperature and the Austenite phase at temperatures greater than 70° Celsius. When in the Martensite phase the shape of the wire is freely deformable, when however heated, in the Austenite phase, it returns to its original shape. The shape it returns to in its Austenite phase can be set by heating the Memory Wire to over 540° Celsius.

When heated the memory wire has enough force to bend a second memory wire. This enables counteracting actuators and is the basic principle of our animated paper. We heat individual strands of memory wire by connecting them to a high powered battery for a brief moment. As we do not add resistors to this circuit, the power spike and, more specifically, the heat which is created in the process induces the phase shift in the memory wire. This setup can easily fold and unfold a piece of paper as can be seen in the above video or in the following picture: 

This mechanism can be used to create paper which can autonomously fold and unfold into simple shapes, reminiscent of folding origami.

Input for interaction

The folding and unfolding will be initiated by various different inputs. We have a variety of sensors available, microphones for measuring sound intensity and frequency, photo resistors which would react to shadows, Sharp IR range-sensors (as seen in the Theremin demo which we brought to class) or kinect for proximity sensing.  Using a combination of these sensors we will attempt to create an intuitive way of interacting with the paper. Proximity and sound amplitude may set off a folding motion when a certain threshold is reached, however it may be interesting to have interaction between the dominant frequency of the sound and the threshold levels. This may enable ‘talking’ to the paper to change its reaction to proximity, volume levels or light levels. If threshold levels have not been surpassed over a certain amount of time, the paper will unfold again. The sensors will not be physically placed on the paper, as the paper needs to be as light as possible and will most likely be prone to interfere with readings due to the temperature changes and the high amount of mA which will be sent through the circuitry on the paper.

A demo of IR range-sensors as Input can be found here:


 All sensors (except, if we use it, the kinect) will be connected to an arduino which in turn will communicate to Max/MSP/Jitter via serial connection.  The computer-vision aspects will also be handled in Max using cv.jitt.

Power supply

As the paper will be animated using a very high power circuit, we do not want to directly connect any microcontroller to it. All sensor readings will be done using an arduino, powered by a secondary electrical circuit.  The memory wire will then be activated by transistors or possibly even by relays. The primary circuit for animating the Paper is currently using a 9.6v DC power supply intended for RC cars; however for the final installation we will use a laptop power supply.  The sensors and Arduino will be powered by USB.

Tracking Projection Area

For projecting onto the paper we will use an active approach. Surface mount infrared LEDs combined with an IR webcam will be used to track the projection area. Surface mount LEDs are optimal due to their small form factor and can be placed directly onto the paper.  This is similar to the setup used by David Holman on his paper computer project. (

Set up & Miscellaneous

As the memory wire can only support paper up to a certain size, we are limited in how large our animated paper can be. It may be interesting creating a whole array or series of small objects instead of one large object. For displaying it most effectively a high contrast setting will be essential. Also, the input which activates the motion of the paper should also activate sounds, creating an ambient atmosphere.


  1. Hi. I made some very similar experimentations a few years ago, and I have a few questions :

    -You seems to use nitinol which change shape at 70°C. For quick response, this mean that you have to heat it to 80-100°C, which might be a bit dangerous for children. Why did you not use one that change shape at about 30°C ?
    -What diameter and current did you use ?
    -My next expriment will be to try to incorporate paper accelerometer in the circuit so the paper react if you touch it (paper dionea !). Whant to join ?

  2. If you can point me to memory wire which changes shape at 30°C you would be my hero.

    Please send me an e-mail paul.strohmeier(at)