robotany: April 2006
Sunday, April 02, 2006
-<><><>- Using Flexinol Wire -<><><>-

These projects use the chemical properites of certain metal alloys to create work from heat. These metal alloys are called shape memory alloys and can be created from different formulations. Wikipedia has a good article on how shape memory alloys work. The most practical shape memory alloy to use for our purposes is Flexinol. Flexinol is a trademark of Dynalloy Corporation. Flexinol is a nickel titanium alloy, called nitinol (say "night-in-awl"), optimized to contract in length when heated. Flexinol wire is rated at 1,000,000 cycles.

When thinking about nitinol and Flexinol, it is helpful to understand that its heated shape is created by forming the alloy and heating it to an annealing temperature of about 540 degrees Celsius. When it is cooled to about room temperature, the metal reaches its low temperature, or martensite, phase. In the martensite phase, the alloy can be deformed into another shape. When heated to a certain temperature (typically between 70 - 90 degrees Celsius) it reaches its austenite phase and returns to the shape it was annealed to.

Shape memory alloys like Flexinol are promising actuators. They can be heated through electrical current, are low power, silent, light weight and have a high strength to weight ratio. The force applied can be compounded by utilizing multiple wires operating in parallel. Shape memory alloys are very delicate to create and to work with. For example, nitinol is created from nearly equal amounts of nickel and titanium atoms. Differences of less than one percent in this ratio can change their activation temperature by 200 degrees Celsius. Similar sensitivities exist when operating nitinol. Nitinol must be protected from exerting too much force, being subject to too much stretching, and protected from overheating. Another thing to keep in mind is that shape memory alloys will typically be manufactured to change shape upon heating, but will not necessarily be configured to return to the initial shape upon cooling. Most nitinol requires a force to bring it back to it's initial, pre-heated state. Note that nitinol is practically non-magnetic.

Nitinol is produced in bars, rods, sheets, and springs. For spring-shaped nitinol, look up BioMetal. Flexinol is produced as wires. Do not attempt to machine or drill nitinol, as the heat generated from friction will cause the metal to react.

Properties of commercial nitinol which are important to their operation include:
minimum bend radius
strength and force thresholds
electrical resistance
recommended current level
activation temperatures

Diameter Flexinol wires range from 25 to 375 microns in diameter. The wire's diameter is directly related to its tensile strength, electrical resistence per unit length, the force it is capable of exerting, the force needed to bring it back to its state before heating (bias force), its cooling rate, and its minimum bend radius.

Strength and related forces To find the force value for your wire, multiply the given value by the wire's cross sectional area.
Maximum recovery force: the greatest force a wire can exert when heated. For nitinol, this is about 600 megapascals or million Newtons per square meter. You actually don't want to approach this value, as you risk burning out your wire.
Recommended recovery force: the force recommended for repeated use. This value is about 30% of the maximum recovery force (i.e. ~ 200 MPa).
Bias force: the force needed to extend a cooled wire, typically 10 - 20% of the maximum recovery force (i.e. ~ 30 - 60 MPa).

Recovery ratios the ratio between the low temperature (martensite) and high temperature (austenite) lengths of Flexinol wire.
Maximum recovery strain/ratio: maximum change in length of a wire. This value is about 8% for nitinol. Most wires will only perform a few cycles at this maximum.
Recommended recovery strain/ratio: recommended change in length of wire for maximum wire life. This value is 3 - 5 % for nitinol wires.
Breaking strength: the force a wire can withstand before breaking. As mentioned above, this value is directly related to wire diameter, as well as composition, temperature and processing. For nitinol, this value is about ten times the maximum recovery force, or 1000 MPa. Expect deformation and damage at about 15 - 30 % before breakage.

Electrical properties:
Resistance: typical electrical linear resistence. Thinner wires have higher resistances than thicker wires. Also remember, the higher the resistance, the more easily heat will be generated when applying a current.
Recommended current level: For Flexinol, this is a typical current level which will activate but not *quickly* overheat the wire in still air at room temperature. Flexinol engineer Jeff Brown recommends using 50 - 80% of maximum current. Use this as a starting point for your requirements. Then adjust the power level according to actual performance.

If you know or measure a wire's resistance (which varies directly with its length) and the recommended current level to activate the wire, you can calculate the voltage required using Ohm's Law.

Transition temperature: the temperature at which the wire changes from austenite to martensite phases, or vice versa.
Austenite start temperature: the temperature at which the crystals in the wire begin changing from the low temperature martensite phase to the higher temperature austenite phase. At this temperature, Flexinol begins to contract and return to its learned, annealed shape.
Austenite finish temperature: the temperature at which all crystals in the wire have transformed to the high temperature austenite phase.
Martensite start temperature: the temperature at which crystals in the wire begin transforming from the austenite to martensite phase.
Martensite finish temperature: the temperature at which all crystals have transformed to the martensite phase.
Temperature hysterisis: difference in starting and ending temperatures listed above. The temperature values listed above are different depending upon whether the alloy is heating up from a martensite to austenite phase or cooling from austenite to martensite.
Annealing temperature: as mentioned above, the temperature at which residual strains are removed and the crystal structure of the metal reforms, erasing previous training.
Thermal conductivity: the ability of a material to transfer heat energy. For nitinol, this value is approx. 0.08 watt/cm degrees Celsius at the martensite (low temperature) phase and 0.18 watt/cm degrees Celsius in the austenite (high temperature) phase.

Operation tips

Flexinol can contract as quickly as one millisecond. The time necessary for a complete cycle of contraction and relaxation depends upon the bias force and how fast it cools. Thin wires cool faster than thicker ones. For this reason, if you want to increase both the strength of the wire and its cycle speed, consider using more than one thinner wire in parallel.

Ways to cool a wire faster:

Use aluminum crimps and fittings: but a stronger heat sink will make the power requirements higher in order to reach the austenite transition temperature, and higher power levels can increase the risk of overheating, causing permanent wire damage.

Run the wire through a small diameter rubber tubing: the tube insulates the wire as it heats, allowing faster heating and thus less time power is applied, and the rubber acts as a heat sink to help the wire cool faster.

These are the things you want to avoid:

If any of these stresses occur, you will want to consider replacing the wire.
Here you will find documentation and tips on how to build ambient robots using smart materials. This work is supported by a generous grant from Belluard Bollwerk International (BBI), producers of the Fribourg Festival, an annual contemporary arts festival held in Fribourg, Switzerland at the beginning of every summer. This production of BBI is carried out thanks to a contribution from the Canton of Fribourg and the support of Fondation Nestlé pour l'Art.

Breeze, our ambient robot, will appear at the Fribourg Festival in June and early July and travel to the ISEA/Zero 1 arts festival in San Diego, USA in early August.

Robotany Team Members: Jill Coffin, John Taylor and Daniel Bauen

Acknowledgements: Zachary Coffin, machining and Tom Clarkson, blacksmithing

April 2006 / May 2006 / July 2006 /

Powered by Blogger