Degree Name

Master of Engineering - Research


School of Mechanical, Materials, and Mechatronic, and Biomedical Engineering


Artificial muscles show great potential in several applications, particularly medical prosthetics and robotics. Thermally-activated stretched rubbers and thermally-activated twisted thermoplastic fibres have not received much attention in comparison to other polymers but show promise in these applications. The aim of this thesis was to address the lack of data on thermally driven polymer fibre actuators. Two types were considered: thermally-activated stretched rubbers and thermally-activated twisted thermoplastic fibres.

Spandex polyurethane was plied with steel, copper or carbon fibre and thermally actuated by passing electrical current through the conductor. The effect of the power input, temperature, stiffness and the force applied during sample manufacture and actuation were investigated. It was proven that as temperature increases the contraction of carbon fibre and polyurethane samples increases. This is consistent with the mathematically derived thermodynamics of heating rubber which state when rubber is strained and is then heated the rubber will contract in the direction of the loading. It was also shown that as the power through the conductive component of the sample increases, the temperature increases linearly. Therefore it can be concluded that as the input electrical power increases the contraction of the sample will increase. It was found that a sample manufactured with more weight applied during plying has significantly lower stiffness resulting in higher contractions compared to a sample manufactured with less weight, and that this relation holds across all actuation weights. The actuation of the spandex occurred by contraction of both the spandex and the carbon fibre coil wrapped around the spandex, and when the carbon fibre coil was stiffer the contraction of the spandex was reduced. The higher weight during manufacture decreases the bias angle during testing which increases the spring index of the carbon fibre coil resulting in decreased stiffness. Maximum contraction is achieved with minimum actuation weight. The maximum contraction observed was 10.2 %.

Overtwisted coiled nylon samples were produced by twisting nylon with carbon fibre until completely coiled. Mandrel wrapped samples were twisted nylon only and coiled around a rod, with heating applied via a furnace. Power input, temperature, temperature change, stress applied during manufacture and actuation, fibre diameter, sample length and coil pitch were investigated. The contraction of electrothermally heated overtwisted nylon samples increased linearly as the power through the conductor, and thus the heat generated, increased. For 0.45 mm diameter monofilament nylon, maximum contraction occurs when the weight on the samples is somewhere around 100 g and 150 g. Maximum contractions occur in furnace heated mandrel wrapped nylon with maximum temperature changes. Maximum contractions also occur with minimum coil pitch and sample length. The largest contraction observed was 9.6 %.

A variance in twist upon actuation was observed with shorter samples resulting in larger heatinduced untwist. This is unexplainable based on the current literature and future work should be carried out to investigate this. These experiments are unique in that they investigated larger coil pitches. This may fall outside of the coil contraction regime wherein untwist remains constant, or more likely the variation of twist with sample length is a result of variation in manufacture of samples. Contractions around 5 % of the initial length were observed, with a maximum of 5.1 %. This is less than other results using smaller pitch coils, however these results confirm that smaller pitch coils produce higher contractions.

The ability to store nylon samples is simpler than stretched rubber as nylon samples do not require constant application of force, allowing work to take place in stages rather than a continuous process. Although the manufacturing process for nylon samples takes much more time, it is evident that the results are repeatable and there are opportunities to investigate further to gain larger contractions. The contraction achieved for stretched polyurethane was larger than nylon; however work carried out by others has resulted in larger strains of 33 % compared to 28.5 % for latex. The simplicity of nylon actuator manufacture coupled with the low cost and high strains position nylon as a high potential material for commercial artificial muscle applications.

FoR codes (2008)




Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.