the deformation and relaxation cycle [106–108]. When the sliding speed is low, the loss through the hysteric component of the friction mechanism is small. However, as the sliding speed increases, the deformation rate increases and the recovery is slow to recoup the elastic energy. The hysteric friction increases as more energy is lost in elastic hysteresis [45]. Recently, numerical simulations and analytical solutions based on Persson’s theory (Section4.2.6) show the increase in friction with sliding speed [109].
3.5
Conclusion
Touch friction is complicated and it can be affected by many factors such as surface material, surface properties, skin conditions and test conditions. An artificial finger can potentially eliminate much of the variability of real human fingers and also allow a much more automated iterative design process for touch-feel optimised materials. Expanding on the work performed in the research group on the friction aspect of touch-feel perception, an anthropomorphic artificial finger had been developed. The artificial finger is multilayered, including a bone support and two layers of silicone rubber. The rubber that makes up the cover layer was chosen to mimic the stiffness and the thickness of the epidermis, whereas the rubber that makes up the inner filler layer was chosen to mimic the stiffness of the dermis. The artificial finger was mounted on an improved friction measurement rig that is capable of measuring normal forces and friction forces simultaneously, with a load cell and a capacitive sensor, respectively. Measurements were taken on a set of aluminium and steel specimens with different roughness. The results of the artificial finger on the new friction test rig show a much-improved correlation between real human finger, with coefficients of determination (r2) between 0.5 to 0.91 compared to that of an
older roller-on-block rig (r2<0.5) [32].
The friction coefficient measurement obtained through the artificial finger is very similar to that from a real human fingertip, as evidence from Section3.4.2shows. Overall the setting of 5 mm s−1 sliding speed and 0.5 N normal force achieved the
highest r2 correlation with human finger results, while the setting of 8 mm s−1 and
1 N resulted in the lowest root mean squared errors (RMSE), followed by 5 mm s−1
and 0.5 N. A trend can be observed with the RMSE that it seems to decrease with increasing sliding speed. This is reasonable as human finger typically stroke a sample during touch-feel at more than 50 mm s−1 [99]. Normal force applied seemed to have
no effect on the RMSE, more contact forces should be tested to investigate this further.
3.5. Conclusion The results in Section 3.4show that, even with the same material, different experiment conditions produces different friction results. Section 3.4.3shows that rougher surfaces tend to have lower friction coefficients for the materials tested. Section3.4.4 looked at the effect of the presence of a fingerprint on friction measure- ments and found that the fingerprint increases friction except for the roughest steel samples tested (S4 andS5). Ridges on the fingerprint increases the apparent contact area leading for a given pressure or loading force (see Section 4.5.2), and therefore increases friction.
Analyses were performed in Section3.4.5 on the relationship between surface friction coefficients and the contact forces. It is concluded that as the contact forces increase, the friction coefficients decrease. Section3.4.6 investigates the relationship between surface friction coefficients and the linear stage sliding speeds. When the linear stage sliding speeds increases, the obtained friction coefficients increase.
Future work is required on looking at natural variability of different human fingertips and how to best design an artificial finger that can emulate these differences in a controlled manner. Increasing the capability of the apparatus to operate at higher sliding speed may improve the matching with human fingers.
Chapter
4
Measurement and Theoretical
Analysis of Contact Area
C
hapter3 introduced a multi-layered artificial fingertip design mimicking the structure, shape, stiffness and the friction properties of real fingertips in order to facilitate human touch-feeling studies. Experiment results confirmed that the friction properties of the artificial fingertip are close to that of a real human fingertip. In order to understand the contact mechanism, a suitable theoretical model of the friction mechanism, in particular, how the normal force and contact area relates to the friction perceived by people is essential in understanding touch-feel perception.4.1
Introduction
There is a growing interest in quantifying and modelling touch-feel perception over the past decade in order to understand customers’ needs in product design [4–12]. Touch-feel perception can only be evaluated when a person performs skin contact or strokes over the surface of an object. Designing a surface material for desirable touch-perception is often left to trial-and-error and hence a time-consuming process. From Chapter 3, an artificial fingertip can replace a human fingertip for friction characterisation with much higher repeatability (precision), controllability and lower turn-around time. Identifying the relationship between the friction measurement and touch-feel perception is not enough—a designer can only design a material against material properties quantified by surface topographical parameters and mechanical parameters. Hence, it is essential to first identify the links between the friction
4.2. Contact Mechanics theories