On the limitations of TENG testing

Titelangaben

Hilgert, Annika ; Orgeldinger, Christian ; Mattauch, Philipp ; Fischerauer, Gerhard ; Tremmel, Stephan:
On the limitations of TENG testing.
In: Amanov, Auezhan ; Juoksukangas, Janne ; Kati, Valtonen ; Lamminen, Saara (Hrsg.): NORDTRIB 2026 : Book of Abstracts. - Tampere, Finnland : Tampere University , 2026 . - S. 53
ISBN 978-952-03-4666-9

Volltext

Link zum Volltext (externe URL):

Abstract

Triboelectric nanogenerators (TENGs) are devices capable of harvesting energy from mechanical motions such as vibrations or movement of the human body. They operate based on the effect of contact electrification, typically involving either two insulating materials or one insulator and one conductor to generate triboelectric charges. Due to the continuous contact and separation of the surfaces involved, TENGs are prone to wear, which can significantly reduce their operational lifetime. To address this issue, one of the contacting surfaces was coated with a diamond-like carbon (DLC) coating to enhance wear resistance.
To investigate the triboelectric performance, a test bench was built to conduct TENG tests in lateral sliding mode. In this configuration, one surface remained fixed while the other surface performed a relative sliding motion under a constant normal force and sliding frequency. For each material pairing, three tests were conducted with a duration of 15 min, a sliding frequency of 2 Hz, and a normal force of 2 N. Prior to each test, both surfaces were electrically discharged using an anti-static frame. Additionally, the environmental conditions were monitored.
Consistent with findings from conventional tribological testing, the three repetitions per material pairing showed a high degree of variance, indicating that this number of tests was insufficient for reliable characterization. In addition, a time-dependent behavior was observed, which is due to a smoothing of the surfaces during the initial periods of the sliding process. To reduce external and systematic influences, the test bench was subsequently improved by implementing a more rigid fixation of the samples, a ball head to ensure precise parallel alignment of the contacting surfaces, and the placement of the entire setup inside a climate chamber to guarantee strictly controlled ambient conditions.
With the improved setup, more reproducible results were obtained. However, a measurable variance between individual tests remained. Consequently, it can be concluded that a large number of TENG tests is required to reliably characterize a given material pairing for sliding-mode TENG applications.
We chose to use polyimide (PI) and polytetrafluorethylene (PTFE) as triboelectric materials because both have been reported to act as negatively charged partners in the triboelectric series, although with different charge densities. For the experiments, PI foils with a thickness of 75 μm and PTFE foils with a thickness of 80 μm were used and attached to an aluminum tape with conductive adhesive. A sliding mode test bench was used to measure the short circuit current between the two aluminum electrodes due to electrostatic induction. To correlate the electric signal to the process conditions, the acceleration, normal force and friction force were recorded simultaneously. In addition, ambient temperature and humidity were monitored.
A total of 30 repetitions were performed for each material. Each individual test lasted for one hour, with an applied normal force of 3 N, a sliding distance of 15 mm and a sliding rate of 3 Hz. In addition to the long-term measurements, a full factorial test matrix was conducted, varying the applied force from 1 N to 5 N and the sliding rate from 1 Hz to 5 Hz.
In general, the results suggest a direct relationship between the short circuit current and the sliding velocity. On the other hand, the surface charge density seems to be independent of the sliding velocity, which is backed up by the theoretical model. However, an increasing normal force led to a slightly increasing surface charge density. Long-term tests showed that running-in behavior is always observable. Nevertheless, differences in amplitude and signal shape were observed between individual samples of identical material.