Characterization of the pressure sensor 66 Characterization of the bending sensor 69 5.7 Functional components of the Pick-and-Place system 71. 3.2 (a) Simulation of the electric potential distribution in the triboelectric layers of water and PDMS using COMSOL Multiphysics. b) Manufacturing process of soft pressure sensor.
Background
Studies have shown that the mechanical compliance of the TENG with the soft and irregular surfaces can be improved by building it with stretchable materials (Pan et al., 2020) such as Polydimethylsiloxane (PDMS) and Ecoflex. Furthermore, the microporous in the triboelectric layer can also help to trap more triboelectric charges (Ma et al., 2018).
Problem Statements
In the soft robotic gripper, the compatibility of the sensor with the soft structure is essential to prevent damage to the grasped objects and to improve the grasping performance (preventing the object from falling). Therefore, it is imperative to explore a softer material to develop a triboelectric-based stretchable pressure sensor for wearable devices and soft robotics applications.
Research Questions
The use of a Kirigami-based structure could improve material elasticity by lowering Young's modulus and is widely applied for sensor design (Park et al., 2019;). The triboelectric sensor adapts to soft and irregular surfaces, which can be applied to wearable devices and soft robotic applications.
Dissertation Outline
Studies have shown that the mechanical compliance of the TENG with the irregular surface can be improved by constructing it with stretchable material (Pan et al., 2020), such as silicone rubber, or amorphous material such as water. The end of the chapter demonstrates the wearable human-machine interaction (HMI) pressure sensor for controlling the light intensity and game character.
Self-Powered Technology
The TENG provides high flexibility of material selection (Wu et al., 2019) and can build a stretchable platform with simple fabrication methods. In addition to the mentioned advantages, TENG generates higher output (about twice the power density) than PG (Ahmed et al., 2020) based on energy output per metric unit of mass.
Introduction of Triboelectric Nanogenerators (TENGs)
Single Electrode Mode
As shown in Figure 2.5, the aluminum acts as one of the triboelectric layers and the primary electrode. In contrast, when the FEP layer contacts back to the aluminum layer (Figure 2.5. c)-(d)), the charge will be transferred in an inverse manner.
Freestanding Triboelectric Layer Mode
Comparison of the four operation methods of TENG
Modification to improve TENG’s performance
Physical modification, such as introducing microstructures on the surface of the TENG layer, has been proven to be an effective method in increasing the contact area of triboelectric layers. However, these surface type modifications are prone to wear due to contact friction between the triboelectric layers. The performance of the TENG is closely related to the materials listed in the triboelectric series.
One of the major limitations of the TENG sensor output performance is the air strike effect.
Stretchable Materials in TENG Sensor
However, some literature reported that it is necessary to maintain significant humidity conditions for polymer-based TENGs (Baytekin et al., 2011). It helps to maintain a layer of water on the surface of the polymer, where charges can be stabilized (Baytekin et al., 2011). It limits the ability of the TENG sensor to maintain its charge density when operated in a vertical contact separation mode (Zi et al., 2017).
A promising solution to solve the leakage problem is to embed EGaIn into Ecoflex (Park et al., 2021) before it solidifies to create stretchable electrodes.
Artificial Intelligence (A.I.) in TENG
The random forest (RF), on the other hand, is an algorithm that works by combining multiple decision trees to increase the accuracy of the thematic maps generated. Moreover, the RF is one of the most widely used learning algorithms (Belgiu and Drăgu, 2016) due to its robustness against overfitting and ease of parameterization (Kavzoglu, 2017). A comparison of the performance of the three classifiers is presented in Table 2.5 (Kavzoglu et al., 2020).
After training, the developed system provided 83.6% accuracy in recognizing different machine operating conditions.
Overview
Design and Working Principle
The pressure sensor operates on the basis of contact electrification during the PDMS-water interaction to produce triboelectric output. The positive charges balance the negative charges of the PDMS surface in the DI water. It creates compressive pressure to move the DI water flow from the chamber to the microchannel (Figure 3.1(b)-(ii)).
When the external pressure is released, the DI water in the microchannel is siphoned back into the chamber.
Numerical Simulation and Fabrication of Sensor
The simulation showed that the microfluidic-based TENG could replace the spatial separation distance in a conventional TENG pressure sensor with a charge transfer mechanism between the liquid-solid interfaces. Then, the PDMS solution was degassed in a vacuum chamber and allowed to cure at 75 °C in an oven for 60 min. The fabricated PDMS microchannel was removed from the 3D printed mold and a closed microchannel was formed by bonding it with another 0.1 mm thick PDMS layer using some ARclad® IS-8026 silicone transfer adhesive.
Finally, a pair of interconnected electrodes was formed by etching a copper-clad polyimide flexible printed circuit board (PCB) (AP9121, Pyralux Polyimide Film, DuPont, Singapore).
Results and Discussion
To suit the purposes of the HMI, the output voltage of the soft pressure sensor was obtained using an NI Elvis plate (NI-ELVIS Series II, USA) for better correlation with the input pressure. As shown in Figure 3.4(a)-(i), the repetitive triboelectric output at each operating frequency indicates that the output performance of the pressure sensor is reliable. The magnitude of the triboelectric output can determine the pressure applied to the soft pressure sensor.
As shown in Figure 3.6(b)-(ii), by touching the pressure sensor with a greater pressure, the ninja jumps from 16 mm.
Summary
Overview
Design of SBTENG
The SBTENG consists of two triboelectric layers, human skin (positive triboelectric layer) and EGaIn-Ecoflex composite layer (negative triboelectric layer). When the human skin presses the SBTENG surface in the normal direction (shown in Figure 4.1(b)(ii)), electrons are transferred from the human skin to the EGaIn-Ecoflex composite layer and form a positively charged layer on the human skin surface. Based on the triboelectric series (Zou et al., 2019), silicon rubber is more electronegative than human skin.
The human skin layer is set to zero charges as the reference layer in the simulation.
Fabrication of Sensor
The porosity of the structure is highly dependent on the volume of the NaCl particles, according to the porosity equation. The fabrication steps were repeated with 0% (as control) and 65% NaCl concentrations, to study the dependence of the output power on the pore concentration. Another experiment was performed to determine the electrical output performance of the sensor with different porosity.
To further explore the stability of the sensor, the output voltage of the sensor is increased by 65%.
Summary
The circuit connection for the demonstration is shown in Figure 4.5 (b), consisting of 5 LEDs, the SBTENG and a 50 cm × 50 cm copper plate as a reference electrode. Its dependence on porosity was characterized and it has been numerically proven that the potential difference of the structure with a porous base is higher than that with a non-porous base. Current power density is more important than reported porous polydimethylsiloxane/lead zirconate energy harvesters (Ma et al., 2018).
Finally, the developed stretchable TENG can power an LED by harvesting energy from hand movement.
Overview
Design of Sensors and Soft Gripper Pick-and-Place System
The fingers of the soft gripper are controlled by the same valve connected to a custom motor-driven pneumatic actuator. The mixing of EGaIn into the Ecoflex layer provides electrical conductivity to the sponge-based Ecoflex layer and transfers triboelectric charges. As a proof of concept, three basic symmetrical shapes of objects, including cylinder, cuboid and pyramid prism, were used as the pick-and-place product (Figure 5.1 (e)).
Device Fabrication
Each of the fingers was attached to the chamber patterns of the surface with two bend sensors. It is placed between the intervals of the two chambers to detect the angle change on the tip and middle part of the finger as it bends. For the pressure sensors, the conductive Ecoflex and copper-clad polyimide were cut into 1.5 cm x 1 cm, encapsulated in an Ecoflex mold and attached to the tip of the soft fingers.
Working Principle of Sensors
For the pressure sensors, the separation distance between the triboelectric layers will be reduced by the input pressure from the target object's surface. To have a more quantitative illustration of the potential differences between the triboelectric layers, a finite element simulation was performed using COMSOL Multiphysics (Figure 5.4). As mentioned in the previous section, the separation distance between the triboelectric layers increases when the fingers bend and induce (state "B").
The simulation shows that a gradient in the charge density produces a flow of electric current according to the separation distance between the two layers.
Results and Discussions
Time response of the triboelectric output at different deflections on the pressure sensor ranging from 50 kPa to 275 kPa in steps of 25 kPa. a)-(ii). The repetitive triboelectric output at an operating frequency of 1 Hz shown in Figure 5.5 (a)-(ii) demonstrates the consistency and reliability of the pressure sensor in sensing an applied pressure of 150 kPa. To further investigate the stability of the sensor, the output voltage of the tip sensor at different bending angles was measured for 1 hour, 2 hours, 168 hours (7 days), and 336 hours (14 days).
As shown in Figure 5.7 (c), the average output of the sensor has been found to be consistent within the 14 days, ensuring good stability and reliability.
The Functional Components of the Pick-and-Place System
As shown in Figure 5.9, the output signal from the hard surface shows a higher peak than the soft surface in a grip and release action. Our TENG sensors are attached directly to the soft gripper, with their positions indicated in Figure 5.10 (a). The signal collected from the nine sensors for the three objects is visualized in Figure 5.10 (b), and the data length for each sensor is 70.
The confusion matrix in Figure 5.10 (c) shows a high accuracy of 91.76%, which is higher than several reported studies with only approx. 84-90%.
Conclusions
Therefore, to extend the performance of the soft pressure sensor, it is necessary to improve the sensitivity of the pressure sensor. After proving its ability in sensing human pressure, the mushroom-based TENG pressure sensor is further applied to future smart manufacturing. The bend sensor is made using the same approach and embedded in the joint of the soft gripper.
The bending sensor can detect the bending angle of the soft grip (up to 60 °), while the pressure sensor works to detect the gripped object.
Future Works
Single electrode triboelectric nanogenerator based on economical graphite coated paper for waste environmental energy harvesting. Triboelectric nanogenerator based on harmonic resonators as a renewable energy source and a self-powered active vibration sensor. Liquid metal electrode for powerful triboelectric nanogenerator with an instant energy conversion efficiency of 70.6%.
An electrospun triboelectric nanogenerator based on nanowires and its application in a fully self-powered UV detector.
