Elsevier

Measurement

Volume 177, June 2021, 109252
Measurement

Design and research of a novel non-contact vertical inductive torque sensor

https://doi.org/10.1016/j.measurement.2021.109252Get rights and content

Highlights

  • A non-contact vertical torque sensor with small size and light weight is designed.

  • Analyze the nonlinearity of the torque sensor according to different rotors.

  • Establish a linear relationship between the output signal and the measured value.

  • Experimental results show that the nonlinearity of the sensor is 1.36% F.S.

Abstract

In this study, a new type of non-contact vertical inductive torque sensor was built and investigated. The sensor is small and easy to install and uses the principle of electromagnetic induction to measure the angular displacement of a rotor. More precisely, based on the relative displacement of the two rotors, the torque of the torsion bar is determined. Since the eddy-current distribution on the rotor surface is uneven, and the shape of the rotor affects the eddy-current path, four different shapes of the rotor are used for the analysis. Using Ansys Maxwell finite-element software, the change of the induced voltage of the sensor’s receiving coil, when the torsion bar rotates, is simulated. Furthermore, the effect of different rotors on the nonlinear characteristics of the sensor is analyzed. Based on the simulation results, a prototype of the sensor is manufactured. The experimentally obtained nonlinearity is 1.36%F.S.

Introduction

Torque sensors have important roles in industrial production-processes, space development, bioengineering, health care, and other fields. The methods to measure torque can be divided into two categories: dynamic measurement and static measurement. Static torque measurements use the principle of balancing the measured parts to determine the torque.

In dynamic torque measurements, the measured torsion bar is in a rotating state, and the rotation-speed changes from time to time, which makes the measurement difficult. Finding ways to measure dynamic torque quickly and accurately has become an important research focus. At present, the commonly used dynamic torque sensor can be further categorized into resistive sensors, photoelectric sensors [1], and magnetoelectric sensors. Resistive sensors include Resistance-strain sensors [2] and Potentiometer type torque sensors [3]. Resistance-strain torque sensors are simple, lighter, with a small strain gauge size. In addition, there is no effect on the operating conditions and stress distribution of the measured parts. However, because their structure is connected with the rotating shaft, dynamic balance problems can easily occur during rotations at high speed. In addition, time- and temperature-drift are large, the output signal is weak, the reliability of the product is low, and the general accuracy is not high. Potentiometer type torque sensors are small, simple, easy to install, feature a large output signal and a stable performance. However, both the brush and resistive element using the brush mechanism tend to show wear, which limits the service life and makes them not suitable for high-speed use. Photoelectric sensors include the photoelectric encoder [4] and grating sensors [5]. Photoelectric sensor have the advantages of high resolution and high precision [6], [7] but they are also complex, expensive, with limited applications in real-life environments [8]. Magnetoelectric sensors include hall sensors [9], capacitive sensors [10], magnetostrictive sensors [11], [12], [13], [14], magnetic resistance sensors [15], and electromagnetic inductive sensors [16]. Hall sensors have a large volume and contain permanent magnets to generate a predetermined magnetic field [17]. Therefore temperature compensation is needed in the application, and the accuracy is low. Capacitive sensors have a higher sensitivity and lower power-consumption [18], [19], [20] but they also have parasitic capacitances, which affect the measurement accuracy [21]. Magnetostrictive sensors have poor anti-interference capabilities and cannot be used with magnetic conductive materials. Magnetic resistance sensors have an AC zero signal, which cannot be dynamically measured at high frequency. Electromagnetic inductive sensors are mainly of the eddy-current sensor type [22], which are simpler, inexpensive, with strong anti-interference capability, and no need for temperature-compensation [23]. They have many applications in the automobile [24] and robot industry. Torque sensors have the following four advantages: 1) Low-cost: 3D printers can be used to simplify the manufacturing process of the sensor parts [25], [26]. 2) High performance: The surface method can be used to optimize the structural parameters of the sensor to improve their sensitivity [27]. 3) Wireless: Bluetooth can be used to transmit torque signals [28], wireless transmission technologies have been used to transmit torque [29], other using surface acoustic wave technologies have been achieved torque transmission [30]. 4) Miniaturization: The polyimide material can be used to manufacture the circuit board of the sensor [31], so that the circuit board of the sensor can be freely bent and wound to achieve any arrangement according to the space layout requirements.

The traditional eddy current sensor is planar [32], [33], [34], and it is widely used for angle- and torque-measurements. The applied principle of torque measurement uses the amount of deformation between two sections at a certain distance on the torsion bar. This allows the determinations of the torque based on the amount of distortion [35]. The larger the ratio of the length of the torsion bar covered by the sensor, the higher is the measurement accuracy. In recent studies, inductive sensors typically use printed circuit boards (PCB) [36], [37] to reduce the size of the sensor. PCB technology can easily control the coil layout to meet the special requirements of traditional coil manufacturing methods [38]. In the column-assist electric power-steering system (C-EPS) of automobiles, a planar inductive torque sensor is commonly used. Its function is to detect both magnitude and direction of the torque generated by the steering wheel, when the driver is steering. Planar inductive torque sensors have the following three major problems: First, PCB technology is only suitable for planar layouts, which results in a large sensor volume and difficult installation. Second, the proportion covered by the planar inductive torque sensor is limited, which leads to relatively low accuracy. Third, the rotor of the sensor is made of thick aluminum plates, which produces a large moment of inertia during rotation with the shaft. Long-term use can cause poor contact between the rotor and the torsion bar. Moreover, only the PCB is separated between the upper and lower rotors of the sensor, and the eddy-current fields of the upper and lower rotors affect each other. This reduces the measurement accuracy. Based on the above factors, a non-contact vertical inductive torque sensor is investigated in this paper. The stator and rotor of the sensor are made using Flexible Printed Circuit (FPC) technology. Compared with traditional PCB, FPC boards are light weight, thin, and flexible, which helps realize the layout for a vertical structure. Moreover, the excitation coil and the receiving coil are integrated into one FPC. The sensors are arranged along the axis of the torsion bar, which reduces the volume of the sensor. The vertical sensor covers a longer proportion of the torsion bar and eliminates interaction between the upper and lower rotors. Therefore, the vertical structures of inductive torque sensors have important advantages for C-EPS systems.

In this paper, by taking into account the nonlinearity of the sensor, the linear relationship between the induced voltage and the torsion angle is determined. The relationship between the torsion angle and torque is also linear. In other words, there is a linear relationship between induced voltage and torque. The structure of the sensor affects the performance of the sensor [39], while the shape of the rotor affects the nonlinearity of the sensor. In this paper, the influences of the rectangular rotor, circular rotor, triangular rotor, and trapezoid rotor on the nonlinearity of the sensor are studied, the optimal design is found, and the sensor is manufactured and tested.

The structure of this paper is as follows: Section 2 introduces the sensor structure and operating principles. Section 3 describes the influence of the rotor shape on the nonlinearity of the sensor. Section 4 describes the experiment, sensor prototype, and results. Conclusions are summarized and discussed in Section 5.

Section snippets

Structure of the sensor

The principle of the sensor studied in this paper is similar to that used by an eddy-current sensor [40]. The excitation coil is connected to a high-frequency AC excitation source. Under excitation, a continuous fluctuating magnetic field is generated, and the surface of the metal rotor induces an eddy current in the fluctuating magnetic field. At the same time, due to changes in magnetic flux, the receiving coil generates an induced voltage. The magnetic effect of the eddy current on the

The rotor model

The structure of the sensor affects the performance of the sensor [43]. As shown in Eq. (2), the amplitude of the induced voltage in the receiving coil is determined in two ways. One is via the coupling area between the rotor and the receiving coil. The other method used the mutual inductance between the rotor and the receiving coil. The smaller the distance from the excitation coil, the larger is the mutual inductance, and the larger is the eddy current. The lower end of the rotor is close to

Experiments and results

According to the simulation parameters, the sensor prototype was trial-produced, as shown in Fig. 9(a). The sensor stator was made using FPC technology and 3D printing technology, including a circuit module, receiving coil, and excitation coil. The circuit module includes a high-frequency oscillation signal generation circuit, a signal amplification circuit, a digital-to-analog conversion circuit, and a data-processing circuit. The excitation coil consisted of ten oblique copper wires parallel

Conclusion

In this paper, a non-contact vertical inductive torque sensor was built and investigated. The sensor used the spatial magnetic field, which was generated by the excitation coil. Furthermore, the coupling areas between the rhombus receiving coil and the rotor were sinusoidal. By collecting and processing the induced voltage of the receiving coil, the sine curve, which contains the angular displacement information was resolved. The feasibility of the operation principle was analyzed

CRediT authorship contribution statement

Chao Zhang: Conceptualization, Visualization, Methodology, Formal analysis, Writing - original draft. Zhipeng Li: Data curation, Writing - review & editing, Supervision. Jie Chen: Resources, Supervision. Feng Qiu: Software, Validation. Shaodan Na: Supervision.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgment

This research is supported by grants from the National Science Foundation of China (Grant Number: 51575097), the Fundamental Research Funds for the Central Universities (No.2572019CP04).

References (43)

  • X. D. Zhang, L. Y. Kang, W. F. Diao. The principle of the potentiometer and its applications in the vehicle steering,...
  • H. Yu et al.

    Drive system of dynamic detection equipment for photoelectric encoders using space-vector torque combination

    J. Opt. Precision Eng.

    (2014)
  • Y.F. Wu et al.

    Application of WNN in Ring Ball Gratings Torque Sensor's Signal Demodulation

    J. Appl. Mech. Mater.

    (2012)
  • L. Wang et al.

    A proposed optical-based sensor for assessment of hand movement

    (2015)
  • S. Das et al.

    Design and Realization of an Optical Rotary Sensor

    J. IEEE Sensors J.

    (2018)
  • Y. Liu et al.

    A Highly Reliable Embedded Optical Torque Sensor Based on Flexure Spring

    J. Appl. Bionics Biomech.

    (2018)
  • M. Gasulla et al.

    A contactless capacitive angular-position sensor

    J. IEEE Sensors  Journal

    (2003)
  • F.T. Calkins et al.

    Overview of Magnetostrictive Sensor Technology

    J. J. Intelligent Mater. Syst. Struct.

    (2007)
  • B.P. Reddy et al.

    Low-cost planar coil structure for inductive sensors to measure absolute angular position

    C. International Conference on Frontiers of Sensors Technologies

    (2017)
  • N. Zurauskiene et al.

    Compact Manganite-Graphene Magnetoresistive Sensor

    J. IEEE Magnet. Lett.

    (2019)
  • H. Yu et al.

    Design and electromagnetic analysis of ring-space array torque sensor

    J. Chinese J. Sci. Instrument

    (2009)
  • Cited by (0)

    View full text