**1. Introduction**

Grinding is a precision abrasive machining process used extensively for producing components with fine tolerances and high surface quality [1]. Among all the variables measured during the grinding process, the surface integrity is one of the most important parameters of any machined surface and is a decisive factor in the evaluation of a successful grinding, especially when high surface integrity requirements need to be met for some applications [2]. The monitoring of the grinding process is very complex because of the high number of influencing parameters, such as the workpiece, the grinding machine, and the process parameters. The control and monitoring of the grinding process allow the improvement of the process performance and the reduction of defects to a possible minimum to guarantee high precision and quality [3,4].

According to Teti et al. [5], the monitoring of machining operations has traditionally been categorized into two approaches: Direct and indirect. In the direct approach, the actual quantity of the

variable, e.g., tool wear, is measured. On the other hand, the indirect approach uses auxiliary quantities, e.g., sensor signals (acoustic emission (AE), power, vibration, and force), to estimate a variable quantity. According to [6], among all the signals of the mentioned sensors, the acoustic emission (AE) is considered the most sensitive signal because its frequency range is beyond mechanical vibrations and electrical noises and, therefore, these noises can be easily filtered. According to Lopes et al. [7], the acoustic emission signal is defined as the transient elastic waves generated by the rapid release of energy within a material. Regarding this subject, AE techniques have been widely employed to monitor engineering applications [8–12]. For example, in He et al. [13], the AE method for detecting metallic grinding burn was presented as a nondestructive detection method. Lopes et al. [14] studied the influence of the temperature on the frequency response of an AE sensor. In Badger et al. [15], the fundamental relationships between the AE signal and the dressing variables were proposed.

Recently, piezoelectric diaphragms of lead zirconate titanate (PZT) have become popular due to the fact that it has been successfully used in many scientific applications in view of its low cost and excellent sensing capacities [16,17]. The PZT diaphragms have a simple construction consisting of a brass plate on which a ceramic disc is fixed. In the past, these acoustic components were only used for sound generation (buzzers and telephone receivers). However, the considerable potential of these transducers makes them attractive to other applications besides producing sound [18]. In this context, PZT diaphragms have been successfully used in the areas of structural health monitoring (SHM) to detect structural damages in engineering projects by signals related to the piezoelectric effect [19]. In addition, they have also been used as ultrasonic actuators [20], as the active element in acoustic position encoder [21], and as an acoustic sensor for partial discharge monitoring in power transformers [22].

The present work differs from other studies because it proposes a non-invasive and low-cost monitoring technique of the ceramic grinding process through the implementation of the PZT diaphragm and digital signal processing in the time–frequency domain, which allows a better interpretation of the results when compared to the traditional techniques. An important aspect of this research work is the use of the PZT diaphragm, which is a low-cost sensor compared to sensors such as acoustic emission, dynamometer, accelerometer, and power. PZT diaphragms can operate in both active and passive configurations and have an average cost of a few cents versus the high average cost of AE sensors, which range from hundreds to thousands of dollars [23]. In addition, they are compact, flexible, lightweight, and simple acoustic components widely used in various electronic devices to produce sound (alarm, ringing, and beep) [18,24]. Some preliminary applications of the PZT diaphragm emerged in the grinding field focusing on workpiece surface integrity and tool condition monitoring. For example, in Batista da Silva et al.'s study [25], the electromechanical impedance (EMI) method with low-cost PZT diaphragms were used to monitor the surface damage of ground steel workpieces. Microhardness and roughness measurements were compared with the results of the proposed technique. In Marchi et al. [26], two PZT diaphragms were used to detect the workpiece wear by means of the EMI method. The workpiece wear was correlated to the calculated damage statistics. Ribeiro et al. [27] proposed a new technique for monitoring the surface burning on steel workpieces using the PZT diaphragm and two types of grinding wheels (cubic boron nitride and aluminum oxide). An acoustic emission sensor was used to verify the efficiency of the PZT diaphragm. In Junior et al. [28], an approach for monitoring the dressing operation by means of PZT diaphragm-based impedance was proposed. The authors validated the proposed approach based on artificial neural networks (ANN), which selected the most damage-sensitive features based on the optimal frequency band.

The present work is an expansion of the research activities described in Viera et al. [16,29], where initial results were presented. Therefore, the scope of this research work is to illustrate a new broader approach for the real-time monitoring system of the surface quality of ground ceramics using the low-cost PZT diaphragm. In this context, the results of the time–frequency analysis, obtained through the short-time Fourier transform (STFT), are used as the basis for the computation of the ratio of power (ROP) metric, which is considered an important metric in the frequency domain. However, in the present work, the ROP was computed in the time domain based on the study of Thomazella et al. [30], who conducted a pioneer work with the time–frequency approach along with the time domain ROP for monitoring self-vibrations on ground steel workpieces. The application of the technique in the monitoring of the ceramic grinding process in order to estimate the workpiece surface quality represents a novel approach for the manufacturing field. Moreover, this work differs from [30] because of the use of the low-cost piezoelectric diaphragm instead of the accelerometer. In addition, in order to verify the effectiveness and reliability of the proposed approach, an acoustic emission (AE) sensor, consolidated in the monitoring of manufacturing processes, was used. This paper is organized as follows: A ceramic grinding overview is presented in Section 2. Sections 3 and 4 describe the piezoelectric diaphragms and the signal processing techniques used in this study. The experimental setup is shown in Section 5. Section 6 presents the results and discussion. Finally, the conclusions are presented in Section 7.
