Temperature monitoring of milling processes using a directional-spectral thermal radiation heat transfer formulation and thermography

Highlights

• We present a methodology for temperature monitoring during the face milling of metals.

• The methodology is based on directional-spectral radiative heat transfer.

• We used an infrared thermal imager, and we pre-evaluate the spectral emissivity.

• We estimate the temperature during the face milling using the same spectral approach.

• The diffuse-gray model underestimated the temperature, compared to the proposed approach.

Abstract

Temperature is an important property to be monitored during cutting operations, such as in the milling of metals, because it affects the workpiece mechanical properties, the machining tool performance, and the process efficiency. For this reason, authors have used thermography to monitor thermal and spatial gradients, and to estimate temperature. The problem is that IR cameras native applications are mostly based on the diffuse-gray approximation, with emissivity set as a constant value. The diffuse-gray approach is a reasonable choice for opaque surfaces with emissivity approximately constant in respect to the wavelength, which is not the case in most of the time for metals under cutting operation. In this work we propose a more suitable methodology that uses radiative heat transfer directional-spectral relations to estimate cutting process temperatures during face milling of metals, using the full electrical response of commercial infrared cameras as an input. AISI H13 steel was used as the workpiece material. In the first part of the experiments, emissivity was estimated for eight different temperatures from 50 °C to 250 °C, using both directional-spectral and diffuse-gray approaches. Temperature was measured using a thermocouple type T, calibrated using a PT-100 reference from 50 °C to 250 °C. In the second part, we used the emissivity obtained in the first part to estimate the temperature for twelve different cutting conditions, again using both directional-spectral and diffuse-gray approaches. We compared the results of both methods, also with the direct application of emissivity provided in infrared cameras manuals. The temperature deviation between the different approaches were up to 41%, which demonstrates that the temperature estimation procedure affects the results substantially.

Introduction

During the metal cutting, mechanical work is transformed into heat. The force applied against the workpiece causes friction, shear, and strain, ultimately raising the temperature [1]. According to Shaw [2], almost 90% of all mechanical energy becomes thermal energy, during machining. Thermal load affects the machining efficiency [3], [4], tool wear [1], [5], [6], workpiece microstructure, shape, dimension, and roughness accuracy [4], [7], [8], and also the residual stress in the workpiece surface [9], [10]. As a consequence, temperature monitoring is required, although being challenging [11].

Temperature measurement in cutting processes is mostly done with embedded thermocouples and IR systems. Infrared thermography provides visualization of thermal gradients and fast response [12], [13], [14]. However, care must be taken since the accuracy of this technique depends highly on the optical properties assumed to the target object [3], [14], [15]. Surface emissivity depends on temperature, directionality, and spectrum [16], [17], [18]. For metallic surfaces, emissivity is spectral dependent, as a consequence a simple diffuse-gray approach sounds not to be appropriate for IR measurements [16], [19].

Despite the fact that thermography is widely used, emissivity is rarely considered dependent of spectrum, temperature, and direction. Many studies that used thermal imagers for temperature measurement on metal cutting processes considered emissivity as a constant [4], [5], [7], [20], [21], [22]. Emissivity can be estimated through theoretical models, semi-empirical relations, tables, or experimental measurements [14], [15], [17]. Mathematical models based on the electromagnetic wave theory can be used to predict spectral emissivities [16], [23], [24], [25]. To estimate the emissivity experimentally, at least two measurements are needed: (i) the surface temperature, used to estimate the blackbody radiative emission; and (ii) the radiative heat flux emitted by the surface. The ratio between the latter and the first emissive power at a specific wavelength and direction will lead to the directional-spectral emissivity [26]. The use of embedded thermocouples is the most common approach to measure surface temperature, while different types of infrared sensors are used to measure the surface emission: infrared thermometers [26], [27], thermal imagers [28], [29], [30], and spectrometers [31], [32], [33].

In this study, we estimated the spectral emissivity of an AISI H13 steel sample for eight different temperatures from 50 °C to 250 °C, including the impacts of the wavelength on the measurements. We used the spectral emissivity results into a directional-spectral approach to estimate the workpiece temperature, in the interface region during face milling, for twelve different cutting conditions. We did not find in the literature a work that used a similar methodology as we used here.

We divided this article into five sections. The first one is the introduction section. In Section 2, we detail the measurement process with all assumptions made, and the uncertainty analysis performed. Section 3 contains the procedure for the workpiece temperature estimation. We presented our results and findings in Section 4. These article conclusions are in Section 5.